From 4bdafb9ddfa4b3d970e2194d00e1c6d5002f513f Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 26 Sep 2018 21:12:22 +0100 Subject: drm/i915: Remove i915.enable_ppgtt override Now that we are confident in providing full-ppgtt where supported, remove the ability to override the context isolation. v2: Remove faked aliasing-ppgtt for testing as it no longer is accepted. v3: s/USES/HAS/ to match usage and reject attempts to load the module on old GVT-g setups that do not provide support for full-ppgtt. v4: Insulate ABI ppGTT values from our internal enum (later plans involve moving ppGTT depth out of the enum, thus potentially breaking ABI unless we document the current values). Signed-off-by: Chris Wilson Cc: Joonas Lahtinen Cc: Matthew Auld Reviewed-by: Joonas Lahtinen Acked-by: Zhi Wang Link: https://patchwork.freedesktop.org/patch/msgid/20180926201222.5643-1-chris@chris-wilson.co.uk --- include/uapi/drm/i915_drm.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'include') diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h index a4446f452040..298b2e197744 100644 --- a/include/uapi/drm/i915_drm.h +++ b/include/uapi/drm/i915_drm.h @@ -412,6 +412,14 @@ typedef struct drm_i915_irq_wait { int irq_seq; } drm_i915_irq_wait_t; +/* + * Different modes of per-process Graphics Translation Table, + * see I915_PARAM_HAS_ALIASING_PPGTT + */ +#define I915_GEM_PPGTT_NONE 0 +#define I915_GEM_PPGTT_ALIASING 1 +#define I915_GEM_PPGTT_FULL 2 + /* Ioctl to query kernel params: */ #define I915_PARAM_IRQ_ACTIVE 1 -- cgit v1.2.3 From c1c8f6fa731bc850d571d9c32016c3bf9ed1332e Mon Sep 17 00:00:00 2001 From: Rodrigo Vivi Date: Mon, 24 Sep 2018 16:43:12 -0700 Subject: drm/i915: Redefine some Whiskey Lake SKUs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit 'b9be78531d27 ("drm/i915/whl: Introducing Whiskey Lake platform")' introduced WHL by moving some of CFL IDs here and using the Spec information of "U43" for most of IDs what appeared to be GT3. However when propagating the change to Mesa, Lionel noticed that based on number of execution unities the classification here seems at least strange. So, let's move for now with the information we trust more: the number of EUs. So we are able to propagate this change across the stack without getting stuck forever. Reference: https://patchwork.freedesktop.org/patch/246695/ Fixes: b9be78531d27 ("drm/i915/whl: Introducing Whiskey Lake platform") Cc: Lionel Landwerlin Cc: José Roberto de Souza Cc: David Airlie Signed-off-by: Rodrigo Vivi Reviewed-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20180924234312.15017-1-rodrigo.vivi@intel.com --- include/drm/i915_pciids.h | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/drm/i915_pciids.h b/include/drm/i915_pciids.h index fd965ffbb92e..c8d3d541ad01 100644 --- a/include/drm/i915_pciids.h +++ b/include/drm/i915_pciids.h @@ -407,17 +407,17 @@ /* WHL/CFL U GT1 */ #define INTEL_WHL_U_GT1_IDS(info) \ - INTEL_VGA_DEVICE(0x3EA1, info) + INTEL_VGA_DEVICE(0x3EA1, info), \ + INTEL_VGA_DEVICE(0x3EA4, info) /* WHL/CFL U GT2 */ #define INTEL_WHL_U_GT2_IDS(info) \ - INTEL_VGA_DEVICE(0x3EA0, info) + INTEL_VGA_DEVICE(0x3EA0, info), \ + INTEL_VGA_DEVICE(0x3EA3, info) /* WHL/CFL U GT3 */ #define INTEL_WHL_U_GT3_IDS(info) \ - INTEL_VGA_DEVICE(0x3EA2, info), \ - INTEL_VGA_DEVICE(0x3EA3, info), \ - INTEL_VGA_DEVICE(0x3EA4, info) + INTEL_VGA_DEVICE(0x3EA2, info) #define INTEL_CFL_IDS(info) \ INTEL_CFL_S_GT1_IDS(info), \ -- cgit v1.2.3 From c0c46ca461f136a0ae1ed69da6c874e850aeeb53 Mon Sep 17 00:00:00 2001 From: José Roberto de Souza Date: Wed, 26 Sep 2018 18:06:50 -0700 Subject: drm/i915/aml: Add new Amber Lake PCI ID MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This new AML PCI ID uses the same gen graphics as Coffe Lake not a Kaby Lake one like the other AMLs. So to make it more explicit renaming INTEL_AML_GT2_IDS to INTEL_AML_KBL_GT2_IDS and naming this id as INTEL_AML_CFL_GT2_IDS. v2: - missed add new AML macro to INTEL_CFL_IDS() - added derivated platform initials to AML macros Reviewed-by: Rodrigo Vivi Signed-off-by: José Roberto de Souza Signed-off-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20180927010650.22731-1-jose.souza@intel.com --- drivers/gpu/drm/i915/i915_pci.c | 3 ++- include/drm/i915_pciids.h | 11 ++++++++--- 2 files changed, 10 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c index 9ddd2db906ce..0a05cc7ace14 100644 --- a/drivers/gpu/drm/i915/i915_pci.c +++ b/drivers/gpu/drm/i915/i915_pci.c @@ -660,7 +660,7 @@ static const struct pci_device_id pciidlist[] = { INTEL_KBL_GT2_IDS(&intel_kabylake_gt2_info), INTEL_KBL_GT3_IDS(&intel_kabylake_gt3_info), INTEL_KBL_GT4_IDS(&intel_kabylake_gt3_info), - INTEL_AML_GT2_IDS(&intel_kabylake_gt2_info), + INTEL_AML_KBL_GT2_IDS(&intel_kabylake_gt2_info), INTEL_CFL_S_GT1_IDS(&intel_coffeelake_gt1_info), INTEL_CFL_S_GT2_IDS(&intel_coffeelake_gt2_info), INTEL_CFL_H_GT2_IDS(&intel_coffeelake_gt2_info), @@ -668,6 +668,7 @@ static const struct pci_device_id pciidlist[] = { INTEL_CFL_U_GT3_IDS(&intel_coffeelake_gt3_info), INTEL_WHL_U_GT1_IDS(&intel_coffeelake_gt1_info), INTEL_WHL_U_GT2_IDS(&intel_coffeelake_gt2_info), + INTEL_AML_CFL_GT2_IDS(&intel_coffeelake_gt2_info), INTEL_WHL_U_GT3_IDS(&intel_coffeelake_gt3_info), INTEL_CNL_IDS(&intel_cannonlake_info), INTEL_ICL_11_IDS(&intel_icelake_11_info), diff --git a/include/drm/i915_pciids.h b/include/drm/i915_pciids.h index c8d3d541ad01..192667144693 100644 --- a/include/drm/i915_pciids.h +++ b/include/drm/i915_pciids.h @@ -365,16 +365,20 @@ INTEL_VGA_DEVICE(0x593B, info) /* Halo GT4 */ /* AML/KBL Y GT2 */ -#define INTEL_AML_GT2_IDS(info) \ +#define INTEL_AML_KBL_GT2_IDS(info) \ INTEL_VGA_DEVICE(0x591C, info), /* ULX GT2 */ \ INTEL_VGA_DEVICE(0x87C0, info) /* ULX GT2 */ +/* AML/CFL Y GT2 */ +#define INTEL_AML_CFL_GT2_IDS(info) \ + INTEL_VGA_DEVICE(0x87CA, info) + #define INTEL_KBL_IDS(info) \ INTEL_KBL_GT1_IDS(info), \ INTEL_KBL_GT2_IDS(info), \ INTEL_KBL_GT3_IDS(info), \ INTEL_KBL_GT4_IDS(info), \ - INTEL_AML_GT2_IDS(info) + INTEL_AML_KBL_GT2_IDS(info) /* CFL S */ #define INTEL_CFL_S_GT1_IDS(info) \ @@ -427,7 +431,8 @@ INTEL_CFL_U_GT3_IDS(info), \ INTEL_WHL_U_GT1_IDS(info), \ INTEL_WHL_U_GT2_IDS(info), \ - INTEL_WHL_U_GT3_IDS(info) + INTEL_WHL_U_GT3_IDS(info), \ + INTEL_AML_CFL_GT2_IDS(info) /* CNL */ #define INTEL_CNL_IDS(info) \ -- cgit v1.2.3 From 39b50c603878f4f8ae541ac4088a805d588abc79 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Tue, 16 Oct 2018 16:39:46 -0400 Subject: drm/atomic_helper: Stop modesets on unregistered connectors harder MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Unfortunately, it appears our fix in: commit b5d29843d8ef ("drm/atomic_helper: Allow DPMS On<->Off changes for unregistered connectors") Which attempted to work around the problems introduced by: commit 4d80273976bf ("drm/atomic_helper: Disallow new modesets on unregistered connectors") Is still not the right solution, as modesets can still be triggered outside of drm_atomic_set_crtc_for_connector(). So in order to fix this, while still being careful that we don't break modesets that a driver may perform before being registered with userspace, we replace connector->registered with a tristate member, connector->registration_state. This allows us to keep track of whether or not a connector is still initializing and hasn't been exposed to userspace, is currently registered and exposed to userspace, or has been legitimately removed from the system after having once been present. Using this info, we can prevent userspace from performing new modesets on unregistered connectors while still allowing the driver to perform modesets on unregistered connectors before the driver has finished being registered. Changes since v1: - Fix WARN_ON() in drm_connector_cleanup() that CI caught with this patchset in igt@drv_module_reload@basic-reload-inject and igt@drv_module_reload@basic-reload by checking if the connector is registered instead of unregistered, as calling drm_connector_cleanup() on a connector that hasn't been registered with userspace yet should stay valid. - Remove unregistered_connector_check(), and just go back to what we were doing before in commit 4d80273976bf ("drm/atomic_helper: Disallow new modesets on unregistered connectors") except replacing READ_ONCE(connector->registered) with drm_connector_is_unregistered(). This gets rid of the behavior of allowing DPMS On<->Off, but that should be fine as it's more consistent with the UAPI we had before - danvet - s/drm_connector_unregistered/drm_connector_is_unregistered/ - danvet - Update documentation, fix some typos. Fixes: b5d29843d8ef ("drm/atomic_helper: Allow DPMS On<->Off changes for unregistered connectors") Cc: Ville Syrjälä Cc: Daniel Vetter Cc: Rodrigo Vivi Cc: stable@vger.kernel.org Cc: David Airlie Signed-off-by: Lyude Paul Reviewed-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/20181016203946.9601-1-lyude@redhat.com --- drivers/gpu/drm/drm_atomic_helper.c | 21 ++++++++++- drivers/gpu/drm/drm_atomic_uapi.c | 21 ----------- drivers/gpu/drm/drm_connector.c | 11 +++--- drivers/gpu/drm/i915/intel_dp_mst.c | 8 ++--- include/drm/drm_connector.h | 71 +++++++++++++++++++++++++++++++++++-- 5 files changed, 99 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 3cf1aa132778..4225b1ca407b 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -308,6 +308,26 @@ update_connector_routing(struct drm_atomic_state *state, return 0; } + crtc_state = drm_atomic_get_new_crtc_state(state, + new_connector_state->crtc); + /* + * For compatibility with legacy users, we want to make sure that + * we allow DPMS On->Off modesets on unregistered connectors. Modesets + * which would result in anything else must be considered invalid, to + * avoid turning on new displays on dead connectors. + * + * Since the connector can be unregistered at any point during an + * atomic check or commit, this is racy. But that's OK: all we care + * about is ensuring that userspace can't do anything but shut off the + * display on a connector that was destroyed after its been notified, + * not before. + */ + if (drm_connector_is_unregistered(connector) && crtc_state->active) { + DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] is not registered\n", + connector->base.id, connector->name); + return -EINVAL; + } + funcs = connector->helper_private; if (funcs->atomic_best_encoder) @@ -352,7 +372,6 @@ update_connector_routing(struct drm_atomic_state *state, set_best_encoder(state, new_connector_state, new_encoder); - crtc_state = drm_atomic_get_new_crtc_state(state, new_connector_state->crtc); crtc_state->connectors_changed = true; DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] using [ENCODER:%d:%s] on [CRTC:%d:%s]\n", diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c index a22d6f269b07..d5b7f315098c 100644 --- a/drivers/gpu/drm/drm_atomic_uapi.c +++ b/drivers/gpu/drm/drm_atomic_uapi.c @@ -299,27 +299,6 @@ drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state, struct drm_connector *connector = conn_state->connector; struct drm_crtc_state *crtc_state; - /* - * For compatibility with legacy users, we want to make sure that - * we allow DPMS On<->Off modesets on unregistered connectors, since - * legacy modesetting users will not be expecting these to fail. We do - * not however, want to allow legacy users to assign a connector - * that's been unregistered from sysfs to another CRTC, since doing - * this with a now non-existent connector could potentially leave us - * in an invalid state. - * - * Since the connector can be unregistered at any point during an - * atomic check or commit, this is racy. But that's OK: all we care - * about is ensuring that userspace can't use this connector for new - * configurations after it's been notified that the connector is no - * longer present. - */ - if (!READ_ONCE(connector->registered) && crtc) { - DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] is not registered\n", - connector->base.id, connector->name); - return -EINVAL; - } - if (conn_state->crtc == crtc) return 0; diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c index 1e40e5decbe9..4943cef178be 100644 --- a/drivers/gpu/drm/drm_connector.c +++ b/drivers/gpu/drm/drm_connector.c @@ -379,7 +379,8 @@ void drm_connector_cleanup(struct drm_connector *connector) /* The connector should have been removed from userspace long before * it is finally destroyed. */ - if (WARN_ON(connector->registered)) + if (WARN_ON(connector->registration_state == + DRM_CONNECTOR_REGISTERED)) drm_connector_unregister(connector); if (connector->tile_group) { @@ -436,7 +437,7 @@ int drm_connector_register(struct drm_connector *connector) return 0; mutex_lock(&connector->mutex); - if (connector->registered) + if (connector->registration_state != DRM_CONNECTOR_INITIALIZING) goto unlock; ret = drm_sysfs_connector_add(connector); @@ -456,7 +457,7 @@ int drm_connector_register(struct drm_connector *connector) drm_mode_object_register(connector->dev, &connector->base); - connector->registered = true; + connector->registration_state = DRM_CONNECTOR_REGISTERED; goto unlock; err_debugfs: @@ -478,7 +479,7 @@ EXPORT_SYMBOL(drm_connector_register); void drm_connector_unregister(struct drm_connector *connector) { mutex_lock(&connector->mutex); - if (!connector->registered) { + if (connector->registration_state != DRM_CONNECTOR_REGISTERED) { mutex_unlock(&connector->mutex); return; } @@ -489,7 +490,7 @@ void drm_connector_unregister(struct drm_connector *connector) drm_sysfs_connector_remove(connector); drm_debugfs_connector_remove(connector); - connector->registered = false; + connector->registration_state = DRM_CONNECTOR_UNREGISTERED; mutex_unlock(&connector->mutex); } EXPORT_SYMBOL(drm_connector_unregister); diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index b268bdd71bd3..8b71d64ebd9d 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -78,7 +78,7 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, pipe_config->pbn = mst_pbn; /* Zombie connectors can't have VCPI slots */ - if (READ_ONCE(connector->registered)) { + if (!drm_connector_is_unregistered(connector)) { slots = drm_dp_atomic_find_vcpi_slots(state, &intel_dp->mst_mgr, port, @@ -314,7 +314,7 @@ static int intel_dp_mst_get_ddc_modes(struct drm_connector *connector) struct edid *edid; int ret; - if (!READ_ONCE(connector->registered)) + if (drm_connector_is_unregistered(connector)) return intel_connector_update_modes(connector, NULL); edid = drm_dp_mst_get_edid(connector, &intel_dp->mst_mgr, intel_connector->port); @@ -330,7 +330,7 @@ intel_dp_mst_detect(struct drm_connector *connector, bool force) struct intel_connector *intel_connector = to_intel_connector(connector); struct intel_dp *intel_dp = intel_connector->mst_port; - if (!READ_ONCE(connector->registered)) + if (drm_connector_is_unregistered(connector)) return connector_status_disconnected; return drm_dp_mst_detect_port(connector, &intel_dp->mst_mgr, intel_connector->port); @@ -361,7 +361,7 @@ intel_dp_mst_mode_valid(struct drm_connector *connector, int bpp = 24; /* MST uses fixed bpp */ int max_rate, mode_rate, max_lanes, max_link_clock; - if (!READ_ONCE(connector->registered)) + if (drm_connector_is_unregistered(connector)) return MODE_ERROR; if (mode->flags & DRM_MODE_FLAG_DBLSCAN) diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h index 91a877fa00cb..9ccad6b062f2 100644 --- a/include/drm/drm_connector.h +++ b/include/drm/drm_connector.h @@ -82,6 +82,53 @@ enum drm_connector_status { connector_status_unknown = 3, }; +/** + * enum drm_connector_registration_status - userspace registration status for + * a &drm_connector + * + * This enum is used to track the status of initializing a connector and + * registering it with userspace, so that DRM can prevent bogus modesets on + * connectors that no longer exist. + */ +enum drm_connector_registration_state { + /** + * @DRM_CONNECTOR_INITIALIZING: The connector has just been created, + * but has yet to be exposed to userspace. There should be no + * additional restrictions to how the state of this connector may be + * modified. + */ + DRM_CONNECTOR_INITIALIZING = 0, + + /** + * @DRM_CONNECTOR_REGISTERED: The connector has been fully initialized + * and registered with sysfs, as such it has been exposed to + * userspace. There should be no additional restrictions to how the + * state of this connector may be modified. + */ + DRM_CONNECTOR_REGISTERED = 1, + + /** + * @DRM_CONNECTOR_UNREGISTERED: The connector has either been exposed + * to userspace and has since been unregistered and removed from + * userspace, or the connector was unregistered before it had a chance + * to be exposed to userspace (e.g. still in the + * @DRM_CONNECTOR_INITIALIZING state). When a connector is + * unregistered, there are additional restrictions to how its state + * may be modified: + * + * - An unregistered connector may only have its DPMS changed from + * On->Off. Once DPMS is changed to Off, it may not be switched back + * to On. + * - Modesets are not allowed on unregistered connectors, unless they + * would result in disabling its assigned CRTCs. This means + * disabling a CRTC on an unregistered connector is OK, but enabling + * one is not. + * - Removing a CRTC from an unregistered connector is OK, but new + * CRTCs may never be assigned to an unregistered connector. + */ + DRM_CONNECTOR_UNREGISTERED = 2, +}; + enum subpixel_order { SubPixelUnknown = 0, SubPixelHorizontalRGB, @@ -853,10 +900,12 @@ struct drm_connector { bool ycbcr_420_allowed; /** - * @registered: Is this connector exposed (registered) with userspace? + * @registration_state: Is this connector initializing, exposed + * (registered) with userspace, or unregistered? + * * Protected by @mutex. */ - bool registered; + enum drm_connector_registration_state registration_state; /** * @modes: @@ -1166,6 +1215,24 @@ static inline void drm_connector_unreference(struct drm_connector *connector) drm_connector_put(connector); } +/** + * drm_connector_is_unregistered - has the connector been unregistered from + * userspace? + * @connector: DRM connector + * + * Checks whether or not @connector has been unregistered from userspace. + * + * Returns: + * True if the connector was unregistered, false if the connector is + * registered or has not yet been registered with userspace. + */ +static inline bool +drm_connector_is_unregistered(struct drm_connector *connector) +{ + return READ_ONCE(connector->registration_state) == + DRM_CONNECTOR_UNREGISTERED; +} + const char *drm_get_connector_status_name(enum drm_connector_status status); const char *drm_get_subpixel_order_name(enum subpixel_order order); const char *drm_get_dpms_name(int val); -- cgit v1.2.3 From cd956bfcd0f58d20485ac0a785415f7d9327a95f Mon Sep 17 00:00:00 2001 From: Lionel Landwerlin Date: Tue, 23 Oct 2018 11:07:07 +0100 Subject: drm/i915/perf: add a parameter to control the size of OA buffer The way our hardware is designed doesn't seem to let us use the MI_RECORD_PERF_COUNT command without setting up a circular buffer. In the case where the user didn't request OA reports to be available through the i915 perf stream, we can set the OA buffer to the minimum size to avoid consuming memory which won't be used by the driver. v2: Simplify oa buffer size exponent selection (Chris) Reuse vma size field (Lionel) v3: Restrict size opening parameter to values supported by HW (Chris) v4: Drop out of date comment (Matt) Add debug message when buffer size is rejected (Matt) Signed-off-by: Lionel Landwerlin Reviewed-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20181023100707.31738-5-lionel.g.landwerlin@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/i915_perf.c | 99 ++++++++++++++++++++++++++-------------- drivers/gpu/drm/i915/i915_reg.h | 2 + include/uapi/drm/i915_drm.h | 7 +++ 4 files changed, 76 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 135a8522a803..22cf3c75558c 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2006,6 +2006,7 @@ struct drm_i915_private { u32 last_ctx_id; int format; int format_size; + int size_exponent; /** * Locks reads and writes to all head/tail state diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 4529edfdcfc8..1712b68de8f5 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -212,13 +212,7 @@ #include "i915_oa_icl.h" #include "intel_lrc_reg.h" -/* HW requires this to be a power of two, between 128k and 16M, though driver - * is currently generally designed assuming the largest 16M size is used such - * that the overflow cases are unlikely in normal operation. - */ -#define OA_BUFFER_SIZE SZ_16M - -#define OA_TAKEN(tail, head) ((tail - head) & (OA_BUFFER_SIZE - 1)) +#define OA_TAKEN(tail, head) (((tail) - (head)) & (dev_priv->perf.oa.oa_buffer.vma->size - 1)) /** * DOC: OA Tail Pointer Race @@ -361,6 +355,7 @@ struct perf_open_properties { int oa_format; bool oa_periodic; int oa_period_exponent; + u32 oa_buffer_size_exponent; }; static void free_oa_config(struct drm_i915_private *dev_priv, @@ -523,7 +518,7 @@ static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) * could put the tail out of bounds... */ if (hw_tail >= gtt_offset && - hw_tail < (gtt_offset + OA_BUFFER_SIZE)) { + hw_tail < (gtt_offset + dev_priv->perf.oa.oa_buffer.vma->size)) { dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = aging_tail = hw_tail; dev_priv->perf.oa.oa_buffer.aging_timestamp = now; @@ -652,7 +647,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, int report_size = dev_priv->perf.oa.oa_buffer.format_size; u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr; u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); - u32 mask = (OA_BUFFER_SIZE - 1); + u32 mask = (dev_priv->perf.oa.oa_buffer.vma->size - 1); size_t start_offset = *offset; unsigned long flags; unsigned int aged_tail_idx; @@ -692,8 +687,8 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, * only be incremented by multiples of the report size (notably also * all a power of two). */ - if (WARN_ONCE(head > OA_BUFFER_SIZE || head % report_size || - tail > OA_BUFFER_SIZE || tail % report_size, + if (WARN_ONCE(head > dev_priv->perf.oa.oa_buffer.vma->size || head % report_size || + tail > dev_priv->perf.oa.oa_buffer.vma->size || tail % report_size, "Inconsistent OA buffer pointers: head = %u, tail = %u\n", head, tail)) return -EIO; @@ -716,7 +711,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, * here would imply a driver bug that would result * in an overrun. */ - if (WARN_ON((OA_BUFFER_SIZE - head) < report_size)) { + if (WARN_ON((dev_priv->perf.oa.oa_buffer.vma->size - head) < report_size)) { DRM_ERROR("Spurious OA head ptr: non-integral report offset\n"); break; } @@ -875,11 +870,6 @@ static int gen8_oa_read(struct i915_perf_stream *stream, * automatically triggered reports in this condition and so we * have to assume that old reports are now being trampled * over. - * - * Considering how we don't currently give userspace control - * over the OA buffer size and always configure a large 16MB - * buffer, then a buffer overflow does anyway likely indicate - * that something has gone quite badly wrong. */ if (oastatus & GEN8_OASTATUS_OABUFFER_OVERFLOW) { ret = append_oa_status(stream, buf, count, offset, @@ -941,7 +931,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, int report_size = dev_priv->perf.oa.oa_buffer.format_size; u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr; u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); - u32 mask = (OA_BUFFER_SIZE - 1); + u32 mask = (dev_priv->perf.oa.oa_buffer.vma->size - 1); size_t start_offset = *offset; unsigned long flags; unsigned int aged_tail_idx; @@ -978,8 +968,8 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, * only be incremented by multiples of the report size (notably also * all a power of two). */ - if (WARN_ONCE(head > OA_BUFFER_SIZE || head % report_size || - tail > OA_BUFFER_SIZE || tail % report_size, + if (WARN_ONCE(head > dev_priv->perf.oa.oa_buffer.vma->size || head % report_size || + tail > dev_priv->perf.oa.oa_buffer.vma->size || tail % report_size, "Inconsistent OA buffer pointers: head = %u, tail = %u\n", head, tail)) return -EIO; @@ -999,7 +989,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, * here would imply a driver bug that would result * in an overrun. */ - if (WARN_ON((OA_BUFFER_SIZE - head) < report_size)) { + if (WARN_ON((dev_priv->perf.oa.oa_buffer.vma->size - head) < report_size)) { DRM_ERROR("Spurious OA head ptr: non-integral report offset\n"); break; } @@ -1394,7 +1384,9 @@ static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv) I915_WRITE(GEN7_OABUFFER, gtt_offset); - I915_WRITE(GEN7_OASTATUS1, gtt_offset | OABUFFER_SIZE_16M); /* tail */ + I915_WRITE(GEN7_OASTATUS1, gtt_offset | + ((dev_priv->perf.oa.oa_buffer.size_exponent - 17) << + GEN7_OASTATUS1_BUFFER_SIZE_SHIFT)); /* tail */ /* Mark that we need updated tail pointers to read from... */ dev_priv->perf.oa.oa_buffer.tails[0].offset = INVALID_TAIL_PTR; @@ -1419,7 +1411,8 @@ static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv) * the assumption that new reports are being written to zeroed * memory... */ - memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, OA_BUFFER_SIZE); + memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, + dev_priv->perf.oa.oa_buffer.vma->size); /* Maybe make ->pollin per-stream state if we support multiple * concurrent streams in the future. @@ -1449,7 +1442,9 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) * bit." */ I915_WRITE(GEN8_OABUFFER, gtt_offset | - OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT); + ((dev_priv->perf.oa.oa_buffer.size_exponent - 17) << + GEN8_OABUFFER_BUFFER_SIZE_SHIFT) | + GEN8_OABUFFER_MEM_SELECT_GGTT); I915_WRITE(GEN8_OATAILPTR, gtt_offset & GEN8_OATAILPTR_MASK); /* Mark that we need updated tail pointers to read from... */ @@ -1477,7 +1472,8 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) * the assumption that new reports are being written to zeroed * memory... */ - memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, OA_BUFFER_SIZE); + memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, + dev_priv->perf.oa.oa_buffer.vma->size); /* * Maybe make ->pollin per-stream state if we support multiple @@ -1486,23 +1482,24 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) dev_priv->perf.oa.pollin = false; } -static int alloc_oa_buffer(struct drm_i915_private *dev_priv) +static int alloc_oa_buffer(struct drm_i915_private *dev_priv, int size_exponent) { struct drm_i915_gem_object *bo; struct i915_vma *vma; + size_t size = 1U << size_exponent; int ret; if (WARN_ON(dev_priv->perf.oa.oa_buffer.vma)) return -ENODEV; + if (WARN_ON(size < SZ_128K || size > SZ_16M)) + return -EINVAL; + ret = i915_mutex_lock_interruptible(&dev_priv->drm); if (ret) return ret; - BUILD_BUG_ON_NOT_POWER_OF_2(OA_BUFFER_SIZE); - BUILD_BUG_ON(OA_BUFFER_SIZE < SZ_128K || OA_BUFFER_SIZE > SZ_16M); - - bo = i915_gem_object_create(dev_priv, OA_BUFFER_SIZE); + bo = i915_gem_object_create(dev_priv, size); if (IS_ERR(bo)) { DRM_ERROR("Failed to allocate OA buffer\n"); ret = PTR_ERR(bo); @@ -1520,6 +1517,7 @@ static int alloc_oa_buffer(struct drm_i915_private *dev_priv) goto err_unref; } dev_priv->perf.oa.oa_buffer.vma = vma; + dev_priv->perf.oa.oa_buffer.size_exponent = size_exponent; dev_priv->perf.oa.oa_buffer.vaddr = i915_gem_object_pin_map(bo, I915_MAP_WB); @@ -1528,9 +1526,10 @@ static int alloc_oa_buffer(struct drm_i915_private *dev_priv) goto err_unpin; } - DRM_DEBUG_DRIVER("OA Buffer initialized, gtt offset = 0x%x, vaddr = %p\n", + DRM_DEBUG_DRIVER("OA Buffer initialized, gtt offset = 0x%x, vaddr = %p, size = %llu\n", i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma), - dev_priv->perf.oa.oa_buffer.vaddr); + dev_priv->perf.oa.oa_buffer.vaddr, + dev_priv->perf.oa.oa_buffer.vma->size); goto unlock; @@ -2090,7 +2089,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, intel_runtime_pm_get(dev_priv); intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); - ret = alloc_oa_buffer(dev_priv); + ret = alloc_oa_buffer(dev_priv, props->oa_buffer_size_exponent); if (ret) goto err_oa_buf_alloc; @@ -2649,6 +2648,26 @@ static u64 oa_exponent_to_ns(struct drm_i915_private *dev_priv, int exponent) 1000ULL * INTEL_INFO(dev_priv)->cs_timestamp_frequency_khz); } +static int +select_oa_buffer_exponent(struct drm_i915_private *i915, + u64 requested_size) +{ + int order; + + /* + * When no size is specified, use the largest size supported by all + * generations. + */ + if (!requested_size) + return order_base_2(SZ_16M); + + order = order_base_2(clamp_t(u64, requested_size, SZ_128K, SZ_16M)); + if (requested_size != (1UL << order)) + return -EINVAL; + + return order; +} + /** * read_properties_unlocked - validate + copy userspace stream open properties * @dev_priv: i915 device instance @@ -2776,6 +2795,14 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, props->oa_periodic = true; props->oa_period_exponent = value; break; + case DRM_I915_PERF_PROP_OA_BUFFER_SIZE: + ret = select_oa_buffer_exponent(dev_priv, value); + if (ret < 0) { + DRM_DEBUG("OA buffer size invalid %llu\n", value); + return ret; + } + props->oa_buffer_size_exponent = ret; + break; case DRM_I915_PERF_PROP_MAX: MISSING_CASE(id); return -EINVAL; @@ -2784,6 +2811,12 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, uprop += 2; } + /* If no buffer size was requested, select the default one. */ + if (!props->oa_buffer_size_exponent) { + props->oa_buffer_size_exponent = + select_oa_buffer_exponent(dev_priv, 0); + } + return 0; } diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 8bd61f946714..4f9c2fe51f27 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -569,12 +569,14 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg) #define GEN8_OABUFFER_UDW _MMIO(0x23b4) #define GEN8_OABUFFER _MMIO(0x2b14) #define GEN8_OABUFFER_MEM_SELECT_GGTT (1 << 0) /* 0: PPGTT, 1: GGTT */ +#define GEN8_OABUFFER_BUFFER_SIZE_SHIFT 3 #define GEN7_OASTATUS1 _MMIO(0x2364) #define GEN7_OASTATUS1_TAIL_MASK 0xffffffc0 #define GEN7_OASTATUS1_COUNTER_OVERFLOW (1 << 2) #define GEN7_OASTATUS1_OABUFFER_OVERFLOW (1 << 1) #define GEN7_OASTATUS1_REPORT_LOST (1 << 0) +#define GEN7_OASTATUS1_BUFFER_SIZE_SHIFT 3 #define GEN7_OASTATUS2 _MMIO(0x2368) #define GEN7_OASTATUS2_HEAD_MASK 0xffffffc0 diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h index 298b2e197744..e477ef8c644e 100644 --- a/include/uapi/drm/i915_drm.h +++ b/include/uapi/drm/i915_drm.h @@ -1540,6 +1540,13 @@ enum drm_i915_perf_property_id { */ DRM_I915_PERF_PROP_OA_EXPONENT, + /** + * Specify a global OA buffer size to be allocated in bytes. The size + * specified must be supported by HW (currently supported sizes are + * powers of 2 ranging from 128Kb to 16Mb). + */ + DRM_I915_PERF_PROP_OA_BUFFER_SIZE, + DRM_I915_PERF_PROP_MAX /* non-ABI */ }; -- cgit v1.2.3 From 5a8de47b3c250521dd632cdedaac6db88367defa Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Wed, 24 Oct 2018 13:54:03 +0200 Subject: netfilter: bridge: define INT_MIN & INT_MAX in userspace MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit With 4.19, programs like ebtables fail to build when they include "linux/netfilter_bridge.h". It is caused by commit 94276fa8a2a4 which added a use of INT_MIN and INT_MAX to the header: : In file included from /usr/include/linux/netfilter_bridge/ebtables.h:18, : from include/ebtables_u.h:28, : from communication.c:23: : /usr/include/linux/netfilter_bridge.h:30:20: error: 'INT_MIN' undeclared here (not in a function) : NF_BR_PRI_FIRST = INT_MIN, : ^~~~~~~ Define these constants by including "limits.h" when !__KERNEL__ (the same way as for other netfilter_* headers). Fixes: 94276fa8a2a4 ("netfilter: bridge: Expose nf_tables bridge hook priorities through uapi") Signed-off-by: Jiri Slaby Acked-by: Máté Eckl Signed-off-by: Pablo Neira Ayuso --- include/uapi/linux/netfilter_bridge.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include') diff --git a/include/uapi/linux/netfilter_bridge.h b/include/uapi/linux/netfilter_bridge.h index 156ccd089df1..1610fdbab98d 100644 --- a/include/uapi/linux/netfilter_bridge.h +++ b/include/uapi/linux/netfilter_bridge.h @@ -11,6 +11,10 @@ #include #include +#ifndef __KERNEL__ +#include /* for INT_MIN, INT_MAX */ +#endif + /* Bridge Hooks */ /* After promisc drops, checksum checks. */ #define NF_BR_PRE_ROUTING 0 -- cgit v1.2.3 From 4269fea768a11a447d8de620ce420f2214d4685c Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 26 Oct 2018 11:14:28 +0200 Subject: Revert "netfilter: nft_numgen: add map lookups for numgen random operations" Laura found a better way to do this from userspace without requiring kernel infrastructure, revert this. Fixes: 978d8f9055c3 ("netfilter: nft_numgen: add map lookups for numgen random operations") Signed-off-by: Laura Garcia Liebana Signed-off-by: Pablo Neira Ayuso --- include/uapi/linux/netfilter/nf_tables.h | 4 +- net/netfilter/nft_numgen.c | 127 ------------------------------- 2 files changed, 2 insertions(+), 129 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/netfilter/nf_tables.h b/include/uapi/linux/netfilter/nf_tables.h index 579974b0bf0d..7de4f1bdaf06 100644 --- a/include/uapi/linux/netfilter/nf_tables.h +++ b/include/uapi/linux/netfilter/nf_tables.h @@ -1635,8 +1635,8 @@ enum nft_ng_attributes { NFTA_NG_MODULUS, NFTA_NG_TYPE, NFTA_NG_OFFSET, - NFTA_NG_SET_NAME, - NFTA_NG_SET_ID, + NFTA_NG_SET_NAME, /* deprecated */ + NFTA_NG_SET_ID, /* deprecated */ __NFTA_NG_MAX }; #define NFTA_NG_MAX (__NFTA_NG_MAX - 1) diff --git a/net/netfilter/nft_numgen.c b/net/netfilter/nft_numgen.c index 649d1700ec5b..3cc1b3dc3c3c 100644 --- a/net/netfilter/nft_numgen.c +++ b/net/netfilter/nft_numgen.c @@ -24,7 +24,6 @@ struct nft_ng_inc { u32 modulus; atomic_t counter; u32 offset; - struct nft_set *map; }; static u32 nft_ng_inc_gen(struct nft_ng_inc *priv) @@ -48,34 +47,11 @@ static void nft_ng_inc_eval(const struct nft_expr *expr, regs->data[priv->dreg] = nft_ng_inc_gen(priv); } -static void nft_ng_inc_map_eval(const struct nft_expr *expr, - struct nft_regs *regs, - const struct nft_pktinfo *pkt) -{ - struct nft_ng_inc *priv = nft_expr_priv(expr); - const struct nft_set *map = priv->map; - const struct nft_set_ext *ext; - u32 result; - bool found; - - result = nft_ng_inc_gen(priv); - found = map->ops->lookup(nft_net(pkt), map, &result, &ext); - - if (!found) - return; - - nft_data_copy(®s->data[priv->dreg], - nft_set_ext_data(ext), map->dlen); -} - static const struct nla_policy nft_ng_policy[NFTA_NG_MAX + 1] = { [NFTA_NG_DREG] = { .type = NLA_U32 }, [NFTA_NG_MODULUS] = { .type = NLA_U32 }, [NFTA_NG_TYPE] = { .type = NLA_U32 }, [NFTA_NG_OFFSET] = { .type = NLA_U32 }, - [NFTA_NG_SET_NAME] = { .type = NLA_STRING, - .len = NFT_SET_MAXNAMELEN - 1 }, - [NFTA_NG_SET_ID] = { .type = NLA_U32 }, }; static int nft_ng_inc_init(const struct nft_ctx *ctx, @@ -101,22 +77,6 @@ static int nft_ng_inc_init(const struct nft_ctx *ctx, NFT_DATA_VALUE, sizeof(u32)); } -static int nft_ng_inc_map_init(const struct nft_ctx *ctx, - const struct nft_expr *expr, - const struct nlattr * const tb[]) -{ - struct nft_ng_inc *priv = nft_expr_priv(expr); - u8 genmask = nft_genmask_next(ctx->net); - - nft_ng_inc_init(ctx, expr, tb); - - priv->map = nft_set_lookup_global(ctx->net, ctx->table, - tb[NFTA_NG_SET_NAME], - tb[NFTA_NG_SET_ID], genmask); - - return PTR_ERR_OR_ZERO(priv->map); -} - static int nft_ng_dump(struct sk_buff *skb, enum nft_registers dreg, u32 modulus, enum nft_ng_types type, u32 offset) { @@ -143,27 +103,10 @@ static int nft_ng_inc_dump(struct sk_buff *skb, const struct nft_expr *expr) priv->offset); } -static int nft_ng_inc_map_dump(struct sk_buff *skb, - const struct nft_expr *expr) -{ - const struct nft_ng_inc *priv = nft_expr_priv(expr); - - if (nft_ng_dump(skb, priv->dreg, priv->modulus, - NFT_NG_INCREMENTAL, priv->offset) || - nla_put_string(skb, NFTA_NG_SET_NAME, priv->map->name)) - goto nla_put_failure; - - return 0; - -nla_put_failure: - return -1; -} - struct nft_ng_random { enum nft_registers dreg:8; u32 modulus; u32 offset; - struct nft_set *map; }; static u32 nft_ng_random_gen(struct nft_ng_random *priv) @@ -183,25 +126,6 @@ static void nft_ng_random_eval(const struct nft_expr *expr, regs->data[priv->dreg] = nft_ng_random_gen(priv); } -static void nft_ng_random_map_eval(const struct nft_expr *expr, - struct nft_regs *regs, - const struct nft_pktinfo *pkt) -{ - struct nft_ng_random *priv = nft_expr_priv(expr); - const struct nft_set *map = priv->map; - const struct nft_set_ext *ext; - u32 result; - bool found; - - result = nft_ng_random_gen(priv); - found = map->ops->lookup(nft_net(pkt), map, &result, &ext); - if (!found) - return; - - nft_data_copy(®s->data[priv->dreg], - nft_set_ext_data(ext), map->dlen); -} - static int nft_ng_random_init(const struct nft_ctx *ctx, const struct nft_expr *expr, const struct nlattr * const tb[]) @@ -226,21 +150,6 @@ static int nft_ng_random_init(const struct nft_ctx *ctx, NFT_DATA_VALUE, sizeof(u32)); } -static int nft_ng_random_map_init(const struct nft_ctx *ctx, - const struct nft_expr *expr, - const struct nlattr * const tb[]) -{ - struct nft_ng_random *priv = nft_expr_priv(expr); - u8 genmask = nft_genmask_next(ctx->net); - - nft_ng_random_init(ctx, expr, tb); - priv->map = nft_set_lookup_global(ctx->net, ctx->table, - tb[NFTA_NG_SET_NAME], - tb[NFTA_NG_SET_ID], genmask); - - return PTR_ERR_OR_ZERO(priv->map); -} - static int nft_ng_random_dump(struct sk_buff *skb, const struct nft_expr *expr) { const struct nft_ng_random *priv = nft_expr_priv(expr); @@ -249,22 +158,6 @@ static int nft_ng_random_dump(struct sk_buff *skb, const struct nft_expr *expr) priv->offset); } -static int nft_ng_random_map_dump(struct sk_buff *skb, - const struct nft_expr *expr) -{ - const struct nft_ng_random *priv = nft_expr_priv(expr); - - if (nft_ng_dump(skb, priv->dreg, priv->modulus, - NFT_NG_RANDOM, priv->offset) || - nla_put_string(skb, NFTA_NG_SET_NAME, priv->map->name)) - goto nla_put_failure; - - return 0; - -nla_put_failure: - return -1; -} - static struct nft_expr_type nft_ng_type; static const struct nft_expr_ops nft_ng_inc_ops = { .type = &nft_ng_type, @@ -274,14 +167,6 @@ static const struct nft_expr_ops nft_ng_inc_ops = { .dump = nft_ng_inc_dump, }; -static const struct nft_expr_ops nft_ng_inc_map_ops = { - .type = &nft_ng_type, - .size = NFT_EXPR_SIZE(sizeof(struct nft_ng_inc)), - .eval = nft_ng_inc_map_eval, - .init = nft_ng_inc_map_init, - .dump = nft_ng_inc_map_dump, -}; - static const struct nft_expr_ops nft_ng_random_ops = { .type = &nft_ng_type, .size = NFT_EXPR_SIZE(sizeof(struct nft_ng_random)), @@ -290,14 +175,6 @@ static const struct nft_expr_ops nft_ng_random_ops = { .dump = nft_ng_random_dump, }; -static const struct nft_expr_ops nft_ng_random_map_ops = { - .type = &nft_ng_type, - .size = NFT_EXPR_SIZE(sizeof(struct nft_ng_random)), - .eval = nft_ng_random_map_eval, - .init = nft_ng_random_map_init, - .dump = nft_ng_random_map_dump, -}; - static const struct nft_expr_ops * nft_ng_select_ops(const struct nft_ctx *ctx, const struct nlattr * const tb[]) { @@ -312,12 +189,8 @@ nft_ng_select_ops(const struct nft_ctx *ctx, const struct nlattr * const tb[]) switch (type) { case NFT_NG_INCREMENTAL: - if (tb[NFTA_NG_SET_NAME]) - return &nft_ng_inc_map_ops; return &nft_ng_inc_ops; case NFT_NG_RANDOM: - if (tb[NFTA_NG_SET_NAME]) - return &nft_ng_random_map_ops; return &nft_ng_random_ops; } -- cgit v1.2.3 From af5aad059885b0bec3a23b9ea61c822357b6dd73 Mon Sep 17 00:00:00 2001 From: Ramalingam C Date: Mon, 29 Oct 2018 15:15:49 +0530 Subject: drm: hdcp2.2 authentication msg definitions This patch defines the hdcp2.2 protocol messages for authentication. v2: bit_fields are removed. Instead bitmasking used. [Tomas and Jani] prefix HDCP_2_2_ is added to the macros. [Tomas] v3: No Changes. v4: Style and spellings are fixed [Uma] v5: Fix for macros. v6: comment for Type is improved [Sean Paul] v7: %s/HDCP_2_2_LPRIME_HALF_LEN/HDCP_2_2_V_PRIME_HALF_LEN [Uma] %s/uintxx_t/uxx v8: %s/eceiver_id/receiver_id Signed-off-by: Ramalingam C Reviewed-by: Uma Shankar Acked-by: Sean Paul Signed-off-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/1540806351-7137-5-git-send-email-ramalingam.c@intel.com --- include/drm/drm_hdcp.h | 184 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 184 insertions(+) (limited to 'include') diff --git a/include/drm/drm_hdcp.h b/include/drm/drm_hdcp.h index 98e63d870139..7cd7269c7c7b 100644 --- a/include/drm/drm_hdcp.h +++ b/include/drm/drm_hdcp.h @@ -38,4 +38,188 @@ #define DRM_HDCP_DDC_BSTATUS 0x41 #define DRM_HDCP_DDC_KSV_FIFO 0x43 +#define DRM_HDCP_1_4_SRM_ID 0x8 +#define DRM_HDCP_1_4_VRL_LENGTH_SIZE 3 +#define DRM_HDCP_1_4_DCP_SIG_SIZE 40 + +/* Protocol message definition for HDCP2.2 specification */ +/* + * Protected content streams are classified into 2 types: + * - Type0: Can be transmitted with HDCP 1.4+ + * - Type1: Can be transmitted with HDCP 2.2+ + */ +#define HDCP_STREAM_TYPE0 0x00 +#define HDCP_STREAM_TYPE1 0x01 + +/* HDCP2.2 Msg IDs */ +#define HDCP_2_2_NULL_MSG 1 +#define HDCP_2_2_AKE_INIT 2 +#define HDCP_2_2_AKE_SEND_CERT 3 +#define HDCP_2_2_AKE_NO_STORED_KM 4 +#define HDCP_2_2_AKE_STORED_KM 5 +#define HDCP_2_2_AKE_SEND_HPRIME 7 +#define HDCP_2_2_AKE_SEND_PAIRING_INFO 8 +#define HDCP_2_2_LC_INIT 9 +#define HDCP_2_2_LC_SEND_LPRIME 10 +#define HDCP_2_2_SKE_SEND_EKS 11 +#define HDCP_2_2_REP_SEND_RECVID_LIST 12 +#define HDCP_2_2_REP_SEND_ACK 15 +#define HDCP_2_2_REP_STREAM_MANAGE 16 +#define HDCP_2_2_REP_STREAM_READY 17 +#define HDCP_2_2_ERRATA_DP_STREAM_TYPE 50 + +#define HDCP_2_2_RTX_LEN 8 +#define HDCP_2_2_RRX_LEN 8 + +#define HDCP_2_2_K_PUB_RX_MOD_N_LEN 128 +#define HDCP_2_2_K_PUB_RX_EXP_E_LEN 3 +#define HDCP_2_2_K_PUB_RX_LEN (HDCP_2_2_K_PUB_RX_MOD_N_LEN + \ + HDCP_2_2_K_PUB_RX_EXP_E_LEN) + +#define HDCP_2_2_DCP_LLC_SIG_LEN 384 + +#define HDCP_2_2_E_KPUB_KM_LEN 128 +#define HDCP_2_2_E_KH_KM_M_LEN (16 + 16) +#define HDCP_2_2_H_PRIME_LEN 32 +#define HDCP_2_2_E_KH_KM_LEN 16 +#define HDCP_2_2_RN_LEN 8 +#define HDCP_2_2_L_PRIME_LEN 32 +#define HDCP_2_2_E_DKEY_KS_LEN 16 +#define HDCP_2_2_RIV_LEN 8 +#define HDCP_2_2_SEQ_NUM_LEN 3 +#define HDCP_2_2_V_PRIME_HALF_LEN (HDCP_2_2_L_PRIME_LEN / 2) +#define HDCP_2_2_RECEIVER_ID_LEN DRM_HDCP_KSV_LEN +#define HDCP_2_2_MAX_DEVICE_COUNT 31 +#define HDCP_2_2_RECEIVER_IDS_MAX_LEN (HDCP_2_2_RECEIVER_ID_LEN * \ + HDCP_2_2_MAX_DEVICE_COUNT) +#define HDCP_2_2_MPRIME_LEN 32 + +/* Following Macros take a byte at a time for bit(s) masking */ +/* + * TODO: This has to be changed for DP MST, as multiple stream on + * same port is possible. + * For HDCP2.2 on HDMI and DP SST this value is always 1. + */ +#define HDCP_2_2_MAX_CONTENT_STREAMS_CNT 1 +#define HDCP_2_2_TXCAP_MASK_LEN 2 +#define HDCP_2_2_RXCAPS_LEN 3 +#define HDCP_2_2_RX_REPEATER(x) ((x) & BIT(0)) +#define HDCP_2_2_DP_HDCP_CAPABLE(x) ((x) & BIT(1)) +#define HDCP_2_2_RXINFO_LEN 2 + +/* HDCP1.x compliant device in downstream */ +#define HDCP_2_2_HDCP1_DEVICE_CONNECTED(x) ((x) & BIT(0)) + +/* HDCP2.0 Compliant repeater in downstream */ +#define HDCP_2_2_HDCP_2_0_REP_CONNECTED(x) ((x) & BIT(1)) +#define HDCP_2_2_MAX_CASCADE_EXCEEDED(x) ((x) & BIT(2)) +#define HDCP_2_2_MAX_DEVS_EXCEEDED(x) ((x) & BIT(3)) +#define HDCP_2_2_DEV_COUNT_LO(x) (((x) & (0xF << 4)) >> 4) +#define HDCP_2_2_DEV_COUNT_HI(x) ((x) & BIT(0)) +#define HDCP_2_2_DEPTH(x) (((x) & (0x7 << 1)) >> 1) + +struct hdcp2_cert_rx { + u8 receiver_id[HDCP_2_2_RECEIVER_ID_LEN]; + u8 kpub_rx[HDCP_2_2_K_PUB_RX_LEN]; + u8 reserved[2]; + u8 dcp_signature[HDCP_2_2_DCP_LLC_SIG_LEN]; +} __packed; + +struct hdcp2_streamid_type { + u8 stream_id; + u8 stream_type; +} __packed; + +/* + * The TxCaps field specified in the HDCP HDMI, DP specs + * This field is big endian as specified in the errata. + */ +struct hdcp2_tx_caps { + /* Transmitter must set this to 0x2 */ + u8 version; + + /* Reserved for HDCP and DP Spec. Read as Zero */ + u8 tx_cap_mask[HDCP_2_2_TXCAP_MASK_LEN]; +} __packed; + +/* Main structures for HDCP2.2 protocol communication */ +struct hdcp2_ake_init { + u8 msg_id; + u8 r_tx[HDCP_2_2_RTX_LEN]; + struct hdcp2_tx_caps tx_caps; +} __packed; + +struct hdcp2_ake_send_cert { + u8 msg_id; + struct hdcp2_cert_rx cert_rx; + u8 r_rx[HDCP_2_2_RRX_LEN]; + u8 rx_caps[HDCP_2_2_RXCAPS_LEN]; +} __packed; + +struct hdcp2_ake_no_stored_km { + u8 msg_id; + u8 e_kpub_km[HDCP_2_2_E_KPUB_KM_LEN]; +} __packed; + +struct hdcp2_ake_stored_km { + u8 msg_id; + u8 e_kh_km_m[HDCP_2_2_E_KH_KM_M_LEN]; +} __packed; + +struct hdcp2_ake_send_hprime { + u8 msg_id; + u8 h_prime[HDCP_2_2_H_PRIME_LEN]; +} __packed; + +struct hdcp2_ake_send_pairing_info { + u8 msg_id; + u8 e_kh_km[HDCP_2_2_E_KH_KM_LEN]; +} __packed; + +struct hdcp2_lc_init { + u8 msg_id; + u8 r_n[HDCP_2_2_RN_LEN]; +} __packed; + +struct hdcp2_lc_send_lprime { + u8 msg_id; + u8 l_prime[HDCP_2_2_L_PRIME_LEN]; +} __packed; + +struct hdcp2_ske_send_eks { + u8 msg_id; + u8 e_dkey_ks[HDCP_2_2_E_DKEY_KS_LEN]; + u8 riv[HDCP_2_2_RIV_LEN]; +} __packed; + +struct hdcp2_rep_send_receiverid_list { + u8 msg_id; + u8 rx_info[HDCP_2_2_RXINFO_LEN]; + u8 seq_num_v[HDCP_2_2_SEQ_NUM_LEN]; + u8 v_prime[HDCP_2_2_V_PRIME_HALF_LEN]; + u8 receiver_ids[HDCP_2_2_RECEIVER_IDS_MAX_LEN]; +} __packed; + +struct hdcp2_rep_send_ack { + u8 msg_id; + u8 v[HDCP_2_2_V_PRIME_HALF_LEN]; +} __packed; + +struct hdcp2_rep_stream_manage { + u8 msg_id; + u8 seq_num_m[HDCP_2_2_SEQ_NUM_LEN]; + __be16 k; + struct hdcp2_streamid_type streams[HDCP_2_2_MAX_CONTENT_STREAMS_CNT]; +} __packed; + +struct hdcp2_rep_stream_ready { + u8 msg_id; + u8 m_prime[HDCP_2_2_MPRIME_LEN]; +} __packed; + +struct hdcp2_dp_errata_stream_type { + u8 msg_id; + u8 stream_type; +} __packed; + #endif -- cgit v1.2.3 From 8b44fefee694ef6095683d91d6f06d23f02a6212 Mon Sep 17 00:00:00 2001 From: Ramalingam C Date: Mon, 29 Oct 2018 15:15:50 +0530 Subject: drm: HDMI and DP specific HDCP2.2 defines This patch adds HDCP register definitions for HDMI and DP HDCP adaptations. HDMI specific HDCP2.2 register definitions are added into drm_hdcp.h, where as HDCP2.2 register offsets in DPCD offsets are defined at drm_dp_helper.h. v2: bit_field definitions are replaced by macros. [Tomas and Jani] v3: No Changes. v4: Comments style and typos are fixed [Uma] v5: Fix for macros. v6: Adds _MS to the timeouts to represent units [Sean Paul] v7: Macro DP_HDCP_2_2_REG_EKH_KM_OFFSET renamed [Uma] Redundant macro is removed [Uma] Signed-off-by: Ramalingam C Reviewed-by: Sean Paul Acked-by: Sean Paul (for merging through drm-intel) Signed-off-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/1540806351-7137-6-git-send-email-ramalingam.c@intel.com --- include/drm/drm_dp_helper.h | 51 +++++++++++++++++++++++++++++++++++++++++++++ include/drm/drm_hdcp.h | 28 +++++++++++++++++++++++++ 2 files changed, 79 insertions(+) (limited to 'include') diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h index 2a3843f248cf..8920ce12650c 100644 --- a/include/drm/drm_dp_helper.h +++ b/include/drm/drm_dp_helper.h @@ -905,6 +905,57 @@ #define DP_AUX_HDCP_KSV_FIFO 0x6802C #define DP_AUX_HDCP_AINFO 0x6803B +/* DP HDCP2.2 parameter offsets in DPCD address space */ +#define DP_HDCP_2_2_REG_RTX_OFFSET 0x69000 +#define DP_HDCP_2_2_REG_TXCAPS_OFFSET 0x69008 +#define DP_HDCP_2_2_REG_CERT_RX_OFFSET 0x6900B +#define DP_HDCP_2_2_REG_RRX_OFFSET 0x69215 +#define DP_HDCP_2_2_REG_RX_CAPS_OFFSET 0x6921D +#define DP_HDCP_2_2_REG_EKPUB_KM_OFFSET 0x69220 +#define DP_HDCP_2_2_REG_EKH_KM_WR_OFFSET 0x692A0 +#define DP_HDCP_2_2_REG_M_OFFSET 0x692B0 +#define DP_HDCP_2_2_REG_HPRIME_OFFSET 0x692C0 +#define DP_HDCP_2_2_REG_EKH_KM_RD_OFFSET 0x692E0 +#define DP_HDCP_2_2_REG_RN_OFFSET 0x692F0 +#define DP_HDCP_2_2_REG_LPRIME_OFFSET 0x692F8 +#define DP_HDCP_2_2_REG_EDKEY_KS_OFFSET 0x69318 +#define DP_HDCP_2_2_REG_RIV_OFFSET 0x69328 +#define DP_HDCP_2_2_REG_RXINFO_OFFSET 0x69330 +#define DP_HDCP_2_2_REG_SEQ_NUM_V_OFFSET 0x69332 +#define DP_HDCP_2_2_REG_VPRIME_OFFSET 0x69335 +#define DP_HDCP_2_2_REG_RECV_ID_LIST_OFFSET 0x69345 +#define DP_HDCP_2_2_REG_V_OFFSET 0x693E0 +#define DP_HDCP_2_2_REG_SEQ_NUM_M_OFFSET 0x693F0 +#define DP_HDCP_2_2_REG_K_OFFSET 0x693F3 +#define DP_HDCP_2_2_REG_STREAM_ID_TYPE_OFFSET 0x693F5 +#define DP_HDCP_2_2_REG_MPRIME_OFFSET 0x69473 +#define DP_HDCP_2_2_REG_RXSTATUS_OFFSET 0x69493 +#define DP_HDCP_2_2_REG_STREAM_TYPE_OFFSET 0x69494 +#define DP_HDCP_2_2_REG_DBG_OFFSET 0x69518 + +/* DP HDCP message start offsets in DPCD address space */ +#define DP_HDCP_2_2_AKE_INIT_OFFSET DP_HDCP_2_2_REG_RTX_OFFSET +#define DP_HDCP_2_2_AKE_SEND_CERT_OFFSET DP_HDCP_2_2_REG_CERT_RX_OFFSET +#define DP_HDCP_2_2_AKE_NO_STORED_KM_OFFSET DP_HDCP_2_2_REG_EKPUB_KM_OFFSET +#define DP_HDCP_2_2_AKE_STORED_KM_OFFSET DP_HDCP_2_2_REG_EKH_KM_WR_OFFSET +#define DP_HDCP_2_2_AKE_SEND_HPRIME_OFFSET DP_HDCP_2_2_REG_HPRIME_OFFSET +#define DP_HDCP_2_2_AKE_SEND_PAIRING_INFO_OFFSET \ + DP_HDCP_2_2_REG_EKH_KM_RD_OFFSET +#define DP_HDCP_2_2_LC_INIT_OFFSET DP_HDCP_2_2_REG_RN_OFFSET +#define DP_HDCP_2_2_LC_SEND_LPRIME_OFFSET DP_HDCP_2_2_REG_LPRIME_OFFSET +#define DP_HDCP_2_2_SKE_SEND_EKS_OFFSET DP_HDCP_2_2_REG_EDKEY_KS_OFFSET +#define DP_HDCP_2_2_REP_SEND_RECVID_LIST_OFFSET DP_HDCP_2_2_REG_RXINFO_OFFSET +#define DP_HDCP_2_2_REP_SEND_ACK_OFFSET DP_HDCP_2_2_REG_V_OFFSET +#define DP_HDCP_2_2_REP_STREAM_MANAGE_OFFSET DP_HDCP_2_2_REG_SEQ_NUM_M_OFFSET +#define DP_HDCP_2_2_REP_STREAM_READY_OFFSET DP_HDCP_2_2_REG_MPRIME_OFFSET + +#define HDCP_2_2_DP_RXSTATUS_LEN 1 +#define HDCP_2_2_DP_RXSTATUS_READY(x) ((x) & BIT(0)) +#define HDCP_2_2_DP_RXSTATUS_H_PRIME(x) ((x) & BIT(1)) +#define HDCP_2_2_DP_RXSTATUS_PAIRING(x) ((x) & BIT(2)) +#define HDCP_2_2_DP_RXSTATUS_REAUTH_REQ(x) ((x) & BIT(3)) +#define HDCP_2_2_DP_RXSTATUS_LINK_FAILED(x) ((x) & BIT(4)) + /* DP 1.2 Sideband message defines */ /* peer device type - DP 1.2a Table 2-92 */ #define DP_PEER_DEVICE_NONE 0x0 diff --git a/include/drm/drm_hdcp.h b/include/drm/drm_hdcp.h index 7cd7269c7c7b..a6de09c5e47f 100644 --- a/include/drm/drm_hdcp.h +++ b/include/drm/drm_hdcp.h @@ -222,4 +222,32 @@ struct hdcp2_dp_errata_stream_type { u8 stream_type; } __packed; +/* HDCP2.2 TIMEOUTs in mSec */ +#define HDCP_2_2_CERT_TIMEOUT_MS 100 +#define HDCP_2_2_HPRIME_NO_PAIRED_TIMEOUT_MS 1000 +#define HDCP_2_2_HPRIME_PAIRED_TIMEOUT_MS 200 +#define HDCP_2_2_PAIRING_TIMEOUT_MS 200 +#define HDCP_2_2_HDMI_LPRIME_TIMEOUT_MS 20 +#define HDCP_2_2_DP_LPRIME_TIMEOUT_MS 7 +#define HDCP_2_2_RECVID_LIST_TIMEOUT_MS 3000 +#define HDCP_2_2_STREAM_READY_TIMEOUT_MS 100 + +/* HDMI HDCP2.2 Register Offsets */ +#define HDCP_2_2_HDMI_REG_VER_OFFSET 0x50 +#define HDCP_2_2_HDMI_REG_WR_MSG_OFFSET 0x60 +#define HDCP_2_2_HDMI_REG_RXSTATUS_OFFSET 0x70 +#define HDCP_2_2_HDMI_REG_RD_MSG_OFFSET 0x80 +#define HDCP_2_2_HDMI_REG_DBG_OFFSET 0xC0 + +#define HDCP_2_2_HDMI_SUPPORT_MASK BIT(2) +#define HDCP_2_2_RX_CAPS_VERSION_VAL 0x02 +#define HDCP_2_2_SEQ_NUM_MAX 0xFFFFFF +#define HDCP_2_2_DELAY_BEFORE_ENCRYPTION_EN 200 + +/* Below macros take a byte at a time and mask the bit(s) */ +#define HDCP_2_2_HDMI_RXSTATUS_LEN 2 +#define HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(x) ((x) & 0x3) +#define HDCP_2_2_HDMI_RXSTATUS_READY(x) ((x) & BIT(2)) +#define HDCP_2_2_HDMI_RXSTATUS_REAUTH_REQ(x) ((x) & BIT(3)) + #endif -- cgit v1.2.3 From ffddc4363c280d56655b5929661def3686a2a6aa Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 30 Oct 2018 17:19:18 -0700 Subject: drm/dp: Add DP DSC DPCD receiver capability size define and missing SHIFT This patch defines the DP DSC receiver capability size that gives total number of DP DSC DPCD registers. This also adds a missing #defines for DP DSC support missed in the commit id (ab6a46ea6842ce "Add DPCD definitions for DP 1.4 DSC feature") v3: * MIN_SLICE_WIDTH = 2560 (Anusha) * Define DP_DSC_SLICE_WIDTH_MULTIPLIER = 320 v2: * Add SHIFT define and DECOMPRESSION_EN define missed in prev patch Cc: dri-devel@lists.freedesktop.org Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Gaurav K Singh Signed-off-by: Manasi Navare Acked-by: Sean Paul (For merging through drm-intel) Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181031001923.31442-2-manasi.d.navare@intel.com --- include/drm/drm_dp_helper.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h index 8920ce12650c..4139738d3024 100644 --- a/include/drm/drm_dp_helper.h +++ b/include/drm/drm_dp_helper.h @@ -231,6 +231,8 @@ #define DP_DSC_MAX_BITS_PER_PIXEL_LOW 0x067 /* eDP 1.4 */ #define DP_DSC_MAX_BITS_PER_PIXEL_HI 0x068 /* eDP 1.4 */ +# define DP_DSC_MAX_BITS_PER_PIXEL_HI_MASK (0x3 << 0) +# define DP_DSC_MAX_BITS_PER_PIXEL_HI_SHIFT 8 #define DP_DSC_DEC_COLOR_FORMAT_CAP 0x069 # define DP_DSC_RGB (1 << 0) @@ -279,6 +281,8 @@ # define DP_DSC_THROUGHPUT_MODE_1_1000 (14 << 4) #define DP_DSC_MAX_SLICE_WIDTH 0x06C +#define DP_DSC_MIN_SLICE_WIDTH_VALUE 2560 +#define DP_DSC_SLICE_WIDTH_MULTIPLIER 320 #define DP_DSC_SLICE_CAP_2 0x06D # define DP_DSC_16_PER_DP_DSC_SINK (1 << 0) @@ -477,6 +481,7 @@ # define DP_AUX_FRAME_SYNC_VALID (1 << 0) #define DP_DSC_ENABLE 0x160 /* DP 1.4 */ +# define DP_DECOMPRESSION_EN (1 << 0) #define DP_PSR_EN_CFG 0x170 /* XXX 1.2? */ # define DP_PSR_ENABLE (1 << 0) @@ -1014,6 +1019,7 @@ u8 drm_dp_get_adjust_request_pre_emphasis(const u8 link_status[DP_LINK_STATUS_SI #define DP_BRANCH_OUI_HEADER_SIZE 0xc #define DP_RECEIVER_CAP_SIZE 0xf +#define DP_DSC_RECEIVER_CAP_SIZE 0xf #define EDP_PSR_RECEIVER_CAP_SIZE 2 #define EDP_DISPLAY_CTL_CAP_SIZE 3 -- cgit v1.2.3 From 0575650077eae2a431f33532de4e15405cc9f653 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 30 Oct 2018 17:19:20 -0700 Subject: drm/dp: DRM DP helper/macros to get DP sink DSC parameters This patch adds inline functions and helpers for obtaining DP sink's supported DSC parameters like DSC sink support, eDP compressed BPP supported, maximum slice count supported by the sink devices, DSC line buffer bit depth supported on DP sink, DSC sink maximum color depth by parsing corresponding DPCD registers. v4: * Add helper to give line buf bit depth (Manasi) * Correct the bit masking in color depth helper (manasi) v3: * Use SLICE_CAP_2 for DP (Anusha) v2: * Add DSC sink support macro (Jani N) Cc: Gaurav K Singh Cc: dri-devel@lists.freedesktop.org Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Reviewed-by: Gaurav K Singh Acked-by: Sean Paul (For merging through drm-intel) Link: https://patchwork.freedesktop.org/patch/msgid/20181031001923.31442-4-manasi.d.navare@intel.com --- drivers/gpu/drm/drm_dp_helper.c | 90 +++++++++++++++++++++++++++++++++++++++++ include/drm/drm_dp_helper.h | 30 ++++++++++++++ 2 files changed, 120 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c index 37c01b6076ec..6d483487f2b4 100644 --- a/drivers/gpu/drm/drm_dp_helper.c +++ b/drivers/gpu/drm/drm_dp_helper.c @@ -1352,3 +1352,93 @@ int drm_dp_read_desc(struct drm_dp_aux *aux, struct drm_dp_desc *desc, return 0; } EXPORT_SYMBOL(drm_dp_read_desc); + +/** + * DRM DP Helpers for DSC + */ +u8 drm_dp_dsc_sink_max_slice_count(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE], + bool is_edp) +{ + u8 slice_cap1 = dsc_dpcd[DP_DSC_SLICE_CAP_1 - DP_DSC_SUPPORT]; + + if (is_edp) { + /* For eDP, register DSC_SLICE_CAPABILITIES_1 gives slice count */ + if (slice_cap1 & DP_DSC_4_PER_DP_DSC_SINK) + return 4; + if (slice_cap1 & DP_DSC_2_PER_DP_DSC_SINK) + return 2; + if (slice_cap1 & DP_DSC_1_PER_DP_DSC_SINK) + return 1; + } else { + /* For DP, use values from DSC_SLICE_CAP_1 and DSC_SLICE_CAP2 */ + u8 slice_cap2 = dsc_dpcd[DP_DSC_SLICE_CAP_2 - DP_DSC_SUPPORT]; + + if (slice_cap2 & DP_DSC_24_PER_DP_DSC_SINK) + return 24; + if (slice_cap2 & DP_DSC_20_PER_DP_DSC_SINK) + return 20; + if (slice_cap2 & DP_DSC_16_PER_DP_DSC_SINK) + return 16; + if (slice_cap1 & DP_DSC_12_PER_DP_DSC_SINK) + return 12; + if (slice_cap1 & DP_DSC_10_PER_DP_DSC_SINK) + return 10; + if (slice_cap1 & DP_DSC_8_PER_DP_DSC_SINK) + return 8; + if (slice_cap1 & DP_DSC_6_PER_DP_DSC_SINK) + return 6; + if (slice_cap1 & DP_DSC_4_PER_DP_DSC_SINK) + return 4; + if (slice_cap1 & DP_DSC_2_PER_DP_DSC_SINK) + return 2; + if (slice_cap1 & DP_DSC_1_PER_DP_DSC_SINK) + return 1; + } + + return 0; +} +EXPORT_SYMBOL(drm_dp_dsc_sink_max_slice_count); + +u8 drm_dp_dsc_sink_line_buf_depth(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) +{ + u8 line_buf_depth = dsc_dpcd[DP_DSC_LINE_BUF_BIT_DEPTH - DP_DSC_SUPPORT]; + + switch (line_buf_depth & DP_DSC_LINE_BUF_BIT_DEPTH_MASK) { + case DP_DSC_LINE_BUF_BIT_DEPTH_9: + return 9; + case DP_DSC_LINE_BUF_BIT_DEPTH_10: + return 10; + case DP_DSC_LINE_BUF_BIT_DEPTH_11: + return 11; + case DP_DSC_LINE_BUF_BIT_DEPTH_12: + return 12; + case DP_DSC_LINE_BUF_BIT_DEPTH_13: + return 13; + case DP_DSC_LINE_BUF_BIT_DEPTH_14: + return 14; + case DP_DSC_LINE_BUF_BIT_DEPTH_15: + return 15; + case DP_DSC_LINE_BUF_BIT_DEPTH_16: + return 16; + case DP_DSC_LINE_BUF_BIT_DEPTH_8: + return 8; + } + + return 0; +} +EXPORT_SYMBOL(drm_dp_dsc_sink_line_buf_depth); + +u8 drm_dp_dsc_sink_max_color_depth(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) +{ + u8 color_depth = dsc_dpcd[DP_DSC_DEC_COLOR_DEPTH_CAP - DP_DSC_SUPPORT]; + + if (color_depth & DP_DSC_12_BPC) + return 12; + if (color_depth & DP_DSC_10_BPC) + return 10; + if (color_depth & DP_DSC_8_BPC) + return 8; + + return 0; +} +EXPORT_SYMBOL(drm_dp_dsc_sink_max_color_depth); diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h index 4139738d3024..636607388a66 100644 --- a/include/drm/drm_dp_helper.h +++ b/include/drm/drm_dp_helper.h @@ -1116,6 +1116,36 @@ drm_dp_is_branch(const u8 dpcd[DP_RECEIVER_CAP_SIZE]) return dpcd[DP_DOWNSTREAMPORT_PRESENT] & DP_DWN_STRM_PORT_PRESENT; } +/* DP/eDP DSC support */ +u8 drm_dp_dsc_sink_max_slice_count(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE], + bool is_edp); +u8 drm_dp_dsc_sink_line_buf_depth(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]); +u8 drm_dp_dsc_sink_max_color_depth(const u8 dsc_dpc[DP_DSC_RECEIVER_CAP_SIZE]); + +static inline bool +drm_dp_sink_supports_dsc(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) +{ + return dsc_dpcd[DP_DSC_SUPPORT - DP_DSC_SUPPORT] & + DP_DSC_DECOMPRESSION_IS_SUPPORTED; +} + +static inline u16 +drm_edp_dsc_sink_output_bpp(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) +{ + return dsc_dpcd[DP_DSC_MAX_BITS_PER_PIXEL_LOW - DP_DSC_SUPPORT] | + (dsc_dpcd[DP_DSC_MAX_BITS_PER_PIXEL_HI - DP_DSC_SUPPORT] & + DP_DSC_MAX_BITS_PER_PIXEL_HI_MASK << + DP_DSC_MAX_BITS_PER_PIXEL_HI_SHIFT); +} + +static inline u32 +drm_dp_dsc_sink_max_slice_width(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) +{ + /* Max Slicewidth = Number of Pixels * 320 */ + return dsc_dpcd[DP_DSC_MAX_SLICE_WIDTH - DP_DSC_SUPPORT] * + DP_DSC_SLICE_WIDTH_MULTIPLIER; +} + /* * DisplayPort AUX channel */ -- cgit v1.2.3 From 6e97272a9aaa5d4ec639c19d7f63a5d9b317dffb Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 30 Oct 2018 17:19:23 -0700 Subject: drm/dp: Define payload size for DP SDP PPS packet DP 1.4 spec defines DP secondary data packet for DSC picture parameter set. This patch defines its payload size according to the DP 1.4 specification. Signed-off-by: Manasi Navare Cc: dri-devel@lists.freedesktop.org Cc: Gaurav K Singh Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Acked-by: Sean Paul (For merging through drm-intel) Reviewed-by: Harry Wentland Link: https://patchwork.freedesktop.org/patch/msgid/20181031001923.31442-7-manasi.d.navare@intel.com --- include/drm/drm_dp_helper.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h index 636607388a66..f82e68db16e0 100644 --- a/include/drm/drm_dp_helper.h +++ b/include/drm/drm_dp_helper.h @@ -1050,6 +1050,7 @@ struct dp_sdp_header { #define EDP_SDP_HEADER_REVISION_MASK 0x1F #define EDP_SDP_HEADER_VALID_PAYLOAD_BYTES 0x1F +#define DP_SDP_PPS_HEADER_PAYLOAD_BYTES_MINUS_1 0x7F struct edp_vsc_psr { struct dp_sdp_header sdp_header; -- cgit v1.2.3 From 439cd39ea136d2c026805264d58a91f36b6b64ca Mon Sep 17 00:00:00 2001 From: Stefano Brivio Date: Sat, 14 Jul 2018 21:59:43 +0200 Subject: netfilter: ipset: list:set: Decrease refcount synchronously on deletion and replace Commit 45040978c899 ("netfilter: ipset: Fix set:list type crash when flush/dump set in parallel") postponed decreasing set reference counters to the RCU callback. An 'ipset del' command can terminate before the RCU grace period is elapsed, and if sets are listed before then, the reference counter shown in userspace will be wrong: # ipset create h hash:ip; ipset create l list:set; ipset add l # ipset del l h; ipset list h Name: h Type: hash:ip Revision: 4 Header: family inet hashsize 1024 maxelem 65536 Size in memory: 88 References: 1 Number of entries: 0 Members: # sleep 1; ipset list h Name: h Type: hash:ip Revision: 4 Header: family inet hashsize 1024 maxelem 65536 Size in memory: 88 References: 0 Number of entries: 0 Members: Fix this by making the reference count update synchronous again. As a result, when sets are listed, ip_set_name_byindex() might now fetch a set whose reference count is already zero. Instead of relying on the reference count to protect against concurrent set renaming, grab ip_set_ref_lock as reader and copy the name, while holding the same lock in ip_set_rename() as writer instead. Reported-by: Li Shuang Fixes: 45040978c899 ("netfilter: ipset: Fix set:list type crash when flush/dump set in parallel") Signed-off-by: Stefano Brivio Signed-off-by: Jozsef Kadlecsik Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/ipset/ip_set.h | 2 +- net/netfilter/ipset/ip_set_core.c | 23 +++++++++++------------ net/netfilter/ipset/ip_set_list_set.c | 17 +++++++++++------ 3 files changed, 23 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/ipset/ip_set.h b/include/linux/netfilter/ipset/ip_set.h index 34fc80f3eb90..1d100efe74ec 100644 --- a/include/linux/netfilter/ipset/ip_set.h +++ b/include/linux/netfilter/ipset/ip_set.h @@ -314,7 +314,7 @@ enum { extern ip_set_id_t ip_set_get_byname(struct net *net, const char *name, struct ip_set **set); extern void ip_set_put_byindex(struct net *net, ip_set_id_t index); -extern const char *ip_set_name_byindex(struct net *net, ip_set_id_t index); +extern void ip_set_name_byindex(struct net *net, ip_set_id_t index, char *name); extern ip_set_id_t ip_set_nfnl_get_byindex(struct net *net, ip_set_id_t index); extern void ip_set_nfnl_put(struct net *net, ip_set_id_t index); diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index bc4bd247bb7d..fa15a831aeee 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -693,21 +693,20 @@ ip_set_put_byindex(struct net *net, ip_set_id_t index) EXPORT_SYMBOL_GPL(ip_set_put_byindex); /* Get the name of a set behind a set index. - * We assume the set is referenced, so it does exist and - * can't be destroyed. The set cannot be renamed due to - * the referencing either. - * + * Set itself is protected by RCU, but its name isn't: to protect against + * renaming, grab ip_set_ref_lock as reader (see ip_set_rename()) and copy the + * name. */ -const char * -ip_set_name_byindex(struct net *net, ip_set_id_t index) +void +ip_set_name_byindex(struct net *net, ip_set_id_t index, char *name) { - const struct ip_set *set = ip_set_rcu_get(net, index); + struct ip_set *set = ip_set_rcu_get(net, index); BUG_ON(!set); - BUG_ON(set->ref == 0); - /* Referenced, so it's safe */ - return set->name; + read_lock_bh(&ip_set_ref_lock); + strncpy(name, set->name, IPSET_MAXNAMELEN); + read_unlock_bh(&ip_set_ref_lock); } EXPORT_SYMBOL_GPL(ip_set_name_byindex); @@ -1153,7 +1152,7 @@ static int ip_set_rename(struct net *net, struct sock *ctnl, if (!set) return -ENOENT; - read_lock_bh(&ip_set_ref_lock); + write_lock_bh(&ip_set_ref_lock); if (set->ref != 0) { ret = -IPSET_ERR_REFERENCED; goto out; @@ -1170,7 +1169,7 @@ static int ip_set_rename(struct net *net, struct sock *ctnl, strncpy(set->name, name2, IPSET_MAXNAMELEN); out: - read_unlock_bh(&ip_set_ref_lock); + write_unlock_bh(&ip_set_ref_lock); return ret; } diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index 072a658fde04..4eef55da0878 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -148,9 +148,7 @@ __list_set_del_rcu(struct rcu_head * rcu) { struct set_elem *e = container_of(rcu, struct set_elem, rcu); struct ip_set *set = e->set; - struct list_set *map = set->data; - ip_set_put_byindex(map->net, e->id); ip_set_ext_destroy(set, e); kfree(e); } @@ -158,15 +156,21 @@ __list_set_del_rcu(struct rcu_head * rcu) static inline void list_set_del(struct ip_set *set, struct set_elem *e) { + struct list_set *map = set->data; + set->elements--; list_del_rcu(&e->list); + ip_set_put_byindex(map->net, e->id); call_rcu(&e->rcu, __list_set_del_rcu); } static inline void -list_set_replace(struct set_elem *e, struct set_elem *old) +list_set_replace(struct ip_set *set, struct set_elem *e, struct set_elem *old) { + struct list_set *map = set->data; + list_replace_rcu(&old->list, &e->list); + ip_set_put_byindex(map->net, old->id); call_rcu(&old->rcu, __list_set_del_rcu); } @@ -298,7 +302,7 @@ list_set_uadd(struct ip_set *set, void *value, const struct ip_set_ext *ext, INIT_LIST_HEAD(&e->list); list_set_init_extensions(set, ext, e); if (n) - list_set_replace(e, n); + list_set_replace(set, e, n); else if (next) list_add_tail_rcu(&e->list, &next->list); else if (prev) @@ -486,6 +490,7 @@ list_set_list(const struct ip_set *set, const struct list_set *map = set->data; struct nlattr *atd, *nested; u32 i = 0, first = cb->args[IPSET_CB_ARG0]; + char name[IPSET_MAXNAMELEN]; struct set_elem *e; int ret = 0; @@ -504,8 +509,8 @@ list_set_list(const struct ip_set *set, nested = ipset_nest_start(skb, IPSET_ATTR_DATA); if (!nested) goto nla_put_failure; - if (nla_put_string(skb, IPSET_ATTR_NAME, - ip_set_name_byindex(map->net, e->id))) + ip_set_name_byindex(map->net, e->id, name); + if (nla_put_string(skb, IPSET_ATTR_NAME, name)) goto nla_put_failure; if (ip_set_put_extensions(skb, set, e, true)) goto nla_put_failure; -- cgit v1.2.3 From a8874e7e8a8896f2b6c641f4b8e2473eafd35204 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Wed, 31 Oct 2018 12:11:48 +0100 Subject: mm: make the __PAGETABLE_PxD_FOLDED defines non-empty Change the currently empty defines for __PAGETABLE_PMD_FOLDED, __PAGETABLE_PUD_FOLDED and __PAGETABLE_P4D_FOLDED to return 1. This makes it possible to use __is_defined() to test if the preprocessor define exists. Acked-by: Kirill A. Shutemov Signed-off-by: Martin Schwidefsky --- arch/arm/include/asm/pgtable-2level.h | 2 +- arch/m68k/include/asm/pgtable_mm.h | 4 ++-- arch/microblaze/include/asm/pgtable.h | 2 +- arch/nds32/include/asm/pgtable.h | 2 +- arch/parisc/include/asm/pgtable.h | 2 +- include/asm-generic/4level-fixup.h | 2 +- include/asm-generic/5level-fixup.h | 2 +- include/asm-generic/pgtable-nop4d-hack.h | 2 +- include/asm-generic/pgtable-nop4d.h | 2 +- include/asm-generic/pgtable-nopmd.h | 2 +- include/asm-generic/pgtable-nopud.h | 2 +- 11 files changed, 12 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/arch/arm/include/asm/pgtable-2level.h b/arch/arm/include/asm/pgtable-2level.h index 92fd2c8a9af0..12659ce5c1f3 100644 --- a/arch/arm/include/asm/pgtable-2level.h +++ b/arch/arm/include/asm/pgtable-2level.h @@ -10,7 +10,7 @@ #ifndef _ASM_PGTABLE_2LEVEL_H #define _ASM_PGTABLE_2LEVEL_H -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 /* * Hardware-wise, we have a two level page table structure, where the first diff --git a/arch/m68k/include/asm/pgtable_mm.h b/arch/m68k/include/asm/pgtable_mm.h index 6181e4134483..fe3ddd73a0cc 100644 --- a/arch/m68k/include/asm/pgtable_mm.h +++ b/arch/m68k/include/asm/pgtable_mm.h @@ -55,12 +55,12 @@ */ #ifdef CONFIG_SUN3 #define PTRS_PER_PTE 16 -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 #define PTRS_PER_PMD 1 #define PTRS_PER_PGD 2048 #elif defined(CONFIG_COLDFIRE) #define PTRS_PER_PTE 512 -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 #define PTRS_PER_PMD 1 #define PTRS_PER_PGD 1024 #else diff --git a/arch/microblaze/include/asm/pgtable.h b/arch/microblaze/include/asm/pgtable.h index f64ebb9c9a41..e14b6621c933 100644 --- a/arch/microblaze/include/asm/pgtable.h +++ b/arch/microblaze/include/asm/pgtable.h @@ -63,7 +63,7 @@ extern int mem_init_done; #include -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 #ifdef __KERNEL__ #ifndef __ASSEMBLY__ diff --git a/arch/nds32/include/asm/pgtable.h b/arch/nds32/include/asm/pgtable.h index d3e19a55cf53..9f52db930c00 100644 --- a/arch/nds32/include/asm/pgtable.h +++ b/arch/nds32/include/asm/pgtable.h @@ -4,7 +4,7 @@ #ifndef _ASMNDS32_PGTABLE_H #define _ASMNDS32_PGTABLE_H -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 #include #include diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h index b941ac7d4e70..c7bb74e22436 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -111,7 +111,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr) #if CONFIG_PGTABLE_LEVELS == 3 #define BITS_PER_PMD (PAGE_SHIFT + PMD_ORDER - BITS_PER_PMD_ENTRY) #else -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 #define BITS_PER_PMD 0 #endif #define PTRS_PER_PMD (1UL << BITS_PER_PMD) diff --git a/include/asm-generic/4level-fixup.h b/include/asm-generic/4level-fixup.h index 89f3b03b1445..e3667c9a33a5 100644 --- a/include/asm-generic/4level-fixup.h +++ b/include/asm-generic/4level-fixup.h @@ -3,7 +3,7 @@ #define _4LEVEL_FIXUP_H #define __ARCH_HAS_4LEVEL_HACK -#define __PAGETABLE_PUD_FOLDED +#define __PAGETABLE_PUD_FOLDED 1 #define PUD_SHIFT PGDIR_SHIFT #define PUD_SIZE PGDIR_SIZE diff --git a/include/asm-generic/5level-fixup.h b/include/asm-generic/5level-fixup.h index 9c2e0708eb82..73474bb52344 100644 --- a/include/asm-generic/5level-fixup.h +++ b/include/asm-generic/5level-fixup.h @@ -3,7 +3,7 @@ #define _5LEVEL_FIXUP_H #define __ARCH_HAS_5LEVEL_HACK -#define __PAGETABLE_P4D_FOLDED +#define __PAGETABLE_P4D_FOLDED 1 #define P4D_SHIFT PGDIR_SHIFT #define P4D_SIZE PGDIR_SIZE diff --git a/include/asm-generic/pgtable-nop4d-hack.h b/include/asm-generic/pgtable-nop4d-hack.h index 0c34215263b8..1d6dd38c0e5e 100644 --- a/include/asm-generic/pgtable-nop4d-hack.h +++ b/include/asm-generic/pgtable-nop4d-hack.h @@ -5,7 +5,7 @@ #ifndef __ASSEMBLY__ #include -#define __PAGETABLE_PUD_FOLDED +#define __PAGETABLE_PUD_FOLDED 1 /* * Having the pud type consist of a pgd gets the size right, and allows diff --git a/include/asm-generic/pgtable-nop4d.h b/include/asm-generic/pgtable-nop4d.h index 1a29b2a0282b..04cb913797bc 100644 --- a/include/asm-generic/pgtable-nop4d.h +++ b/include/asm-generic/pgtable-nop4d.h @@ -4,7 +4,7 @@ #ifndef __ASSEMBLY__ -#define __PAGETABLE_P4D_FOLDED +#define __PAGETABLE_P4D_FOLDED 1 typedef struct { pgd_t pgd; } p4d_t; diff --git a/include/asm-generic/pgtable-nopmd.h b/include/asm-generic/pgtable-nopmd.h index f35f6e8149e4..b85b8271a73d 100644 --- a/include/asm-generic/pgtable-nopmd.h +++ b/include/asm-generic/pgtable-nopmd.h @@ -8,7 +8,7 @@ struct mm_struct; -#define __PAGETABLE_PMD_FOLDED +#define __PAGETABLE_PMD_FOLDED 1 /* * Having the pmd type consist of a pud gets the size right, and allows diff --git a/include/asm-generic/pgtable-nopud.h b/include/asm-generic/pgtable-nopud.h index e950b9c50f34..9bef475db6fe 100644 --- a/include/asm-generic/pgtable-nopud.h +++ b/include/asm-generic/pgtable-nopud.h @@ -9,7 +9,7 @@ #else #include -#define __PAGETABLE_PUD_FOLDED +#define __PAGETABLE_PUD_FOLDED 1 /* * Having the pud type consist of a p4d gets the size right, and allows -- cgit v1.2.3 From 1071fc5779d9846fec56a4ff6089ab08cac1ab72 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 15 Oct 2018 10:25:57 +0200 Subject: mm: introduce mm_[p4d|pud|pmd]_folded Add three architecture overrideable functions to test if the p4d, pud, or pmd layer of a page table is folded or not. Acked-by: Kirill A. Shutemov Signed-off-by: Martin Schwidefsky --- include/asm-generic/pgtable.h | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'include') diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h index 88ebc6102c7c..15fd0277ffa6 100644 --- a/include/asm-generic/pgtable.h +++ b/include/asm-generic/pgtable.h @@ -1127,4 +1127,20 @@ static inline bool arch_has_pfn_modify_check(void) #endif #endif +/* + * On some architectures it depends on the mm if the p4d/pud or pmd + * layer of the page table hierarchy is folded or not. + */ +#ifndef mm_p4d_folded +#define mm_p4d_folded(mm) __is_defined(__PAGETABLE_P4D_FOLDED) +#endif + +#ifndef mm_pud_folded +#define mm_pud_folded(mm) __is_defined(__PAGETABLE_PUD_FOLDED) +#endif + +#ifndef mm_pmd_folded +#define mm_pmd_folded(mm) __is_defined(__PAGETABLE_PMD_FOLDED) +#endif + #endif /* _ASM_GENERIC_PGTABLE_H */ -- cgit v1.2.3 From 6d212db11947ae5464e4717536ed9faf61c01e86 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 15 Oct 2018 10:30:23 +0200 Subject: mm: add mm_pxd_folded checks to pgtable_bytes accounting functions The common mm code calls mm_dec_nr_pmds() and mm_dec_nr_puds() in free_pgtables() if the address range spans a full pud or pmd. If mm_dec_nr_puds/mm_dec_nr_pmds are non-empty due to configuration settings they blindly subtract the size of the pmd or pud table from pgtable_bytes even if the pud or pmd page table layer is folded. Add explicit mm_[pmd|pud]_folded checks to the four pgtable_bytes accounting functions mm_inc_nr_puds, mm_inc_nr_pmds, mm_dec_nr_puds and mm_dec_nr_pmds. As the check for folded page tables can be overwritten by the architecture, this allows to keep a correct pgtable_bytes value for platforms that use a dynamic number of page table levels. Acked-by: Kirill A. Shutemov Signed-off-by: Martin Schwidefsky --- include/linux/mm.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'include') diff --git a/include/linux/mm.h b/include/linux/mm.h index daa2b8f1e9a8..a3701e91bb57 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1742,11 +1742,15 @@ int __pud_alloc(struct mm_struct *mm, p4d_t *p4d, unsigned long address); static inline void mm_inc_nr_puds(struct mm_struct *mm) { + if (mm_pud_folded(mm)) + return; atomic_long_add(PTRS_PER_PUD * sizeof(pud_t), &mm->pgtables_bytes); } static inline void mm_dec_nr_puds(struct mm_struct *mm) { + if (mm_pud_folded(mm)) + return; atomic_long_sub(PTRS_PER_PUD * sizeof(pud_t), &mm->pgtables_bytes); } #endif @@ -1766,11 +1770,15 @@ int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address); static inline void mm_inc_nr_pmds(struct mm_struct *mm) { + if (mm_pmd_folded(mm)) + return; atomic_long_add(PTRS_PER_PMD * sizeof(pmd_t), &mm->pgtables_bytes); } static inline void mm_dec_nr_pmds(struct mm_struct *mm) { + if (mm_pmd_folded(mm)) + return; atomic_long_sub(PTRS_PER_PMD * sizeof(pmd_t), &mm->pgtables_bytes); } #endif -- cgit v1.2.3 From dafb7f9aef2fd44991ff1691721ff765a23be27b Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Fri, 2 Nov 2018 06:36:32 -0400 Subject: v4l2-controls: add a missing include As warned by "make headers_check", the definition for the linux-specific integer types is missing: ./usr/include/linux/v4l2-controls.h:1105: found __[us]{8,16,32,64} type without #include Fixes: c27bb30e7b6d ("media: v4l: Add definitions for MPEG-2 slice format and metadata") Reported-by: Linus Torvalds Reported-by: Stephen Rothwell Signed-off-by: Mauro Carvalho Chehab --- include/uapi/linux/v4l2-controls.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/uapi/linux/v4l2-controls.h b/include/uapi/linux/v4l2-controls.h index 51b095898f4b..86a54916206f 100644 --- a/include/uapi/linux/v4l2-controls.h +++ b/include/uapi/linux/v4l2-controls.h @@ -47,6 +47,8 @@ * videodev2.h. */ +#include + #ifndef __LINUX_V4L2_CONTROLS_H #define __LINUX_V4L2_CONTROLS_H -- cgit v1.2.3 From 47e22ff1a9e0c144611bd063b3e6135f9a269503 Mon Sep 17 00:00:00 2001 From: Radhakrishna Sripada Date: Fri, 12 Oct 2018 11:42:32 -0700 Subject: drm: Add connector property to limit max bpc MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit At times 12bpc HDMI cannot be driven due to faulty cables, dongles level shifters etc. To workaround them we may need to drive the output at a lower bpc. Currently the user space does not have a way to limit the bpc. The default bpc to be programmed is decided by the driver and is run against connector limitations. Creating a new connector property "max bpc" in order to limit the bpc. xrandr can make use of this connector property to make sure that bpc does not exceed the configured value. This property can be used by userspace to set the bpc. V2: Initialize max_bpc to satisfy kms_properties V3: Move the property to drm_connector V4: Split drm and i915 components(Ville) V5: Make the property per connector(Ville) V6: Compare the requested bpc to connector bpc(Daniel) Move the attach_property function to core(Ville) V7: Fix checkpatch warnings V8: Simplify the connector check code(Ville) V9: Const display_info(Ville) V10,V11: Fix CI issues. V12: Add the Kernel documentation(Daniel) V14: Crossreference the function name in the doc(Daniel) Cc: Ville Syrjälä Cc: Daniel Vetter Cc: Kishore Kadiyala Cc: Rodrigo Vivi Cc: Manasi Navare Cc: Stanislav Lisovskiy Cc: Sunpeng Li Acked-by: Daniel Vetter Signed-off-by: Radhakrishna Sripada Reviewed-by: Rodrigo Vivi Signed-off-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20181012184233.29250-1-radhakrishna.sripada@intel.com --- drivers/gpu/drm/drm_atomic.c | 5 +++++ drivers/gpu/drm/drm_atomic_helper.c | 4 ++++ drivers/gpu/drm/drm_atomic_uapi.c | 4 ++++ drivers/gpu/drm/drm_connector.c | 41 +++++++++++++++++++++++++++++++++++++ include/drm/drm_connector.h | 20 ++++++++++++++++++ 5 files changed, 74 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index 2870ae205237..cd8362dc4f74 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -390,6 +390,11 @@ static int drm_atomic_connector_check(struct drm_connector *connector, { struct drm_crtc_state *crtc_state; struct drm_writeback_job *writeback_job = state->writeback_job; + const struct drm_display_info *info = &connector->display_info; + + state->max_bpc = info->bpc ? info->bpc : 8; + if (connector->max_bpc_property) + state->max_bpc = min(state->max_bpc, state->max_requested_bpc); if ((connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK) || !writeback_job) return 0; diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 1cc3a045ec2f..86ac6b5b36ea 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -658,6 +658,10 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, if (old_connector_state->link_status != new_connector_state->link_status) new_crtc_state->connectors_changed = true; + + if (old_connector_state->max_requested_bpc != + new_connector_state->max_requested_bpc) + new_crtc_state->connectors_changed = true; } if (funcs->atomic_check) diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c index d5b7f315098c..86ac33922b09 100644 --- a/drivers/gpu/drm/drm_atomic_uapi.c +++ b/drivers/gpu/drm/drm_atomic_uapi.c @@ -740,6 +740,8 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector, return set_out_fence_for_connector(state->state, connector, fence_ptr); + } else if (property == connector->max_bpc_property) { + state->max_requested_bpc = val; } else if (connector->funcs->atomic_set_property) { return connector->funcs->atomic_set_property(connector, state, property, val); @@ -804,6 +806,8 @@ drm_atomic_connector_get_property(struct drm_connector *connector, *val = 0; } else if (property == config->writeback_out_fence_ptr_property) { *val = 0; + } else if (property == connector->max_bpc_property) { + *val = state->max_requested_bpc; } else if (connector->funcs->atomic_get_property) { return connector->funcs->atomic_get_property(connector, state, property, val); diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c index 4943cef178be..375cdce7df94 100644 --- a/drivers/gpu/drm/drm_connector.c +++ b/drivers/gpu/drm/drm_connector.c @@ -916,6 +916,13 @@ DRM_ENUM_NAME_FN(drm_get_content_protection_name, drm_cp_enum_list) * is no longer protected and userspace should take appropriate action * (whatever that might be). * + * max bpc: + * This range property is used by userspace to limit the bit depth. When + * used the driver would limit the bpc in accordance with the valid range + * supported by the hardware and sink. Drivers to use the function + * drm_connector_attach_max_bpc_property() to create and attach the + * property to the connector during initialization. + * * Connectors also have one standardized atomic property: * * CRTC_ID: @@ -1583,6 +1590,40 @@ void drm_connector_set_link_status_property(struct drm_connector *connector, } EXPORT_SYMBOL(drm_connector_set_link_status_property); +/** + * drm_connector_attach_max_bpc_property - attach "max bpc" property + * @connector: connector to attach max bpc property on. + * @min: The minimum bit depth supported by the connector. + * @max: The maximum bit depth supported by the connector. + * + * This is used to add support for limiting the bit depth on a connector. + * + * Returns: + * Zero on success, negative errno on failure. + */ +int drm_connector_attach_max_bpc_property(struct drm_connector *connector, + int min, int max) +{ + struct drm_device *dev = connector->dev; + struct drm_property *prop; + + prop = connector->max_bpc_property; + if (!prop) { + prop = drm_property_create_range(dev, 0, "max bpc", min, max); + if (!prop) + return -ENOMEM; + + connector->max_bpc_property = prop; + } + + drm_object_attach_property(&connector->base, prop, max); + connector->state->max_requested_bpc = max; + connector->state->max_bpc = max; + + return 0; +} +EXPORT_SYMBOL(drm_connector_attach_max_bpc_property); + /** * drm_connector_init_panel_orientation_property - * initialize the connecters panel_orientation property diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h index 9ccad6b062f2..6cf19984facb 100644 --- a/include/drm/drm_connector.h +++ b/include/drm/drm_connector.h @@ -508,6 +508,18 @@ struct drm_connector_state { * drm_writeback_signal_completion() */ struct drm_writeback_job *writeback_job; + + /** + * @max_requested_bpc: Connector property to limit the maximum bit + * depth of the pixels. + */ + u8 max_requested_bpc; + + /** + * @max_bpc: Connector max_bpc based on the requested max_bpc property + * and the connector bpc limitations obtained from edid. + */ + u8 max_bpc; }; /** @@ -973,6 +985,12 @@ struct drm_connector { */ struct drm_property_blob *path_blob_ptr; + /** + * @max_bpc_property: Default connector property for the max bpc to be + * driven out of the connector. + */ + struct drm_property *max_bpc_property; + #define DRM_CONNECTOR_POLL_HPD (1 << 0) #define DRM_CONNECTOR_POLL_CONNECT (1 << 1) #define DRM_CONNECTOR_POLL_DISCONNECT (1 << 2) @@ -1268,6 +1286,8 @@ void drm_connector_set_link_status_property(struct drm_connector *connector, uint64_t link_status); int drm_connector_init_panel_orientation_property( struct drm_connector *connector, int width, int height); +int drm_connector_attach_max_bpc_property(struct drm_connector *connector, + int min, int max); /** * struct drm_tile_group - Tile group metadata -- cgit v1.2.3 From f9005571701920551bcf54a500973fb61f2e1eda Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Wed, 31 Oct 2018 16:11:49 -0700 Subject: CONFIG_XEN_PV breaks xen_create_contiguous_region on ARM xen_create_contiguous_region has now only an implementation if CONFIG_XEN_PV is defined. However, on ARM we never set CONFIG_XEN_PV but we do have an implementation of xen_create_contiguous_region which is required for swiotlb-xen to work correctly (although it just sets *dma_handle). Cc: # 4.12 Fixes: 16624390816c ("xen: create xen_create/destroy_contiguous_region() stubs for PVHVM only builds") Signed-off-by: Stefano Stabellini Reviewed-by: Juergen Gross CC: Jeff.Kubascik@dornerworks.com CC: Jarvis.Roach@dornerworks.com CC: Nathan.Studer@dornerworks.com CC: vkuznets@redhat.com CC: boris.ostrovsky@oracle.com CC: jgross@suse.com CC: julien.grall@arm.com Signed-off-by: Juergen Gross --- include/xen/xen-ops.h | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h index 18803ff76e27..4969817124a8 100644 --- a/include/xen/xen-ops.h +++ b/include/xen/xen-ops.h @@ -42,16 +42,12 @@ int xen_setup_shutdown_event(void); extern unsigned long *xen_contiguous_bitmap; -#ifdef CONFIG_XEN_PV +#if defined(CONFIG_XEN_PV) || defined(CONFIG_ARM) || defined(CONFIG_ARM64) int xen_create_contiguous_region(phys_addr_t pstart, unsigned int order, unsigned int address_bits, dma_addr_t *dma_handle); void xen_destroy_contiguous_region(phys_addr_t pstart, unsigned int order); - -int xen_remap_pfn(struct vm_area_struct *vma, unsigned long addr, - xen_pfn_t *pfn, int nr, int *err_ptr, pgprot_t prot, - unsigned int domid, bool no_translate, struct page **pages); #else static inline int xen_create_contiguous_region(phys_addr_t pstart, unsigned int order, @@ -63,7 +59,13 @@ static inline int xen_create_contiguous_region(phys_addr_t pstart, static inline void xen_destroy_contiguous_region(phys_addr_t pstart, unsigned int order) { } +#endif +#if defined(CONFIG_XEN_PV) +int xen_remap_pfn(struct vm_area_struct *vma, unsigned long addr, + xen_pfn_t *pfn, int nr, int *err_ptr, pgprot_t prot, + unsigned int domid, bool no_translate, struct page **pages); +#else static inline int xen_remap_pfn(struct vm_area_struct *vma, unsigned long addr, xen_pfn_t *pfn, int nr, int *err_ptr, pgprot_t prot, unsigned int domid, -- cgit v1.2.3 From 857d828374cc84d07c7b94db2cf535aa517444dd Mon Sep 17 00:00:00 2001 From: Anusha Srivatsa Date: Thu, 1 Nov 2018 21:14:55 -0700 Subject: drm/dp/fec: DRM helper for Forward Error Correction DP 1.4 has Forward Error Correction Support(FEC). Add helper function to check if the sink device supports FEC. v2: Separate the helper and the code that uses the helper into two separate patches. (Manasi) v3: - Move the code to drm_dp_helper.c (Manasi) - change the return type, code style changes (Gaurav) - Use drm_dp_dpcd_readb instead of drm_dp_dpcd_read. (Jani) v4: - Avoid aux reads everytime, instead read cached values of dpcd register (jani) - Move helper to drm_dp_helper.h like other dsc helpers.(Anusha) v5: rebased. Change the helper parameter suitably. Cc: Ville Syrjala Cc: Jani Nikula Cc: Manasi Navare Acked-by: Sean Paul (For merging through drm-intel) Signed-off-by: Anusha Srivatsa Reviewed-by: Manasi Navare Signed-off-by: Manasi Navare Link: https://patchwork.freedesktop.org/patch/msgid/20181102041455.15818-2-anusha.srivatsa@intel.com --- include/drm/drm_dp_helper.h | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'include') diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h index f82e68db16e0..3e781dfb6fc1 100644 --- a/include/drm/drm_dp_helper.h +++ b/include/drm/drm_dp_helper.h @@ -1147,6 +1147,13 @@ drm_dp_dsc_sink_max_slice_width(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) DP_DSC_SLICE_WIDTH_MULTIPLIER; } +/* Forward Error Correction Support on DP 1.4 */ +static inline bool +drm_dp_sink_supports_fec(const u8 fec_capable) +{ + return fec_capable & DP_FEC_CAPABLE; +} + /* * DisplayPort AUX channel */ -- cgit v1.2.3 From 2384d02520ff2a916169b2fd85ea50e923ed56c2 Mon Sep 17 00:00:00 2001 From: Jeff Barnhill <0xeffeff@gmail.com> Date: Fri, 2 Nov 2018 20:23:57 +0000 Subject: net/ipv6: Add anycast addresses to a global hashtable MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit icmp6_send() function is expensive on systems with a large number of interfaces. Every time it’s called, it has to verify that the source address does not correspond to an existing anycast address by looping through every device and every anycast address on the device. This can result in significant delays for a CPU when there are a large number of neighbors and ND timers are frequently timing out and calling neigh_invalidate(). Add anycast addresses to a global hashtable to allow quick searching for matching anycast addresses. This is based on inet6_addr_lst in addrconf.c. Signed-off-by: Jeff Barnhill <0xeffeff@gmail.com> Signed-off-by: David S. Miller --- include/net/addrconf.h | 2 ++ include/net/if_inet6.h | 2 ++ net/ipv6/af_inet6.c | 5 ++++ net/ipv6/anycast.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++--- 4 files changed, 85 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/net/addrconf.h b/include/net/addrconf.h index 14b789a123e7..1656c5978498 100644 --- a/include/net/addrconf.h +++ b/include/net/addrconf.h @@ -317,6 +317,8 @@ bool ipv6_chk_acast_addr(struct net *net, struct net_device *dev, const struct in6_addr *addr); bool ipv6_chk_acast_addr_src(struct net *net, struct net_device *dev, const struct in6_addr *addr); +int ipv6_anycast_init(void); +void ipv6_anycast_cleanup(void); /* Device notifier */ int register_inet6addr_notifier(struct notifier_block *nb); diff --git a/include/net/if_inet6.h b/include/net/if_inet6.h index d7578cf49c3a..c9c78c15bce0 100644 --- a/include/net/if_inet6.h +++ b/include/net/if_inet6.h @@ -146,10 +146,12 @@ struct ifacaddr6 { struct in6_addr aca_addr; struct fib6_info *aca_rt; struct ifacaddr6 *aca_next; + struct hlist_node aca_addr_lst; int aca_users; refcount_t aca_refcnt; unsigned long aca_cstamp; unsigned long aca_tstamp; + struct rcu_head rcu; }; #define IFA_HOST IPV6_ADDR_LOOPBACK diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 3f4d61017a69..f0cd291034f0 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -1001,6 +1001,9 @@ static int __init inet6_init(void) err = ip6_flowlabel_init(); if (err) goto ip6_flowlabel_fail; + err = ipv6_anycast_init(); + if (err) + goto ipv6_anycast_fail; err = addrconf_init(); if (err) goto addrconf_fail; @@ -1091,6 +1094,8 @@ ipv6_frag_fail: ipv6_exthdrs_fail: addrconf_cleanup(); addrconf_fail: + ipv6_anycast_cleanup(); +ipv6_anycast_fail: ip6_flowlabel_cleanup(); ip6_flowlabel_fail: ndisc_late_cleanup(); diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c index 4e0ff7031edd..7698637cf827 100644 --- a/net/ipv6/anycast.c +++ b/net/ipv6/anycast.c @@ -44,8 +44,22 @@ #include +#define IN6_ADDR_HSIZE_SHIFT 8 +#define IN6_ADDR_HSIZE BIT(IN6_ADDR_HSIZE_SHIFT) +/* anycast address hash table + */ +static struct hlist_head inet6_acaddr_lst[IN6_ADDR_HSIZE]; +static DEFINE_SPINLOCK(acaddr_hash_lock); + static int ipv6_dev_ac_dec(struct net_device *dev, const struct in6_addr *addr); +static u32 inet6_acaddr_hash(struct net *net, const struct in6_addr *addr) +{ + u32 val = ipv6_addr_hash(addr) ^ net_hash_mix(net); + + return hash_32(val, IN6_ADDR_HSIZE_SHIFT); +} + /* * socket join an anycast group */ @@ -204,16 +218,39 @@ void ipv6_sock_ac_close(struct sock *sk) rtnl_unlock(); } +static void ipv6_add_acaddr_hash(struct net *net, struct ifacaddr6 *aca) +{ + unsigned int hash = inet6_acaddr_hash(net, &aca->aca_addr); + + spin_lock(&acaddr_hash_lock); + hlist_add_head_rcu(&aca->aca_addr_lst, &inet6_acaddr_lst[hash]); + spin_unlock(&acaddr_hash_lock); +} + +static void ipv6_del_acaddr_hash(struct ifacaddr6 *aca) +{ + spin_lock(&acaddr_hash_lock); + hlist_del_init_rcu(&aca->aca_addr_lst); + spin_unlock(&acaddr_hash_lock); +} + static void aca_get(struct ifacaddr6 *aca) { refcount_inc(&aca->aca_refcnt); } +static void aca_free_rcu(struct rcu_head *h) +{ + struct ifacaddr6 *aca = container_of(h, struct ifacaddr6, rcu); + + fib6_info_release(aca->aca_rt); + kfree(aca); +} + static void aca_put(struct ifacaddr6 *ac) { if (refcount_dec_and_test(&ac->aca_refcnt)) { - fib6_info_release(ac->aca_rt); - kfree(ac); + call_rcu(&ac->rcu, aca_free_rcu); } } @@ -229,6 +266,7 @@ static struct ifacaddr6 *aca_alloc(struct fib6_info *f6i, aca->aca_addr = *addr; fib6_info_hold(f6i); aca->aca_rt = f6i; + INIT_HLIST_NODE(&aca->aca_addr_lst); aca->aca_users = 1; /* aca_tstamp should be updated upon changes */ aca->aca_cstamp = aca->aca_tstamp = jiffies; @@ -285,6 +323,8 @@ int __ipv6_dev_ac_inc(struct inet6_dev *idev, const struct in6_addr *addr) aca_get(aca); write_unlock_bh(&idev->lock); + ipv6_add_acaddr_hash(net, aca); + ip6_ins_rt(net, f6i); addrconf_join_solict(idev->dev, &aca->aca_addr); @@ -325,6 +365,7 @@ int __ipv6_dev_ac_dec(struct inet6_dev *idev, const struct in6_addr *addr) else idev->ac_list = aca->aca_next; write_unlock_bh(&idev->lock); + ipv6_del_acaddr_hash(aca); addrconf_leave_solict(idev, &aca->aca_addr); ip6_del_rt(dev_net(idev->dev), aca->aca_rt); @@ -352,6 +393,8 @@ void ipv6_ac_destroy_dev(struct inet6_dev *idev) idev->ac_list = aca->aca_next; write_unlock_bh(&idev->lock); + ipv6_del_acaddr_hash(aca); + addrconf_leave_solict(idev, &aca->aca_addr); ip6_del_rt(dev_net(idev->dev), aca->aca_rt); @@ -390,17 +433,25 @@ static bool ipv6_chk_acast_dev(struct net_device *dev, const struct in6_addr *ad bool ipv6_chk_acast_addr(struct net *net, struct net_device *dev, const struct in6_addr *addr) { + unsigned int hash = inet6_acaddr_hash(net, addr); + struct net_device *nh_dev; + struct ifacaddr6 *aca; bool found = false; rcu_read_lock(); if (dev) found = ipv6_chk_acast_dev(dev, addr); else - for_each_netdev_rcu(net, dev) - if (ipv6_chk_acast_dev(dev, addr)) { + hlist_for_each_entry_rcu(aca, &inet6_acaddr_lst[hash], + aca_addr_lst) { + nh_dev = fib6_info_nh_dev(aca->aca_rt); + if (!nh_dev || !net_eq(dev_net(nh_dev), net)) + continue; + if (ipv6_addr_equal(&aca->aca_addr, addr)) { found = true; break; } + } rcu_read_unlock(); return found; } @@ -539,4 +590,25 @@ void ac6_proc_exit(struct net *net) { remove_proc_entry("anycast6", net->proc_net); } + +/* Init / cleanup code + */ +int __init ipv6_anycast_init(void) +{ + int i; + + for (i = 0; i < IN6_ADDR_HSIZE; i++) + INIT_HLIST_HEAD(&inet6_acaddr_lst[i]); + return 0; +} + +void ipv6_anycast_cleanup(void) +{ + int i; + + spin_lock(&acaddr_hash_lock); + for (i = 0; i < IN6_ADDR_HSIZE; i++) + WARN_ON(!hlist_empty(&inet6_acaddr_lst[i])); + spin_unlock(&acaddr_hash_lock); +} #endif -- cgit v1.2.3 From 17b8b74c0f8dbf9b9e3301f9ca5b65dd1c079951 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Fri, 19 Oct 2018 19:35:19 +0200 Subject: netfilter: ipset: Correct rcu_dereference() call in ip_set_put_comment() The function is called when rcu_read_lock() is held and not when rcu_read_lock_bh() is held. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/ipset/ip_set_comment.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/ipset/ip_set_comment.h b/include/linux/netfilter/ipset/ip_set_comment.h index 8e2bab1e8e90..70877f8de7e9 100644 --- a/include/linux/netfilter/ipset/ip_set_comment.h +++ b/include/linux/netfilter/ipset/ip_set_comment.h @@ -43,11 +43,11 @@ ip_set_init_comment(struct ip_set *set, struct ip_set_comment *comment, rcu_assign_pointer(comment->c, c); } -/* Used only when dumping a set, protected by rcu_read_lock_bh() */ +/* Used only when dumping a set, protected by rcu_read_lock() */ static inline int ip_set_put_comment(struct sk_buff *skb, const struct ip_set_comment *comment) { - struct ip_set_comment_rcu *c = rcu_dereference_bh(comment->c); + struct ip_set_comment_rcu *c = rcu_dereference(comment->c); if (!c) return 0; -- cgit v1.2.3 From a95a7774d51e13f9cf4b7285666829b68852f07a Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 2 Nov 2018 00:11:34 +0100 Subject: netfilter: conntrack: add nf_{tcp,udp,sctp,icmp,dccp,icmpv6,generic}_pernet() Expose these functions to access conntrack protocol tracker netns area, nfnetlink_cttimeout needs this. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_conntrack_l4proto.h | 39 ++++++++++++++++++++++++++++ net/netfilter/nf_conntrack_proto_dccp.c | 13 +++------- net/netfilter/nf_conntrack_proto_generic.c | 11 +++----- net/netfilter/nf_conntrack_proto_icmp.c | 11 +++----- net/netfilter/nf_conntrack_proto_icmpv6.c | 11 +++----- net/netfilter/nf_conntrack_proto_sctp.c | 11 +++----- net/netfilter/nf_conntrack_proto_tcp.c | 15 ++++------- net/netfilter/nf_conntrack_proto_udp.c | 11 +++----- 8 files changed, 63 insertions(+), 59 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h index eed04af9b75e..ae7b86f587f2 100644 --- a/include/net/netfilter/nf_conntrack_l4proto.h +++ b/include/net/netfilter/nf_conntrack_l4proto.h @@ -153,4 +153,43 @@ void nf_ct_l4proto_log_invalid(const struct sk_buff *skb, const char *fmt, ...) { } #endif /* CONFIG_SYSCTL */ +static inline struct nf_generic_net *nf_generic_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.generic; +} + +static inline struct nf_tcp_net *nf_tcp_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.tcp; +} + +static inline struct nf_udp_net *nf_udp_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.udp; +} + +static inline struct nf_icmp_net *nf_icmp_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.icmp; +} + +static inline struct nf_icmp_net *nf_icmpv6_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.icmpv6; +} + +#ifdef CONFIG_NF_CT_PROTO_DCCP +static inline struct nf_dccp_net *nf_dccp_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.dccp; +} +#endif + +#ifdef CONFIG_NF_CT_PROTO_SCTP +static inline struct nf_sctp_net *nf_sctp_pernet(struct net *net) +{ + return &net->ct.nf_ct_proto.sctp; +} +#endif + #endif /*_NF_CONNTRACK_PROTOCOL_H*/ diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 171e9e122e5f..023c1445bc39 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -384,11 +384,6 @@ dccp_state_table[CT_DCCP_ROLE_MAX + 1][DCCP_PKT_SYNCACK + 1][CT_DCCP_MAX + 1] = }, }; -static inline struct nf_dccp_net *dccp_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.dccp; -} - static noinline bool dccp_new(struct nf_conn *ct, const struct sk_buff *skb, const struct dccp_hdr *dh) @@ -401,7 +396,7 @@ dccp_new(struct nf_conn *ct, const struct sk_buff *skb, state = dccp_state_table[CT_DCCP_ROLE_CLIENT][dh->dccph_type][CT_DCCP_NONE]; switch (state) { default: - dn = dccp_pernet(net); + dn = nf_dccp_pernet(net); if (dn->dccp_loose == 0) { msg = "not picking up existing connection "; goto out_invalid; @@ -568,7 +563,7 @@ static int dccp_packet(struct nf_conn *ct, struct sk_buff *skb, timeouts = nf_ct_timeout_lookup(ct); if (!timeouts) - timeouts = dccp_pernet(nf_ct_net(ct))->dccp_timeout; + timeouts = nf_dccp_pernet(nf_ct_net(ct))->dccp_timeout; nf_ct_refresh_acct(ct, ctinfo, skb, timeouts[new_state]); return NF_ACCEPT; @@ -681,7 +676,7 @@ static int nlattr_to_dccp(struct nlattr *cda[], struct nf_conn *ct) static int dccp_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { - struct nf_dccp_net *dn = dccp_pernet(net); + struct nf_dccp_net *dn = nf_dccp_pernet(net); unsigned int *timeouts = data; int i; @@ -814,7 +809,7 @@ static int dccp_kmemdup_sysctl_table(struct net *net, struct nf_proto_net *pn, static int dccp_init_net(struct net *net) { - struct nf_dccp_net *dn = dccp_pernet(net); + struct nf_dccp_net *dn = nf_dccp_pernet(net); struct nf_proto_net *pn = &dn->pn; if (!pn->users) { diff --git a/net/netfilter/nf_conntrack_proto_generic.c b/net/netfilter/nf_conntrack_proto_generic.c index e10e867e0b55..5da19d5fbc76 100644 --- a/net/netfilter/nf_conntrack_proto_generic.c +++ b/net/netfilter/nf_conntrack_proto_generic.c @@ -27,11 +27,6 @@ static bool nf_generic_should_process(u8 proto) } } -static inline struct nf_generic_net *generic_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.generic; -} - static bool generic_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff, struct net *net, struct nf_conntrack_tuple *tuple) @@ -58,7 +53,7 @@ static int generic_packet(struct nf_conn *ct, } if (!timeout) - timeout = &generic_pernet(nf_ct_net(ct))->timeout; + timeout = &nf_generic_pernet(nf_ct_net(ct))->timeout; nf_ct_refresh_acct(ct, ctinfo, skb, *timeout); return NF_ACCEPT; @@ -72,7 +67,7 @@ static int generic_packet(struct nf_conn *ct, static int generic_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { - struct nf_generic_net *gn = generic_pernet(net); + struct nf_generic_net *gn = nf_generic_pernet(net); unsigned int *timeout = data; if (!timeout) @@ -138,7 +133,7 @@ static int generic_kmemdup_sysctl_table(struct nf_proto_net *pn, static int generic_init_net(struct net *net) { - struct nf_generic_net *gn = generic_pernet(net); + struct nf_generic_net *gn = nf_generic_pernet(net); struct nf_proto_net *pn = &gn->pn; gn->timeout = nf_ct_generic_timeout; diff --git a/net/netfilter/nf_conntrack_proto_icmp.c b/net/netfilter/nf_conntrack_proto_icmp.c index 3598520bd19b..de64d8a5fdfd 100644 --- a/net/netfilter/nf_conntrack_proto_icmp.c +++ b/net/netfilter/nf_conntrack_proto_icmp.c @@ -25,11 +25,6 @@ static const unsigned int nf_ct_icmp_timeout = 30*HZ; -static inline struct nf_icmp_net *icmp_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.icmp; -} - static bool icmp_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff, struct net *net, struct nf_conntrack_tuple *tuple) { @@ -103,7 +98,7 @@ static int icmp_packet(struct nf_conn *ct, } if (!timeout) - timeout = &icmp_pernet(nf_ct_net(ct))->timeout; + timeout = &nf_icmp_pernet(nf_ct_net(ct))->timeout; nf_ct_refresh_acct(ct, ctinfo, skb, *timeout); return NF_ACCEPT; @@ -275,7 +270,7 @@ static int icmp_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { unsigned int *timeout = data; - struct nf_icmp_net *in = icmp_pernet(net); + struct nf_icmp_net *in = nf_icmp_pernet(net); if (tb[CTA_TIMEOUT_ICMP_TIMEOUT]) { if (!timeout) @@ -337,7 +332,7 @@ static int icmp_kmemdup_sysctl_table(struct nf_proto_net *pn, static int icmp_init_net(struct net *net) { - struct nf_icmp_net *in = icmp_pernet(net); + struct nf_icmp_net *in = nf_icmp_pernet(net); struct nf_proto_net *pn = &in->pn; in->timeout = nf_ct_icmp_timeout; diff --git a/net/netfilter/nf_conntrack_proto_icmpv6.c b/net/netfilter/nf_conntrack_proto_icmpv6.c index 378618feed5d..a15eefb8e317 100644 --- a/net/netfilter/nf_conntrack_proto_icmpv6.c +++ b/net/netfilter/nf_conntrack_proto_icmpv6.c @@ -30,11 +30,6 @@ static const unsigned int nf_ct_icmpv6_timeout = 30*HZ; -static inline struct nf_icmp_net *icmpv6_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.icmpv6; -} - static bool icmpv6_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff, struct net *net, @@ -87,7 +82,7 @@ static bool icmpv6_invert_tuple(struct nf_conntrack_tuple *tuple, static unsigned int *icmpv6_get_timeouts(struct net *net) { - return &icmpv6_pernet(net)->timeout; + return &nf_icmpv6_pernet(net)->timeout; } /* Returns verdict for packet, or -1 for invalid. */ @@ -286,7 +281,7 @@ static int icmpv6_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { unsigned int *timeout = data; - struct nf_icmp_net *in = icmpv6_pernet(net); + struct nf_icmp_net *in = nf_icmpv6_pernet(net); if (!timeout) timeout = icmpv6_get_timeouts(net); @@ -348,7 +343,7 @@ static int icmpv6_kmemdup_sysctl_table(struct nf_proto_net *pn, static int icmpv6_init_net(struct net *net) { - struct nf_icmp_net *in = icmpv6_pernet(net); + struct nf_icmp_net *in = nf_icmpv6_pernet(net); struct nf_proto_net *pn = &in->pn; in->timeout = nf_ct_icmpv6_timeout; diff --git a/net/netfilter/nf_conntrack_proto_sctp.c b/net/netfilter/nf_conntrack_proto_sctp.c index 3d719d3eb9a3..d53e3e78f605 100644 --- a/net/netfilter/nf_conntrack_proto_sctp.c +++ b/net/netfilter/nf_conntrack_proto_sctp.c @@ -146,11 +146,6 @@ static const u8 sctp_conntracks[2][11][SCTP_CONNTRACK_MAX] = { } }; -static inline struct nf_sctp_net *sctp_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.sctp; -} - #ifdef CONFIG_NF_CONNTRACK_PROCFS /* Print out the private part of the conntrack. */ static void sctp_print_conntrack(struct seq_file *s, struct nf_conn *ct) @@ -480,7 +475,7 @@ static int sctp_packet(struct nf_conn *ct, timeouts = nf_ct_timeout_lookup(ct); if (!timeouts) - timeouts = sctp_pernet(nf_ct_net(ct))->timeouts; + timeouts = nf_sctp_pernet(nf_ct_net(ct))->timeouts; nf_ct_refresh_acct(ct, ctinfo, skb, timeouts[new_state]); @@ -599,7 +594,7 @@ static int sctp_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { unsigned int *timeouts = data; - struct nf_sctp_net *sn = sctp_pernet(net); + struct nf_sctp_net *sn = nf_sctp_pernet(net); int i; /* set default SCTP timeouts. */ @@ -736,7 +731,7 @@ static int sctp_kmemdup_sysctl_table(struct nf_proto_net *pn, static int sctp_init_net(struct net *net) { - struct nf_sctp_net *sn = sctp_pernet(net); + struct nf_sctp_net *sn = nf_sctp_pernet(net); struct nf_proto_net *pn = &sn->pn; if (!pn->users) { diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index 1bcf9984d45e..4dcbd51a8e97 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -272,11 +272,6 @@ static const u8 tcp_conntracks[2][6][TCP_CONNTRACK_MAX] = { } }; -static inline struct nf_tcp_net *tcp_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.tcp; -} - #ifdef CONFIG_NF_CONNTRACK_PROCFS /* Print out the private part of the conntrack. */ static void tcp_print_conntrack(struct seq_file *s, struct nf_conn *ct) @@ -475,7 +470,7 @@ static bool tcp_in_window(const struct nf_conn *ct, const struct tcphdr *tcph) { struct net *net = nf_ct_net(ct); - struct nf_tcp_net *tn = tcp_pernet(net); + struct nf_tcp_net *tn = nf_tcp_pernet(net); struct ip_ct_tcp_state *sender = &state->seen[dir]; struct ip_ct_tcp_state *receiver = &state->seen[!dir]; const struct nf_conntrack_tuple *tuple = &ct->tuplehash[dir].tuple; @@ -767,7 +762,7 @@ static noinline bool tcp_new(struct nf_conn *ct, const struct sk_buff *skb, { enum tcp_conntrack new_state; struct net *net = nf_ct_net(ct); - const struct nf_tcp_net *tn = tcp_pernet(net); + const struct nf_tcp_net *tn = nf_tcp_pernet(net); const struct ip_ct_tcp_state *sender = &ct->proto.tcp.seen[0]; const struct ip_ct_tcp_state *receiver = &ct->proto.tcp.seen[1]; @@ -841,7 +836,7 @@ static int tcp_packet(struct nf_conn *ct, const struct nf_hook_state *state) { struct net *net = nf_ct_net(ct); - struct nf_tcp_net *tn = tcp_pernet(net); + struct nf_tcp_net *tn = nf_tcp_pernet(net); struct nf_conntrack_tuple *tuple; enum tcp_conntrack new_state, old_state; unsigned int index, *timeouts; @@ -1283,7 +1278,7 @@ static unsigned int tcp_nlattr_tuple_size(void) static int tcp_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { - struct nf_tcp_net *tn = tcp_pernet(net); + struct nf_tcp_net *tn = nf_tcp_pernet(net); unsigned int *timeouts = data; int i; @@ -1508,7 +1503,7 @@ static int tcp_kmemdup_sysctl_table(struct nf_proto_net *pn, static int tcp_init_net(struct net *net) { - struct nf_tcp_net *tn = tcp_pernet(net); + struct nf_tcp_net *tn = nf_tcp_pernet(net); struct nf_proto_net *pn = &tn->pn; if (!pn->users) { diff --git a/net/netfilter/nf_conntrack_proto_udp.c b/net/netfilter/nf_conntrack_proto_udp.c index a7aa70370913..c879d8d78cfd 100644 --- a/net/netfilter/nf_conntrack_proto_udp.c +++ b/net/netfilter/nf_conntrack_proto_udp.c @@ -32,14 +32,9 @@ static const unsigned int udp_timeouts[UDP_CT_MAX] = { [UDP_CT_REPLIED] = 180*HZ, }; -static inline struct nf_udp_net *udp_pernet(struct net *net) -{ - return &net->ct.nf_ct_proto.udp; -} - static unsigned int *udp_get_timeouts(struct net *net) { - return udp_pernet(net)->timeouts; + return nf_udp_pernet(net)->timeouts; } static void udp_error_log(const struct sk_buff *skb, @@ -212,7 +207,7 @@ static int udp_timeout_nlattr_to_obj(struct nlattr *tb[], struct net *net, void *data) { unsigned int *timeouts = data; - struct nf_udp_net *un = udp_pernet(net); + struct nf_udp_net *un = nf_udp_pernet(net); if (!timeouts) timeouts = un->timeouts; @@ -292,7 +287,7 @@ static int udp_kmemdup_sysctl_table(struct nf_proto_net *pn, static int udp_init_net(struct net *net) { - struct nf_udp_net *un = udp_pernet(net); + struct nf_udp_net *un = nf_udp_pernet(net); struct nf_proto_net *pn = &un->pn; if (!pn->users) { -- cgit v1.2.3 From 3e59020abf0f88182730527ee5b862e786eb485a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 31 Oct 2018 08:39:12 -0700 Subject: net: bql: add __netdev_tx_sent_queue() When qdisc_run() tries to use BQL budget to bulk-dequeue a batch of packets, GSO can later transform this list in another list of skbs, and each skb is sent to device ndo_start_xmit(), one at a time, with skb->xmit_more being set to one but for last skb. Problem is that very often, BQL limit is hit in the middle of the packet train, forcing dev_hard_start_xmit() to stop the bulk send and requeue the end of the list. BQL role is to avoid head of line blocking, making sure a qdisc can deliver high priority packets before low priority ones. But there is no way requeued packets can be bypassed by fresh packets in the qdisc. Aborting the bulk send increases TX softirqs, and hot cache lines (after skb_segment()) are wasted. Note that for TSO packets, we never split a packet in the middle because of BQL limit being hit. Drivers should be able to update BQL counters without flipping/caring about BQL status, if the current skb has xmit_more set. Upper layers are ultimately responsible to stop sending another packet train when BQL limit is hit. Code template in a driver might look like the following : send_doorbell = __netdev_tx_sent_queue(tx_queue, nr_bytes, skb->xmit_more); Note that __netdev_tx_sent_queue() use is not mandatory, since following patch will change dev_hard_start_xmit() to not care about BQL status. But it is highly recommended so that xmit_more full benefits can be reached (less doorbells sent, and less atomic operations as well) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/netdevice.h | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index dc1d9ed33b31..857f8abf7b91 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -3190,6 +3190,26 @@ static inline void netdev_tx_sent_queue(struct netdev_queue *dev_queue, #endif } +/* Variant of netdev_tx_sent_queue() for drivers that are aware + * that they should not test BQL status themselves. + * We do want to change __QUEUE_STATE_STACK_XOFF only for the last + * skb of a batch. + * Returns true if the doorbell must be used to kick the NIC. + */ +static inline bool __netdev_tx_sent_queue(struct netdev_queue *dev_queue, + unsigned int bytes, + bool xmit_more) +{ + if (xmit_more) { +#ifdef CONFIG_BQL + dql_queued(&dev_queue->dql, bytes); +#endif + return netif_tx_queue_stopped(dev_queue); + } + netdev_tx_sent_queue(dev_queue, bytes); + return true; +} + /** * netdev_sent_queue - report the number of bytes queued to hardware * @dev: network device -- cgit v1.2.3 From fd82d61ba142f0b83463e47064bf5460aac57b6e Mon Sep 17 00:00:00 2001 From: Xin Long Date: Sat, 3 Nov 2018 13:59:45 +0800 Subject: sctp: fix strchange_flags name for Stream Change Event As defined in rfc6525#section-6.1.3, SCTP_STREAM_CHANGE_DENIED and SCTP_STREAM_CHANGE_FAILED should be used instead of SCTP_ASSOC_CHANGE_DENIED and SCTP_ASSOC_CHANGE_FAILED. To keep the compatibility, fix it by adding two macros. Fixes: b444153fb5a6 ("sctp: add support for generating add stream change event notification") Reported-by: Jianwen Ji Signed-off-by: Xin Long Signed-off-by: David S. Miller --- include/uapi/linux/sctp.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/uapi/linux/sctp.h b/include/uapi/linux/sctp.h index 34dd3d497f2c..680ecc3bf2a9 100644 --- a/include/uapi/linux/sctp.h +++ b/include/uapi/linux/sctp.h @@ -568,6 +568,8 @@ struct sctp_assoc_reset_event { #define SCTP_ASSOC_CHANGE_DENIED 0x0004 #define SCTP_ASSOC_CHANGE_FAILED 0x0008 +#define SCTP_STREAM_CHANGE_DENIED SCTP_ASSOC_CHANGE_DENIED +#define SCTP_STREAM_CHANGE_FAILED SCTP_ASSOC_CHANGE_FAILED struct sctp_stream_change_event { __u16 strchange_type; __u16 strchange_flags; -- cgit v1.2.3 From 12480e3b16982c4026de10dd8155823219cd6391 Mon Sep 17 00:00:00 2001 From: Xin Long Date: Sat, 3 Nov 2018 14:01:31 +0800 Subject: sctp: define SCTP_SS_DEFAULT for Stream schedulers According to rfc8260#section-4.3.2, SCTP_SS_DEFAULT is required to defined as SCTP_SS_FCFS or SCTP_SS_RR. SCTP_SS_FCFS is used for SCTP_SS_DEFAULT's value in this patch. Fixes: 5bbbbe32a431 ("sctp: introduce stream scheduler foundations") Reported-by: Jianwen Ji Signed-off-by: Xin Long Signed-off-by: David S. Miller --- include/uapi/linux/sctp.h | 1 + net/sctp/outqueue.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/uapi/linux/sctp.h b/include/uapi/linux/sctp.h index 680ecc3bf2a9..c81feb373d3e 100644 --- a/include/uapi/linux/sctp.h +++ b/include/uapi/linux/sctp.h @@ -1153,6 +1153,7 @@ struct sctp_add_streams { /* SCTP Stream schedulers */ enum sctp_sched_type { SCTP_SS_FCFS, + SCTP_SS_DEFAULT = SCTP_SS_FCFS, SCTP_SS_PRIO, SCTP_SS_RR, SCTP_SS_MAX = SCTP_SS_RR diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c index 9cb854b05342..c37e1c2dec9d 100644 --- a/net/sctp/outqueue.c +++ b/net/sctp/outqueue.c @@ -212,7 +212,7 @@ void sctp_outq_init(struct sctp_association *asoc, struct sctp_outq *q) INIT_LIST_HEAD(&q->retransmit); INIT_LIST_HEAD(&q->sacked); INIT_LIST_HEAD(&q->abandoned); - sctp_sched_set_sched(asoc, SCTP_SS_FCFS); + sctp_sched_set_sched(asoc, SCTP_SS_DEFAULT); } /* Free the outqueue structure and any related pending chunks. -- cgit v1.2.3 From 35b69a420bfb56b7b74cb635ea903db05e357bec Mon Sep 17 00:00:00 2001 From: Michael Kelley Date: Sun, 4 Nov 2018 03:48:54 +0000 Subject: clockevents/drivers/i8253: Add support for PIT shutdown quirk Add support for platforms where pit_shutdown() doesn't work because of a quirk in the PIT emulation. On these platforms setting the counter register to zero causes the PIT to start running again, negating the shutdown. Provide a global variable that controls whether the counter register is zero'ed, which platform specific code can override. Signed-off-by: Michael Kelley Signed-off-by: Thomas Gleixner Cc: "gregkh@linuxfoundation.org" Cc: "devel@linuxdriverproject.org" Cc: "daniel.lezcano@linaro.org" Cc: "virtualization@lists.linux-foundation.org" Cc: "jgross@suse.com" Cc: "akataria@vmware.com" Cc: "olaf@aepfle.de" Cc: "apw@canonical.com" Cc: vkuznets Cc: "jasowang@redhat.com" Cc: "marcelo.cerri@canonical.com" Cc: KY Srinivasan Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/1541303219-11142-2-git-send-email-mikelley@microsoft.com --- drivers/clocksource/i8253.c | 14 ++++++++++++-- include/linux/i8253.h | 1 + 2 files changed, 13 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/clocksource/i8253.c b/drivers/clocksource/i8253.c index 9c38895542f4..d4350bb10b83 100644 --- a/drivers/clocksource/i8253.c +++ b/drivers/clocksource/i8253.c @@ -20,6 +20,13 @@ DEFINE_RAW_SPINLOCK(i8253_lock); EXPORT_SYMBOL(i8253_lock); +/* + * Handle PIT quirk in pit_shutdown() where zeroing the counter register + * restarts the PIT, negating the shutdown. On platforms with the quirk, + * platform specific code can set this to false. + */ +bool i8253_clear_counter_on_shutdown __ro_after_init = true; + #ifdef CONFIG_CLKSRC_I8253 /* * Since the PIT overflows every tick, its not very useful @@ -109,8 +116,11 @@ static int pit_shutdown(struct clock_event_device *evt) raw_spin_lock(&i8253_lock); outb_p(0x30, PIT_MODE); - outb_p(0, PIT_CH0); - outb_p(0, PIT_CH0); + + if (i8253_clear_counter_on_shutdown) { + outb_p(0, PIT_CH0); + outb_p(0, PIT_CH0); + } raw_spin_unlock(&i8253_lock); return 0; diff --git a/include/linux/i8253.h b/include/linux/i8253.h index e6bb36a97519..8336b2f6f834 100644 --- a/include/linux/i8253.h +++ b/include/linux/i8253.h @@ -21,6 +21,7 @@ #define PIT_LATCH ((PIT_TICK_RATE + HZ/2) / HZ) extern raw_spinlock_t i8253_lock; +extern bool i8253_clear_counter_on_shutdown; extern struct clock_event_device i8253_clockevent; extern void clockevent_i8253_init(bool oneshot); -- cgit v1.2.3 From d098093ba06eb032057d1aca1c2e45889e099d00 Mon Sep 17 00:00:00 2001 From: Boris Brezillon Date: Sun, 28 Oct 2018 12:29:55 +0100 Subject: mtd: nand: Fix nanddev_neraseblocks() nanddev_neraseblocks() currently returns the number pages per LUN instead of the total number of eraseblocks. Fixes: 9c3736a3de21 ("mtd: nand: Add core infrastructure to deal with NAND devices") Cc: Signed-off-by: Boris Brezillon Reviewed-by: Miquel Raynal --- include/linux/mtd/nand.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/mtd/nand.h b/include/linux/mtd/nand.h index abe975c87b90..78b86dea2f29 100644 --- a/include/linux/mtd/nand.h +++ b/include/linux/mtd/nand.h @@ -324,9 +324,8 @@ static inline unsigned int nanddev_ntargets(const struct nand_device *nand) */ static inline unsigned int nanddev_neraseblocks(const struct nand_device *nand) { - return (u64)nand->memorg.luns_per_target * - nand->memorg.eraseblocks_per_lun * - nand->memorg.pages_per_eraseblock; + return nand->memorg.ntargets * nand->memorg.luns_per_target * + nand->memorg.eraseblocks_per_lun; } /** -- cgit v1.2.3 From 163c8d54a997153ee1a1e07fcac087492ad85b37 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 5 Nov 2018 07:36:28 +0100 Subject: compiler: remove __no_sanitize_address_or_inline again The __no_sanitize_address_or_inline and __no_kasan_or_inline defines are almost identical. The only difference is that __no_kasan_or_inline does not have the 'notrace' attribute. To be able to replace __no_sanitize_address_or_inline with the older definition, add 'notrace' to __no_kasan_or_inline and change to two users of __no_sanitize_address_or_inline in the s390 code. The 'notrace' option is necessary for e.g. the __load_psw_mask function in arch/s390/include/asm/processor.h. Without the option it is possible to trace __load_psw_mask which leads to kernel stack overflow. Signed-off-by: Martin Schwidefsky Pointed-out-by: Andrey Ryabinin Acked-by: Steven Rostedt (VMware) Signed-off-by: Linus Torvalds --- arch/s390/include/asm/processor.h | 4 ++-- include/linux/compiler-gcc.h | 12 ------------ include/linux/compiler.h | 2 +- 3 files changed, 3 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h index 302795c47c06..81038ab357ce 100644 --- a/arch/s390/include/asm/processor.h +++ b/arch/s390/include/asm/processor.h @@ -236,7 +236,7 @@ static inline unsigned long current_stack_pointer(void) return sp; } -static __no_sanitize_address_or_inline unsigned short stap(void) +static __no_kasan_or_inline unsigned short stap(void) { unsigned short cpu_address; @@ -330,7 +330,7 @@ static inline void __load_psw(psw_t psw) * Set PSW mask to specified value, while leaving the * PSW addr pointing to the next instruction. */ -static __no_sanitize_address_or_inline void __load_psw_mask(unsigned long mask) +static __no_kasan_or_inline void __load_psw_mask(unsigned long mask) { unsigned long addr; psw_t psw; diff --git a/include/linux/compiler-gcc.h b/include/linux/compiler-gcc.h index c0f5db3a9621..2010493e1040 100644 --- a/include/linux/compiler-gcc.h +++ b/include/linux/compiler-gcc.h @@ -143,18 +143,6 @@ #define KASAN_ABI_VERSION 3 #endif -/* - * Because __no_sanitize_address conflicts with inlining: - * https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67368 - * we do one or the other. - */ -#ifdef CONFIG_KASAN -#define __no_sanitize_address_or_inline \ - __no_sanitize_address __maybe_unused notrace -#else -#define __no_sanitize_address_or_inline inline -#endif - #if GCC_VERSION >= 50100 #define COMPILER_HAS_GENERIC_BUILTIN_OVERFLOW 1 #endif diff --git a/include/linux/compiler.h b/include/linux/compiler.h index 18c80cfa4fc4..06396c1cf127 100644 --- a/include/linux/compiler.h +++ b/include/linux/compiler.h @@ -189,7 +189,7 @@ void __read_once_size(const volatile void *p, void *res, int size) * https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67368 * '__maybe_unused' allows us to avoid defined-but-not-used warnings. */ -# define __no_kasan_or_inline __no_sanitize_address __maybe_unused +# define __no_kasan_or_inline __no_sanitize_address notrace __maybe_unused #else # define __no_kasan_or_inline __always_inline #endif -- cgit v1.2.3 From e55a5c9b5f5b80275a38293ac0fd38336dd2efdf Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Tue, 16 Oct 2018 10:04:08 +0200 Subject: drm/ttm: Rename ttm_bo_global_{init,release}() to ttm_bo_global_ref_{,}() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The functions ttm_bo_global_init() and ttm_bo_global_release() do not receive an argument of type struct ttm_bo_global. Both take a struct drm_global_reference that contains points to a struct ttm_bo_global_ref. Renaming them reflects this. Signed-off-by: Thomas Zimmermann Reviewed-by: Christian König Signed-off-by: Alex Deucher --- Documentation/gpu/drm-mm.rst | 4 ++-- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 4 ++-- drivers/gpu/drm/ast/ast_ttm.c | 4 ++-- drivers/gpu/drm/bochs/bochs_mm.c | 4 ++-- drivers/gpu/drm/cirrus/cirrus_ttm.c | 4 ++-- drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c | 4 ++-- drivers/gpu/drm/mgag200/mgag200_ttm.c | 4 ++-- drivers/gpu/drm/nouveau/nouveau_ttm.c | 4 ++-- drivers/gpu/drm/qxl/qxl_ttm.c | 4 ++-- drivers/gpu/drm/radeon/radeon_ttm.c | 4 ++-- drivers/gpu/drm/ttm/ttm_bo.c | 8 ++++---- drivers/gpu/drm/virtio/virtgpu_ttm.c | 4 ++-- drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c | 4 ++-- drivers/staging/vboxvideo/vbox_ttm.c | 4 ++-- include/drm/ttm/ttm_bo_driver.h | 4 ++-- 15 files changed, 32 insertions(+), 32 deletions(-) (limited to 'include') diff --git a/Documentation/gpu/drm-mm.rst b/Documentation/gpu/drm-mm.rst index e725e8449e72..d0f3c6b03200 100644 --- a/Documentation/gpu/drm-mm.rst +++ b/Documentation/gpu/drm-mm.rst @@ -72,8 +72,8 @@ object TTM to provide a pool for buffer object allocation by clients and the kernel itself. The type of this object should be TTM_GLOBAL_TTM_BO, and its size should be sizeof(struct ttm_bo_global). Again, driver-specific init and release functions may -be provided, likely eventually calling ttm_bo_global_init() and -ttm_bo_global_release(), respectively. Also, like the previous +be provided, likely eventually calling ttm_bo_global_ref_init() and +ttm_bo_global_ref_release(), respectively. Also, like the previous object, ttm_global_item_ref() is used to create an initial reference count for the TTM, which will call your initialization function. diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index a44fc12ae1f9..3a6802846698 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -125,8 +125,8 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev) global_ref = &adev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index fe354ebf374d..d21fbd26785a 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -70,8 +70,8 @@ static int ast_ttm_global_init(struct ast_private *ast) global_ref = &ast->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c index a61c1ecb2bdc..2d36179c0e83 100644 --- a/drivers/gpu/drm/bochs/bochs_mm.c +++ b/drivers/gpu/drm/bochs/bochs_mm.c @@ -48,8 +48,8 @@ static int bochs_ttm_global_init(struct bochs_device *bochs) global_ref = &bochs->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c index f21953243790..2e2141f26c5b 100644 --- a/drivers/gpu/drm/cirrus/cirrus_ttm.c +++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c @@ -70,8 +70,8 @@ static int cirrus_ttm_global_init(struct cirrus_device *cirrus) global_ref = &cirrus->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c index 2e3e0bdb8932..0454aa43ffc6 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c @@ -59,8 +59,8 @@ static int hibmc_ttm_global_init(struct hibmc_drm_private *hibmc) hibmc->mem_global_ref.object; hibmc->bo_global_ref.ref.global_type = DRM_GLOBAL_TTM_BO; hibmc->bo_global_ref.ref.size = sizeof(struct ttm_bo_global); - hibmc->bo_global_ref.ref.init = &ttm_bo_global_init; - hibmc->bo_global_ref.ref.release = &ttm_bo_global_release; + hibmc->bo_global_ref.ref.init = &ttm_bo_global_ref_init; + hibmc->bo_global_ref.ref.release = &ttm_bo_global_ref_release; ret = drm_global_item_ref(&hibmc->bo_global_ref.ref); if (ret) { DRM_ERROR("failed setting up TTM BO subsystem: %d\n", ret); diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c index 05570f0de4d7..3444b539e7f4 100644 --- a/drivers/gpu/drm/mgag200/mgag200_ttm.c +++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c @@ -70,8 +70,8 @@ static int mgag200_ttm_global_init(struct mga_device *ast) global_ref = &ast->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index 8edb9f2a4269..a293383c8654 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -209,8 +209,8 @@ nouveau_ttm_global_init(struct nouveau_drm *drm) global_ref = &drm->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; ret = drm_global_item_ref(global_ref); if (unlikely(ret != 0)) { diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index 86a1fb32f6db..db2a0036e9c4 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -80,8 +80,8 @@ static int qxl_ttm_global_init(struct qxl_device *qdev) global_ref = &qdev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index cbb67e9ffb3a..dac4ec5a120b 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -97,8 +97,8 @@ static int radeon_ttm_global_init(struct radeon_device *rdev) global_ref = &rdev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 26b889f86670..9c2bb880491e 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1522,16 +1522,16 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj) kfree(glob); } -void ttm_bo_global_release(struct drm_global_reference *ref) +void ttm_bo_global_ref_release(struct drm_global_reference *ref) { struct ttm_bo_global *glob = ref->object; kobject_del(&glob->kobj); kobject_put(&glob->kobj); } -EXPORT_SYMBOL(ttm_bo_global_release); +EXPORT_SYMBOL(ttm_bo_global_ref_release); -int ttm_bo_global_init(struct drm_global_reference *ref) +int ttm_bo_global_ref_init(struct drm_global_reference *ref) { struct ttm_bo_global_ref *bo_ref = container_of(ref, struct ttm_bo_global_ref, ref); @@ -1564,7 +1564,7 @@ out_no_drp: kfree(glob); return ret; } -EXPORT_SYMBOL(ttm_bo_global_init); +EXPORT_SYMBOL(ttm_bo_global_ref_init); int ttm_bo_device_release(struct ttm_bo_device *bdev) diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c index e3152d45c5f1..526a5e48dc3b 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ttm.c +++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c @@ -84,8 +84,8 @@ static int virtio_gpu_ttm_global_init(struct virtio_gpu_device *vgdev) global_ref = &vgdev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c index 7b1e5a5cbd2c..f3ce43c41978 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c @@ -76,8 +76,8 @@ int vmw_ttm_global_init(struct vmw_private *dev_priv) global_ref = &dev_priv->bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; ret = drm_global_item_ref(global_ref); if (unlikely(ret != 0)) { diff --git a/drivers/staging/vboxvideo/vbox_ttm.c b/drivers/staging/vboxvideo/vbox_ttm.c index 5ecfa7629173..344975579ea5 100644 --- a/drivers/staging/vboxvideo/vbox_ttm.c +++ b/drivers/staging/vboxvideo/vbox_ttm.c @@ -68,8 +68,8 @@ static int vbox_ttm_global_init(struct vbox_private *vbox) global_ref = &vbox->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_init; - global_ref->release = &ttm_bo_global_release; + global_ref->init = &ttm_bo_global_ref_init; + global_ref->release = &ttm_bo_global_ref_release; ret = drm_global_item_ref(global_ref); if (ret) { diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index e4fee8e02559..c3c0751dec63 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -578,8 +578,8 @@ void ttm_bo_mem_put(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); void ttm_bo_mem_put_locked(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); -void ttm_bo_global_release(struct drm_global_reference *ref); -int ttm_bo_global_init(struct drm_global_reference *ref); +void ttm_bo_global_ref_release(struct drm_global_reference *ref); +int ttm_bo_global_ref_init(struct drm_global_reference *ref); int ttm_bo_device_release(struct ttm_bo_device *bdev); -- cgit v1.2.3 From 105f20706fb5df8b763e3d9a9bfbfa73386391c3 Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Tue, 16 Oct 2018 10:04:09 +0200 Subject: drm/ttm: Provide ttm_bo_global_{init/release}() for struct ttm_bo_global MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit So far, struct ttm_bo_global_ref was the only way of initializing a struct ttm_bo_global. Providing separate initializer and release functions for struct ttm_bo_global gives drivers the option of implementing their own init and release callbacks for drm_global_references of type DRM_GLOBAL_TTM_BO. The original functions for initializing and releasing via struct ttm_bo_global_ref are wrappers around the new interfaces. Signed-off-by: Thomas Zimmermann Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/ttm/ttm_bo.c | 16 +++++-------- include/drm/ttm/ttm_bo_driver.h | 53 ++++++++++++++++++++++++++++++++--------- 2 files changed, 48 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 9c2bb880491e..9edece6510d3 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1522,26 +1522,22 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj) kfree(glob); } -void ttm_bo_global_ref_release(struct drm_global_reference *ref) +void ttm_bo_global_release(struct ttm_bo_global *glob) { - struct ttm_bo_global *glob = ref->object; - kobject_del(&glob->kobj); kobject_put(&glob->kobj); } -EXPORT_SYMBOL(ttm_bo_global_ref_release); +EXPORT_SYMBOL(ttm_bo_global_release); -int ttm_bo_global_ref_init(struct drm_global_reference *ref) +int ttm_bo_global_init(struct ttm_bo_global *glob, + struct ttm_mem_global *mem_glob) { - struct ttm_bo_global_ref *bo_ref = - container_of(ref, struct ttm_bo_global_ref, ref); - struct ttm_bo_global *glob = ref->object; int ret; unsigned i; mutex_init(&glob->device_list_mutex); spin_lock_init(&glob->lru_lock); - glob->mem_glob = bo_ref->mem_glob; + glob->mem_glob = mem_glob; glob->mem_glob->bo_glob = glob; glob->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32); @@ -1564,7 +1560,7 @@ out_no_drp: kfree(glob); return ret; } -EXPORT_SYMBOL(ttm_bo_global_ref_init); +EXPORT_SYMBOL(ttm_bo_global_init); int ttm_bo_device_release(struct ttm_bo_device *bdev) diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index c3c0751dec63..c6ee07d10281 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -384,15 +384,6 @@ struct ttm_bo_driver { void *buf, int len, int write); }; -/** - * struct ttm_bo_global_ref - Argument to initialize a struct ttm_bo_global. - */ - -struct ttm_bo_global_ref { - struct drm_global_reference ref; - struct ttm_mem_global *mem_glob; -}; - /** * struct ttm_bo_global - Buffer object driver global data. * @@ -578,8 +569,9 @@ void ttm_bo_mem_put(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); void ttm_bo_mem_put_locked(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); -void ttm_bo_global_ref_release(struct drm_global_reference *ref); -int ttm_bo_global_ref_init(struct drm_global_reference *ref); +void ttm_bo_global_release(struct ttm_bo_global *glob); +int ttm_bo_global_init(struct ttm_bo_global *glob, + struct ttm_mem_global *mem_glob); int ttm_bo_device_release(struct ttm_bo_device *bdev); @@ -897,4 +889,43 @@ pgprot_t ttm_io_prot(uint32_t caching_flags, pgprot_t tmp); extern const struct ttm_mem_type_manager_func ttm_bo_manager_func; +/** + * struct ttm_bo_global_ref - Argument to initialize a struct ttm_bo_global. + */ + +struct ttm_bo_global_ref { + struct drm_global_reference ref; + struct ttm_mem_global *mem_glob; +}; + +/** + * ttm_bo_global_ref_init + * + * @ref: DRM global reference + * + * Helper function that initializes a struct ttm_bo_global. This function + * is used as init call-back function for DRM global references of type + * DRM_GLOBAL_TTM_BO_REF. + */ +static inline int ttm_bo_global_ref_init(struct drm_global_reference *ref) +{ + struct ttm_bo_global_ref *bo_ref = + container_of(ref, struct ttm_bo_global_ref, ref); + return ttm_bo_global_init(ref->object, bo_ref->mem_glob); +} + +/** + * ttm_bo_global_ref_release + * + * @ref: DRM global reference + * + * Helper function that releases a struct ttm_bo_global. This function + * is used as release call-back function for DRM global references of type + * DRM_GLOBAL_TTM_BO_REF. + */ +static inline void ttm_bo_global_ref_release(struct drm_global_reference *ref) +{ + ttm_bo_global_release(ref->object); +} + #endif -- cgit v1.2.3 From 8fe159b0143d817222c8799181deb799472b9339 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 12 Oct 2018 16:47:13 +0200 Subject: drm/sched: add drm_sched_fault MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add a helper to immediately start timeout handling in case of a hardware fault. Signed-off-by: Christian König Reviewed-by: Andrey Grodzovsky Signed-off-by: Alex Deucher --- drivers/gpu/drm/scheduler/sched_main.c | 13 +++++++++++++ include/drm/gpu_scheduler.h | 1 + 2 files changed, 14 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c index 73449c653b6e..63b997d9c562 100644 --- a/drivers/gpu/drm/scheduler/sched_main.c +++ b/drivers/gpu/drm/scheduler/sched_main.c @@ -196,6 +196,19 @@ static void drm_sched_start_timeout(struct drm_gpu_scheduler *sched) schedule_delayed_work(&sched->work_tdr, sched->timeout); } +/** + * drm_sched_fault - immediately start timeout handler + * + * @sched: scheduler where the timeout handling should be started. + * + * Start timeout handling immediately when the driver detects a hardware fault. + */ +void drm_sched_fault(struct drm_gpu_scheduler *sched) +{ + mod_delayed_work(system_wq, &sched->work_tdr, 0); +} +EXPORT_SYMBOL(drm_sched_fault); + /* job_finish is called after hw fence signaled */ static void drm_sched_job_finish(struct work_struct *work) diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index d87b268f1781..0684dcd99c0f 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -299,6 +299,7 @@ void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, void drm_sched_job_recovery(struct drm_gpu_scheduler *sched); bool drm_sched_dependency_optimized(struct dma_fence* fence, struct drm_sched_entity *entity); +void drm_sched_fault(struct drm_gpu_scheduler *sched); void drm_sched_job_kickout(struct drm_sched_job *s_job); void drm_sched_rq_add_entity(struct drm_sched_rq *rq, -- cgit v1.2.3 From ce331f8f7c04186b7b130550250d9c4e752b4f8f Mon Sep 17 00:00:00 2001 From: Nicholas Kazlauskas Date: Tue, 23 Oct 2018 10:04:54 -0400 Subject: drm/amdgpu: Add DCC flags for GFX9 amdgpu_bo MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit [Why] Hardware support for Delta Color Compression (DCC) decompression is available in DC for GFX9 but there's no way for userspace to enable the feature. Enabling the feature can provide improved GFX performance and power savings in many situations. [How] Extend the GFX9 tiling flags to include DCC parameters. These are logically grouped together with tiling flags even if they are technically distinct. This trivially maintains backwards compatibility with existing users of amdgpu_gem_metadata. No new IOCTls or data structures are needed to support DCC. This patch helps expose DCC attributes to both libdrm and amdgpu_dm. Signed-off-by: Nicholas Kazlauskas Reviewed-by: Marek Olšák Signed-off-by: Alex Deucher --- include/uapi/drm/amdgpu_drm.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index 370e9a5536ef..be84e43c1e19 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -326,6 +326,12 @@ struct drm_amdgpu_gem_userptr { /* GFX9 and later: */ #define AMDGPU_TILING_SWIZZLE_MODE_SHIFT 0 #define AMDGPU_TILING_SWIZZLE_MODE_MASK 0x1f +#define AMDGPU_TILING_DCC_OFFSET_256B_SHIFT 5 +#define AMDGPU_TILING_DCC_OFFSET_256B_MASK 0xFFFFFF +#define AMDGPU_TILING_DCC_PITCH_MAX_SHIFT 29 +#define AMDGPU_TILING_DCC_PITCH_MAX_MASK 0x3FFF +#define AMDGPU_TILING_DCC_INDEPENDENT_64B_SHIFT 43 +#define AMDGPU_TILING_DCC_INDEPENDENT_64B_MASK 0x1 /* Set/Get helpers for tiling flags. */ #define AMDGPU_TILING_SET(field, value) \ -- cgit v1.2.3 From 27eb1fa9130a98edd2b321d4dbce5c8b244ee7af Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 19 Oct 2018 13:49:05 +0200 Subject: drm/ttm: use a static ttm_mem_global instance MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As the name says we only need one global instance of ttm_mem_global. Drop all the driver initialization and just use a single exported instance which is initialized during BO global initialization. Signed-off-by: Christian König Reviewed-by: Junwei Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 44 ------------------------- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 1 - drivers/gpu/drm/ast/ast_drv.h | 1 - drivers/gpu/drm/ast/ast_ttm.c | 32 ++---------------- drivers/gpu/drm/bochs/bochs.h | 1 - drivers/gpu/drm/bochs/bochs_mm.c | 30 ++--------------- drivers/gpu/drm/cirrus/cirrus_drv.h | 1 - drivers/gpu/drm/cirrus/cirrus_ttm.c | 32 ++---------------- drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h | 1 - drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c | 31 +++-------------- drivers/gpu/drm/mgag200/mgag200_drv.h | 1 - drivers/gpu/drm/mgag200/mgag200_ttm.c | 32 ++---------------- drivers/gpu/drm/nouveau/nouveau_drv.h | 1 - drivers/gpu/drm/nouveau/nouveau_ttm.c | 34 ++----------------- drivers/gpu/drm/qxl/qxl_drv.h | 1 - drivers/gpu/drm/qxl/qxl_ttm.c | 28 ---------------- drivers/gpu/drm/radeon/radeon.h | 1 - drivers/gpu/drm/radeon/radeon_ttm.c | 26 --------------- drivers/gpu/drm/ttm/ttm_bo.c | 10 ++++-- drivers/gpu/drm/ttm/ttm_memory.c | 5 +-- drivers/gpu/drm/virtio/virtgpu_drv.h | 1 - drivers/gpu/drm/virtio/virtgpu_ttm.c | 27 --------------- drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 4 +-- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 3 +- drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c | 27 --------------- drivers/staging/vboxvideo/vbox_drv.h | 1 - drivers/staging/vboxvideo/vbox_ttm.c | 24 -------------- include/drm/ttm/ttm_bo_driver.h | 8 ++--- include/drm/ttm/ttm_memory.h | 4 +-- 29 files changed, 32 insertions(+), 380 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 3a6802846698..fda252022b15 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -65,33 +65,6 @@ static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev); * Global memory. */ -/** - * amdgpu_ttm_mem_global_init - Initialize and acquire reference to - * memory object - * - * @ref: Object for initialization. - * - * This is called by drm_global_item_ref() when an object is being - * initialized. - */ -static int amdgpu_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -/** - * amdgpu_ttm_mem_global_release - Drop reference to a memory object - * - * @ref: Object being removed - * - * This is called by drm_global_item_unref() when an object is being - * released. - */ -static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - /** * amdgpu_ttm_global_init - Initialize global TTM memory reference structures. * @@ -108,20 +81,6 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev) /* ensure reference is false in case init fails */ adev->mman.mem_global_referenced = false; - global_ref = &adev->mman.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &amdgpu_ttm_mem_global_init; - global_ref->release = &amdgpu_ttm_mem_global_release; - r = drm_global_item_ref(global_ref); - if (r) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - goto error_mem; - } - - adev->mman.bo_global_ref.mem_glob = - adev->mman.mem_global_ref.object; global_ref = &adev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -140,8 +99,6 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev) return 0; error_bo: - drm_global_item_unref(&adev->mman.mem_global_ref); -error_mem: return r; } @@ -150,7 +107,6 @@ static void amdgpu_ttm_global_fini(struct amdgpu_device *adev) if (adev->mman.mem_global_referenced) { mutex_destroy(&adev->mman.gtt_window_lock); drm_global_item_unref(&adev->mman.bo_global_ref.ref); - drm_global_item_unref(&adev->mman.mem_global_ref); adev->mman.mem_global_referenced = false; } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index fe8f276e9811..e114f209b701 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -40,7 +40,6 @@ struct amdgpu_mman { struct ttm_bo_global_ref bo_global_ref; - struct drm_global_reference mem_global_ref; struct ttm_bo_device bdev; bool mem_global_referenced; bool initialized; diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h index e6c4cd3dc50e..6ae11a477643 100644 --- a/drivers/gpu/drm/ast/ast_drv.h +++ b/drivers/gpu/drm/ast/ast_drv.h @@ -104,7 +104,6 @@ struct ast_private { int fb_mtrr; struct { - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index d21fbd26785a..8a59d6fc1160 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -36,37 +36,11 @@ ast_bdev(struct ttm_bo_device *bd) return container_of(bd, struct ast_private, ttm.bdev); } -static int -ast_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void -ast_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int ast_ttm_global_init(struct ast_private *ast) { struct drm_global_reference *global_ref; int r; - global_ref = &ast->ttm.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &ast_ttm_mem_global_init; - global_ref->release = &ast_ttm_mem_global_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - ast->ttm.bo_global_ref.mem_glob = - ast->ttm.mem_global_ref.object; global_ref = &ast->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -75,7 +49,6 @@ static int ast_ttm_global_init(struct ast_private *ast) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&ast->ttm.mem_global_ref); return r; } return 0; @@ -84,12 +57,11 @@ static int ast_ttm_global_init(struct ast_private *ast) static void ast_ttm_global_release(struct ast_private *ast) { - if (ast->ttm.mem_global_ref.release == NULL) + if (ast->ttm.bo_global_ref.ref.release == NULL) return; drm_global_item_unref(&ast->ttm.bo_global_ref.ref); - drm_global_item_unref(&ast->ttm.mem_global_ref); - ast->ttm.mem_global_ref.release = NULL; + ast->ttm.bo_global_ref.ref.release = NULL; } diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h index e7a69077e45a..a035257d1b56 100644 --- a/drivers/gpu/drm/bochs/bochs.h +++ b/drivers/gpu/drm/bochs/bochs.h @@ -76,7 +76,6 @@ struct bochs_device { /* ttm */ struct { - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; bool initialized; diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c index 2d36179c0e83..c697d456656f 100644 --- a/drivers/gpu/drm/bochs/bochs_mm.c +++ b/drivers/gpu/drm/bochs/bochs_mm.c @@ -16,35 +16,11 @@ static inline struct bochs_device *bochs_bdev(struct ttm_bo_device *bd) return container_of(bd, struct bochs_device, ttm.bdev); } -static int bochs_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void bochs_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int bochs_ttm_global_init(struct bochs_device *bochs) { struct drm_global_reference *global_ref; int r; - global_ref = &bochs->ttm.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &bochs_ttm_mem_global_init; - global_ref->release = &bochs_ttm_mem_global_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - bochs->ttm.bo_global_ref.mem_glob = - bochs->ttm.mem_global_ref.object; global_ref = &bochs->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -53,7 +29,6 @@ static int bochs_ttm_global_init(struct bochs_device *bochs) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&bochs->ttm.mem_global_ref); return r; } @@ -62,12 +37,11 @@ static int bochs_ttm_global_init(struct bochs_device *bochs) static void bochs_ttm_global_release(struct bochs_device *bochs) { - if (bochs->ttm.mem_global_ref.release == NULL) + if (bochs->ttm.bo_global_ref.ref.release == NULL) return; drm_global_item_unref(&bochs->ttm.bo_global_ref.ref); - drm_global_item_unref(&bochs->ttm.mem_global_ref); - bochs->ttm.mem_global_ref.release = NULL; + bochs->ttm.bo_global_ref.ref.release = NULL; } diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h index a29f87e98d9d..01852fbda9da 100644 --- a/drivers/gpu/drm/cirrus/cirrus_drv.h +++ b/drivers/gpu/drm/cirrus/cirrus_drv.h @@ -136,7 +136,6 @@ struct cirrus_device { int fb_mtrr; struct { - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c index 2e2141f26c5b..7801c56c3c9b 100644 --- a/drivers/gpu/drm/cirrus/cirrus_ttm.c +++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c @@ -36,37 +36,11 @@ cirrus_bdev(struct ttm_bo_device *bd) return container_of(bd, struct cirrus_device, ttm.bdev); } -static int -cirrus_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void -cirrus_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int cirrus_ttm_global_init(struct cirrus_device *cirrus) { struct drm_global_reference *global_ref; int r; - global_ref = &cirrus->ttm.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &cirrus_ttm_mem_global_init; - global_ref->release = &cirrus_ttm_mem_global_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - cirrus->ttm.bo_global_ref.mem_glob = - cirrus->ttm.mem_global_ref.object; global_ref = &cirrus->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -75,7 +49,6 @@ static int cirrus_ttm_global_init(struct cirrus_device *cirrus) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&cirrus->ttm.mem_global_ref); return r; } return 0; @@ -84,12 +57,11 @@ static int cirrus_ttm_global_init(struct cirrus_device *cirrus) static void cirrus_ttm_global_release(struct cirrus_device *cirrus) { - if (cirrus->ttm.mem_global_ref.release == NULL) + if (cirrus->ttm.bo_global_ref.ref.release == NULL) return; drm_global_item_unref(&cirrus->ttm.bo_global_ref.ref); - drm_global_item_unref(&cirrus->ttm.mem_global_ref); - cirrus->ttm.mem_global_ref.release = NULL; + cirrus->ttm.bo_global_ref.ref.release = NULL; } diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h index 45c25a488f42..60479502e277 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h @@ -49,7 +49,6 @@ struct hibmc_drm_private { bool mode_config_initialized; /* ttm */ - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; bool initialized; diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c index 0454aa43ffc6..14071c849121 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c @@ -29,34 +29,10 @@ hibmc_bdev(struct ttm_bo_device *bd) return container_of(bd, struct hibmc_drm_private, bdev); } -static int -hibmc_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void -hibmc_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int hibmc_ttm_global_init(struct hibmc_drm_private *hibmc) { int ret; - hibmc->mem_global_ref.global_type = DRM_GLOBAL_TTM_MEM; - hibmc->mem_global_ref.size = sizeof(struct ttm_mem_global); - hibmc->mem_global_ref.init = &hibmc_ttm_mem_global_init; - hibmc->mem_global_ref.release = &hibmc_ttm_mem_global_release; - ret = drm_global_item_ref(&hibmc->mem_global_ref); - if (ret) { - DRM_ERROR("could not get ref on ttm global: %d\n", ret); - return ret; - } - - hibmc->bo_global_ref.mem_glob = - hibmc->mem_global_ref.object; hibmc->bo_global_ref.ref.global_type = DRM_GLOBAL_TTM_BO; hibmc->bo_global_ref.ref.size = sizeof(struct ttm_bo_global); hibmc->bo_global_ref.ref.init = &ttm_bo_global_ref_init; @@ -64,7 +40,6 @@ static int hibmc_ttm_global_init(struct hibmc_drm_private *hibmc) ret = drm_global_item_ref(&hibmc->bo_global_ref.ref); if (ret) { DRM_ERROR("failed setting up TTM BO subsystem: %d\n", ret); - drm_global_item_unref(&hibmc->mem_global_ref); return ret; } return 0; @@ -73,9 +48,11 @@ static int hibmc_ttm_global_init(struct hibmc_drm_private *hibmc) static void hibmc_ttm_global_release(struct hibmc_drm_private *hibmc) { + if (hibmc->bo_global_ref.ref.release == NULL) + return; + drm_global_item_unref(&hibmc->bo_global_ref.ref); - drm_global_item_unref(&hibmc->mem_global_ref); - hibmc->mem_global_ref.release = NULL; + hibmc->bo_global_ref.ref.release = NULL; } static void hibmc_bo_ttm_destroy(struct ttm_buffer_object *tbo) diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.h b/drivers/gpu/drm/mgag200/mgag200_drv.h index 04f1dfba12e5..e5348955a3cc 100644 --- a/drivers/gpu/drm/mgag200/mgag200_drv.h +++ b/drivers/gpu/drm/mgag200/mgag200_drv.h @@ -212,7 +212,6 @@ struct mga_device { int fb_mtrr; struct { - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c index 3444b539e7f4..11bdc8121a7d 100644 --- a/drivers/gpu/drm/mgag200/mgag200_ttm.c +++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c @@ -36,37 +36,11 @@ mgag200_bdev(struct ttm_bo_device *bd) return container_of(bd, struct mga_device, ttm.bdev); } -static int -mgag200_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void -mgag200_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int mgag200_ttm_global_init(struct mga_device *ast) { struct drm_global_reference *global_ref; int r; - global_ref = &ast->ttm.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &mgag200_ttm_mem_global_init; - global_ref->release = &mgag200_ttm_mem_global_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - ast->ttm.bo_global_ref.mem_glob = - ast->ttm.mem_global_ref.object; global_ref = &ast->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -75,7 +49,6 @@ static int mgag200_ttm_global_init(struct mga_device *ast) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&ast->ttm.mem_global_ref); return r; } return 0; @@ -84,12 +57,11 @@ static int mgag200_ttm_global_init(struct mga_device *ast) static void mgag200_ttm_global_release(struct mga_device *ast) { - if (ast->ttm.mem_global_ref.release == NULL) + if (ast->ttm.bo_global_ref.ref.release == NULL) return; drm_global_item_unref(&ast->ttm.bo_global_ref.ref); - drm_global_item_unref(&ast->ttm.mem_global_ref); - ast->ttm.mem_global_ref.release = NULL; + ast->ttm.bo_global_ref.ref.release = NULL; } diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index 0b2191fa96f7..0f3bb1a11fc9 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -146,7 +146,6 @@ struct nouveau_drm { /* TTM interface support */ struct { - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; atomic_t validate_sequence; diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index a293383c8654..69448b02649e 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -174,38 +174,12 @@ nouveau_ttm_mmap(struct file *filp, struct vm_area_struct *vma) return ttm_bo_mmap(filp, vma, &drm->ttm.bdev); } -static int -nouveau_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void -nouveau_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - int nouveau_ttm_global_init(struct nouveau_drm *drm) { struct drm_global_reference *global_ref; int ret; - global_ref = &drm->ttm.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &nouveau_ttm_mem_global_init; - global_ref->release = &nouveau_ttm_mem_global_release; - - ret = drm_global_item_ref(global_ref); - if (unlikely(ret != 0)) { - DRM_ERROR("Failed setting up TTM memory accounting\n"); - drm->ttm.mem_global_ref.release = NULL; - return ret; - } - - drm->ttm.bo_global_ref.mem_glob = global_ref->object; global_ref = &drm->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -215,8 +189,7 @@ nouveau_ttm_global_init(struct nouveau_drm *drm) ret = drm_global_item_ref(global_ref); if (unlikely(ret != 0)) { DRM_ERROR("Failed setting up TTM BO subsystem\n"); - drm_global_item_unref(&drm->ttm.mem_global_ref); - drm->ttm.mem_global_ref.release = NULL; + drm->ttm.bo_global_ref.ref.release = NULL; return ret; } @@ -226,12 +199,11 @@ nouveau_ttm_global_init(struct nouveau_drm *drm) void nouveau_ttm_global_release(struct nouveau_drm *drm) { - if (drm->ttm.mem_global_ref.release == NULL) + if (drm->ttm.bo_global_ref.ref.release == NULL) return; drm_global_item_unref(&drm->ttm.bo_global_ref.ref); - drm_global_item_unref(&drm->ttm.mem_global_ref); - drm->ttm.mem_global_ref.release = NULL; + drm->ttm.bo_global_ref.ref.release = NULL; } static int diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h index 8ff70a7281a7..52912e54e990 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.h +++ b/drivers/gpu/drm/qxl/qxl_drv.h @@ -128,7 +128,6 @@ struct qxl_output { struct qxl_mman { struct ttm_bo_global_ref bo_global_ref; - struct drm_global_reference mem_global_ref; bool mem_global_referenced; struct ttm_bo_device bdev; }; diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index db2a0036e9c4..bb8cc9b16780 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -46,37 +46,11 @@ static struct qxl_device *qxl_get_qdev(struct ttm_bo_device *bdev) return qdev; } -static int qxl_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void qxl_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int qxl_ttm_global_init(struct qxl_device *qdev) { struct drm_global_reference *global_ref; int r; - qdev->mman.mem_global_referenced = false; - global_ref = &qdev->mman.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &qxl_ttm_mem_global_init; - global_ref->release = &qxl_ttm_mem_global_release; - - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - qdev->mman.bo_global_ref.mem_glob = - qdev->mman.mem_global_ref.object; global_ref = &qdev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -85,7 +59,6 @@ static int qxl_ttm_global_init(struct qxl_device *qdev) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&qdev->mman.mem_global_ref); return r; } @@ -97,7 +70,6 @@ static void qxl_ttm_global_fini(struct qxl_device *qdev) { if (qdev->mman.mem_global_referenced) { drm_global_item_unref(&qdev->mman.bo_global_ref.ref); - drm_global_item_unref(&qdev->mman.mem_global_ref); qdev->mman.mem_global_referenced = false; } } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 1a6f6edb3515..06fb952b6290 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -449,7 +449,6 @@ struct radeon_surface_reg { */ struct radeon_mman { struct ttm_bo_global_ref bo_global_ref; - struct drm_global_reference mem_global_ref; struct ttm_bo_device bdev; bool mem_global_referenced; bool initialized; diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index dac4ec5a120b..2104429ddbac 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -64,36 +64,12 @@ static struct radeon_device *radeon_get_rdev(struct ttm_bo_device *bdev) /* * Global memory. */ -static int radeon_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void radeon_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int radeon_ttm_global_init(struct radeon_device *rdev) { struct drm_global_reference *global_ref; int r; rdev->mman.mem_global_referenced = false; - global_ref = &rdev->mman.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &radeon_ttm_mem_global_init; - global_ref->release = &radeon_ttm_mem_global_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - rdev->mman.bo_global_ref.mem_glob = - rdev->mman.mem_global_ref.object; global_ref = &rdev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -102,7 +78,6 @@ static int radeon_ttm_global_init(struct radeon_device *rdev) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&rdev->mman.mem_global_ref); return r; } @@ -114,7 +89,6 @@ static void radeon_ttm_global_fini(struct radeon_device *rdev) { if (rdev->mman.mem_global_referenced) { drm_global_item_unref(&rdev->mman.bo_global_ref.ref); - drm_global_item_unref(&rdev->mman.mem_global_ref); rdev->mman.mem_global_referenced = false; } } diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 9edece6510d3..3006050b1720 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1526,18 +1526,22 @@ void ttm_bo_global_release(struct ttm_bo_global *glob) { kobject_del(&glob->kobj); kobject_put(&glob->kobj); + ttm_mem_global_release(&ttm_mem_glob); } EXPORT_SYMBOL(ttm_bo_global_release); -int ttm_bo_global_init(struct ttm_bo_global *glob, - struct ttm_mem_global *mem_glob) +int ttm_bo_global_init(struct ttm_bo_global *glob) { int ret; unsigned i; + ret = ttm_mem_global_init(&ttm_mem_glob); + if (ret) + return ret; + mutex_init(&glob->device_list_mutex); spin_lock_init(&glob->lru_lock); - glob->mem_glob = mem_glob; + glob->mem_glob = &ttm_mem_glob; glob->mem_glob->bo_glob = glob; glob->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32); diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c index 450387c92b63..7704e17c402f 100644 --- a/drivers/gpu/drm/ttm/ttm_memory.c +++ b/drivers/gpu/drm/ttm/ttm_memory.c @@ -41,6 +41,9 @@ #define TTM_MEMORY_ALLOC_RETRIES 4 +struct ttm_mem_global ttm_mem_glob; +EXPORT_SYMBOL(ttm_mem_glob); + struct ttm_mem_zone { struct kobject kobj; struct ttm_mem_global *glob; @@ -464,7 +467,6 @@ out_no_zone: ttm_mem_global_release(glob); return ret; } -EXPORT_SYMBOL(ttm_mem_global_init); void ttm_mem_global_release(struct ttm_mem_global *glob) { @@ -486,7 +488,6 @@ void ttm_mem_global_release(struct ttm_mem_global *glob) kobject_del(&glob->kobj); kobject_put(&glob->kobj); } -EXPORT_SYMBOL(ttm_mem_global_release); static void ttm_check_swapping(struct ttm_mem_global *glob) { diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h index d29f0c7c768c..d9756b59e329 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.h +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h @@ -143,7 +143,6 @@ struct virtio_gpu_fbdev { struct virtio_gpu_mman { struct ttm_bo_global_ref bo_global_ref; - struct drm_global_reference mem_global_ref; bool mem_global_referenced; struct ttm_bo_device bdev; }; diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c index 526a5e48dc3b..8510109e09da 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ttm.c +++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c @@ -50,37 +50,12 @@ virtio_gpu_device *virtio_gpu_get_vgdev(struct ttm_bo_device *bdev) return vgdev; } -static int virtio_gpu_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void virtio_gpu_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - static int virtio_gpu_ttm_global_init(struct virtio_gpu_device *vgdev) { struct drm_global_reference *global_ref; int r; vgdev->mman.mem_global_referenced = false; - global_ref = &vgdev->mman.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &virtio_gpu_ttm_mem_global_init; - global_ref->release = &virtio_gpu_ttm_mem_global_release; - - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM memory accounting " - "subsystem.\n"); - return r; - } - - vgdev->mman.bo_global_ref.mem_glob = - vgdev->mman.mem_global_ref.object; global_ref = &vgdev->mman.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -89,7 +64,6 @@ static int virtio_gpu_ttm_global_init(struct virtio_gpu_device *vgdev) r = drm_global_item_ref(global_ref); if (r != 0) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&vgdev->mman.mem_global_ref); return r; } @@ -101,7 +75,6 @@ static void virtio_gpu_ttm_global_fini(struct virtio_gpu_device *vgdev) { if (vgdev->mman.mem_global_referenced) { drm_global_item_unref(&vgdev->mman.bo_global_ref.ref); - drm_global_item_unref(&vgdev->mman.mem_global_ref); vgdev->mman.mem_global_referenced = false; } } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c index 61a84b958d67..67494148accd 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c @@ -828,8 +828,8 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset) goto out_err4; } - dev_priv->tdev = ttm_object_device_init - (dev_priv->mem_global_ref.object, 12, &vmw_prime_dmabuf_ops); + dev_priv->tdev = ttm_object_device_init(&ttm_mem_glob, 12, + &vmw_prime_dmabuf_ops); if (unlikely(dev_priv->tdev == NULL)) { DRM_ERROR("Unable to initialize TTM object management.\n"); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 59f614225bcd..252f202ae897 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -418,7 +418,6 @@ enum { struct vmw_private { struct ttm_bo_device bdev; struct ttm_bo_global_ref bo_global_ref; - struct drm_global_reference mem_global_ref; struct vmw_fifo_state fifo; @@ -1363,7 +1362,7 @@ vmw_bo_reference(struct vmw_buffer_object *buf) static inline struct ttm_mem_global *vmw_mem_glob(struct vmw_private *dev_priv) { - return (struct ttm_mem_global *) dev_priv->mem_global_ref.object; + return &ttm_mem_glob; } static inline void vmw_fifo_resource_inc(struct vmw_private *dev_priv) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c index f3ce43c41978..0ac473cd5136 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c @@ -43,36 +43,11 @@ int vmw_mmap(struct file *filp, struct vm_area_struct *vma) return ttm_bo_mmap(filp, vma, &dev_priv->bdev); } -static int vmw_ttm_mem_global_init(struct drm_global_reference *ref) -{ - DRM_INFO("global init.\n"); - return ttm_mem_global_init(ref->object); -} - -static void vmw_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - int vmw_ttm_global_init(struct vmw_private *dev_priv) { struct drm_global_reference *global_ref; int ret; - global_ref = &dev_priv->mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &vmw_ttm_mem_global_init; - global_ref->release = &vmw_ttm_mem_global_release; - - ret = drm_global_item_ref(global_ref); - if (unlikely(ret != 0)) { - DRM_ERROR("Failed setting up TTM memory accounting.\n"); - return ret; - } - - dev_priv->bo_global_ref.mem_glob = - dev_priv->mem_global_ref.object; global_ref = &dev_priv->bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -87,12 +62,10 @@ int vmw_ttm_global_init(struct vmw_private *dev_priv) return 0; out_no_bo: - drm_global_item_unref(&dev_priv->mem_global_ref); return ret; } void vmw_ttm_global_release(struct vmw_private *dev_priv) { drm_global_item_unref(&dev_priv->bo_global_ref.ref); - drm_global_item_unref(&dev_priv->mem_global_ref); } diff --git a/drivers/staging/vboxvideo/vbox_drv.h b/drivers/staging/vboxvideo/vbox_drv.h index 73395a7536c5..dc257a892ecb 100644 --- a/drivers/staging/vboxvideo/vbox_drv.h +++ b/drivers/staging/vboxvideo/vbox_drv.h @@ -99,7 +99,6 @@ struct vbox_private { int fb_mtrr; struct { - struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/staging/vboxvideo/vbox_ttm.c b/drivers/staging/vboxvideo/vbox_ttm.c index 344975579ea5..73b895328059 100644 --- a/drivers/staging/vboxvideo/vbox_ttm.c +++ b/drivers/staging/vboxvideo/vbox_ttm.c @@ -35,16 +35,6 @@ static inline struct vbox_private *vbox_bdev(struct ttm_bo_device *bd) return container_of(bd, struct vbox_private, ttm.bdev); } -static int vbox_ttm_mem_global_init(struct drm_global_reference *ref) -{ - return ttm_mem_global_init(ref->object); -} - -static void vbox_ttm_mem_global_release(struct drm_global_reference *ref) -{ - ttm_mem_global_release(ref->object); -} - /** * Adds the vbox memory manager object/structures to the global memory manager. */ @@ -53,18 +43,6 @@ static int vbox_ttm_global_init(struct vbox_private *vbox) struct drm_global_reference *global_ref; int ret; - global_ref = &vbox->ttm.mem_global_ref; - global_ref->global_type = DRM_GLOBAL_TTM_MEM; - global_ref->size = sizeof(struct ttm_mem_global); - global_ref->init = &vbox_ttm_mem_global_init; - global_ref->release = &vbox_ttm_mem_global_release; - ret = drm_global_item_ref(global_ref); - if (ret) { - DRM_ERROR("Failed setting up TTM memory subsystem.\n"); - return ret; - } - - vbox->ttm.bo_global_ref.mem_glob = vbox->ttm.mem_global_ref.object; global_ref = &vbox->ttm.bo_global_ref.ref; global_ref->global_type = DRM_GLOBAL_TTM_BO; global_ref->size = sizeof(struct ttm_bo_global); @@ -74,7 +52,6 @@ static int vbox_ttm_global_init(struct vbox_private *vbox) ret = drm_global_item_ref(global_ref); if (ret) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&vbox->ttm.mem_global_ref); return ret; } @@ -87,7 +64,6 @@ static int vbox_ttm_global_init(struct vbox_private *vbox) static void vbox_ttm_global_release(struct vbox_private *vbox) { drm_global_item_unref(&vbox->ttm.bo_global_ref.ref); - drm_global_item_unref(&vbox->ttm.mem_global_ref); } static void vbox_bo_ttm_destroy(struct ttm_buffer_object *tbo) diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index c6ee07d10281..4ae6fc33f761 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -570,8 +570,7 @@ void ttm_bo_mem_put_locked(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); void ttm_bo_global_release(struct ttm_bo_global *glob); -int ttm_bo_global_init(struct ttm_bo_global *glob, - struct ttm_mem_global *mem_glob); +int ttm_bo_global_init(struct ttm_bo_global *glob); int ttm_bo_device_release(struct ttm_bo_device *bdev); @@ -895,7 +894,6 @@ extern const struct ttm_mem_type_manager_func ttm_bo_manager_func; struct ttm_bo_global_ref { struct drm_global_reference ref; - struct ttm_mem_global *mem_glob; }; /** @@ -909,9 +907,7 @@ struct ttm_bo_global_ref { */ static inline int ttm_bo_global_ref_init(struct drm_global_reference *ref) { - struct ttm_bo_global_ref *bo_ref = - container_of(ref, struct ttm_bo_global_ref, ref); - return ttm_bo_global_init(ref->object, bo_ref->mem_glob); + return ttm_bo_global_init(ref->object); } /** diff --git a/include/drm/ttm/ttm_memory.h b/include/drm/ttm/ttm_memory.h index 737b5fed8003..3ff48a0a2d7b 100644 --- a/include/drm/ttm/ttm_memory.h +++ b/include/drm/ttm/ttm_memory.h @@ -63,7 +63,7 @@ #define TTM_MEM_MAX_ZONES 2 struct ttm_mem_zone; -struct ttm_mem_global { +extern struct ttm_mem_global { struct kobject kobj; struct ttm_bo_global *bo_glob; struct workqueue_struct *swap_queue; @@ -78,7 +78,7 @@ struct ttm_mem_global { #else struct ttm_mem_zone *zone_dma32; #endif -}; +} ttm_mem_glob; extern int ttm_mem_global_init(struct ttm_mem_global *glob); extern void ttm_mem_global_release(struct ttm_mem_global *glob); -- cgit v1.2.3 From 56b3d20413587fab6a790cfc8bc075ca94bc8ed9 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 19 Oct 2018 14:09:24 +0200 Subject: drm/ttm: make the device list mutex static MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This way it can protect the whole BO global state. Signed-off-by: Christian König Reviewed-by: Junwei Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/ttm/ttm_bo.c | 14 +++++++++----- include/drm/ttm/ttm_bo_driver.h | 1 - 2 files changed, 9 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 3006050b1720..4ec368b2555a 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -45,6 +45,11 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj); +/** + * ttm_global_mutex - protecting the global BO state + */ +DEFINE_MUTEX(ttm_global_mutex); + static struct attribute ttm_bo_count = { .name = "bo_count", .mode = S_IRUGO @@ -1539,7 +1544,6 @@ int ttm_bo_global_init(struct ttm_bo_global *glob) if (ret) return ret; - mutex_init(&glob->device_list_mutex); spin_lock_init(&glob->lru_lock); glob->mem_glob = &ttm_mem_glob; glob->mem_glob->bo_glob = glob; @@ -1587,9 +1591,9 @@ int ttm_bo_device_release(struct ttm_bo_device *bdev) } } - mutex_lock(&glob->device_list_mutex); + mutex_lock(&ttm_global_mutex); list_del(&bdev->device_list); - mutex_unlock(&glob->device_list_mutex); + mutex_unlock(&ttm_global_mutex); cancel_delayed_work_sync(&bdev->wq); @@ -1636,9 +1640,9 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev, bdev->dev_mapping = mapping; bdev->glob = glob; bdev->need_dma32 = need_dma32; - mutex_lock(&glob->device_list_mutex); + mutex_lock(&ttm_global_mutex); list_add_tail(&bdev->device_list, &glob->device_list); - mutex_unlock(&glob->device_list_mutex); + mutex_unlock(&ttm_global_mutex); return 0; out_no_sys: diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 4ae6fc33f761..9cec8835b88f 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -407,7 +407,6 @@ struct ttm_bo_global { struct kobject kobj; struct ttm_mem_global *mem_glob; struct page *dummy_read_page; - struct mutex device_list_mutex; spinlock_t lru_lock; /** -- cgit v1.2.3 From 62b53b37e4b1500d4eb4624a44ad861cf8d3cd18 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 19 Oct 2018 15:06:06 +0200 Subject: drm/ttm: use a static ttm_bo_global instance MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As the name says we only need one global instance of ttm_bo_global. Just use a single exported instance which is save to initialize multiple times. Signed-off-by: Christian König Reviewed-by: Junwei Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/ttm/ttm_bo.c | 31 +++++++++++++++++++++++-------- include/drm/ttm/ttm_bo_driver.h | 15 ++++++++------- 2 files changed, 31 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 4ec368b2555a..d89183f95570 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -49,6 +49,9 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj); * ttm_global_mutex - protecting the global BO state */ DEFINE_MUTEX(ttm_global_mutex); +struct ttm_bo_global ttm_bo_glob = { + .use_count = 0 +}; static struct attribute ttm_bo_count = { .name = "bo_count", @@ -1527,22 +1530,35 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj) kfree(glob); } -void ttm_bo_global_release(struct ttm_bo_global *glob) +void ttm_bo_global_release(void) { + struct ttm_bo_global *glob = &ttm_bo_glob; + + mutex_lock(&ttm_global_mutex); + if (--glob->use_count > 0) + goto out; + kobject_del(&glob->kobj); kobject_put(&glob->kobj); ttm_mem_global_release(&ttm_mem_glob); +out: + mutex_unlock(&ttm_global_mutex); } EXPORT_SYMBOL(ttm_bo_global_release); -int ttm_bo_global_init(struct ttm_bo_global *glob) +int ttm_bo_global_init(void) { - int ret; + struct ttm_bo_global *glob = &ttm_bo_glob; + int ret = 0; unsigned i; + mutex_lock(&ttm_global_mutex); + if (++glob->use_count > 1) + goto out; + ret = ttm_mem_global_init(&ttm_mem_glob); if (ret) - return ret; + goto out; spin_lock_init(&glob->lru_lock); glob->mem_glob = &ttm_mem_glob; @@ -1551,7 +1567,7 @@ int ttm_bo_global_init(struct ttm_bo_global *glob) if (unlikely(glob->dummy_read_page == NULL)) { ret = -ENOMEM; - goto out_no_drp; + goto out; } for (i = 0; i < TTM_MAX_BO_PRIORITY; ++i) @@ -1563,9 +1579,8 @@ int ttm_bo_global_init(struct ttm_bo_global *glob) &glob->kobj, &ttm_bo_glob_kobj_type, ttm_get_kobj(), "buffer_objects"); if (unlikely(ret != 0)) kobject_put(&glob->kobj); - return ret; -out_no_drp: - kfree(glob); +out: + mutex_unlock(&ttm_global_mutex); return ret; } EXPORT_SYMBOL(ttm_bo_global_init); diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 9cec8835b88f..26be74939f10 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -398,7 +398,7 @@ struct ttm_bo_driver { * @swap_lru: Lru list of buffer objects used for swapping. */ -struct ttm_bo_global { +extern struct ttm_bo_global { /** * Constant after init. @@ -410,8 +410,9 @@ struct ttm_bo_global { spinlock_t lru_lock; /** - * Protected by device_list_mutex. + * Protected by ttm_global_mutex. */ + unsigned int use_count; struct list_head device_list; /** @@ -423,7 +424,7 @@ struct ttm_bo_global { * Internal protection. */ atomic_t bo_count; -}; +} ttm_bo_glob; #define TTM_NUM_MEM_TYPES 8 @@ -568,8 +569,8 @@ void ttm_bo_mem_put(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); void ttm_bo_mem_put_locked(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); -void ttm_bo_global_release(struct ttm_bo_global *glob); -int ttm_bo_global_init(struct ttm_bo_global *glob); +void ttm_bo_global_release(void); +int ttm_bo_global_init(void); int ttm_bo_device_release(struct ttm_bo_device *bdev); @@ -906,7 +907,7 @@ struct ttm_bo_global_ref { */ static inline int ttm_bo_global_ref_init(struct drm_global_reference *ref) { - return ttm_bo_global_init(ref->object); + return ttm_bo_global_init(); } /** @@ -920,7 +921,7 @@ static inline int ttm_bo_global_ref_init(struct drm_global_reference *ref) */ static inline void ttm_bo_global_ref_release(struct drm_global_reference *ref) { - ttm_bo_global_release(ref->object); + ttm_bo_global_release(); } #endif -- cgit v1.2.3 From a64f784bb14a56bfdfad2dc397dd67e4564e3a29 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 19 Oct 2018 16:55:26 +0200 Subject: drm/ttm: initialize globals during device init (v2) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Make sure that the global BO state is always correctly initialized. This allows removing all the device code to initialize it. v2: fix up vbox (Alex) Signed-off-by: Christian König Reviewed-by: Junwei Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 59 +------------------------ drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 1 - drivers/gpu/drm/ast/ast_drv.h | 1 - drivers/gpu/drm/ast/ast_ttm.c | 36 --------------- drivers/gpu/drm/bochs/bochs.h | 1 - drivers/gpu/drm/bochs/bochs_mm.c | 35 --------------- drivers/gpu/drm/cirrus/cirrus_drv.h | 1 - drivers/gpu/drm/cirrus/cirrus_ttm.c | 36 --------------- drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h | 1 - drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c | 34 -------------- drivers/gpu/drm/mgag200/mgag200_drv.h | 1 - drivers/gpu/drm/mgag200/mgag200_ttm.c | 36 --------------- drivers/gpu/drm/nouveau/nouveau_drv.h | 1 - drivers/gpu/drm/nouveau/nouveau_ttm.c | 39 ---------------- drivers/gpu/drm/qxl/qxl_drv.h | 2 - drivers/gpu/drm/qxl/qxl_ttm.c | 33 -------------- drivers/gpu/drm/radeon/radeon.h | 2 - drivers/gpu/drm/radeon/radeon_ttm.c | 39 ---------------- drivers/gpu/drm/ttm/ttm_bo.c | 19 +++++--- drivers/gpu/drm/virtio/virtgpu_drv.h | 2 - drivers/gpu/drm/virtio/virtgpu_ttm.c | 35 --------------- drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 11 +---- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 3 -- drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c | 27 ----------- drivers/staging/vboxvideo/vbox_drv.h | 1 - drivers/staging/vboxvideo/vbox_ttm.c | 41 +---------------- include/drm/ttm/ttm_bo_driver.h | 41 +---------------- 27 files changed, 17 insertions(+), 521 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index fda252022b15..31fe85dd0b50 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -61,56 +61,6 @@ static int amdgpu_map_buffer(struct ttm_buffer_object *bo, static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev); static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev); -/* - * Global memory. - */ - -/** - * amdgpu_ttm_global_init - Initialize global TTM memory reference structures. - * - * @adev: AMDGPU device for which the global structures need to be registered. - * - * This is called as part of the AMDGPU ttm init from amdgpu_ttm_init() - * during bring up. - */ -static int amdgpu_ttm_global_init(struct amdgpu_device *adev) -{ - struct drm_global_reference *global_ref; - int r; - - /* ensure reference is false in case init fails */ - adev->mman.mem_global_referenced = false; - - global_ref = &adev->mman.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - goto error_bo; - } - - mutex_init(&adev->mman.gtt_window_lock); - - adev->mman.mem_global_referenced = true; - - return 0; - -error_bo: - return r; -} - -static void amdgpu_ttm_global_fini(struct amdgpu_device *adev) -{ - if (adev->mman.mem_global_referenced) { - mutex_destroy(&adev->mman.gtt_window_lock); - drm_global_item_unref(&adev->mman.bo_global_ref.ref); - adev->mman.mem_global_referenced = false; - } -} - static int amdgpu_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags) { return 0; @@ -1714,14 +1664,10 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) int r; u64 vis_vram_limit; - /* initialize global references for vram/gtt */ - r = amdgpu_ttm_global_init(adev); - if (r) { - return r; - } + mutex_init(&adev->mman.gtt_window_lock); + /* No others user of address space so set it to 0 */ r = ttm_bo_device_init(&adev->mman.bdev, - adev->mman.bo_global_ref.ref.object, &amdgpu_bo_driver, adev->ddev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -1878,7 +1824,6 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev) ttm_bo_clean_mm(&adev->mman.bdev, AMDGPU_PL_GWS); ttm_bo_clean_mm(&adev->mman.bdev, AMDGPU_PL_OA); ttm_bo_device_release(&adev->mman.bdev); - amdgpu_ttm_global_fini(adev); adev->mman.initialized = false; DRM_INFO("amdgpu: ttm finalized\n"); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index e114f209b701..b5b2d101f7db 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -39,7 +39,6 @@ #define AMDGPU_GTT_NUM_TRANSFER_WINDOWS 2 struct amdgpu_mman { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; bool mem_global_referenced; bool initialized; diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h index 6ae11a477643..bfc65040dfcb 100644 --- a/drivers/gpu/drm/ast/ast_drv.h +++ b/drivers/gpu/drm/ast/ast_drv.h @@ -104,7 +104,6 @@ struct ast_private { int fb_mtrr; struct { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index 8a59d6fc1160..c168d62fe8f9 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -36,35 +36,6 @@ ast_bdev(struct ttm_bo_device *bd) return container_of(bd, struct ast_private, ttm.bdev); } -static int ast_ttm_global_init(struct ast_private *ast) -{ - struct drm_global_reference *global_ref; - int r; - - global_ref = &ast->ttm.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - return 0; -} - -static void -ast_ttm_global_release(struct ast_private *ast) -{ - if (ast->ttm.bo_global_ref.ref.release == NULL) - return; - - drm_global_item_unref(&ast->ttm.bo_global_ref.ref); - ast->ttm.bo_global_ref.ref.release = NULL; -} - - static void ast_bo_ttm_destroy(struct ttm_buffer_object *tbo) { struct ast_bo *bo; @@ -204,12 +175,7 @@ int ast_mm_init(struct ast_private *ast) struct drm_device *dev = ast->dev; struct ttm_bo_device *bdev = &ast->ttm.bdev; - ret = ast_ttm_global_init(ast); - if (ret) - return ret; - ret = ttm_bo_device_init(&ast->ttm.bdev, - ast->ttm.bo_global_ref.ref.object, &ast_bo_driver, dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -240,8 +206,6 @@ void ast_mm_fini(struct ast_private *ast) ttm_bo_device_release(&ast->ttm.bdev); - ast_ttm_global_release(ast); - arch_phys_wc_del(ast->fb_mtrr); arch_io_free_memtype_wc(pci_resource_start(dev->pdev, 0), pci_resource_len(dev->pdev, 0)); diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h index a035257d1b56..75e4cf6cda5d 100644 --- a/drivers/gpu/drm/bochs/bochs.h +++ b/drivers/gpu/drm/bochs/bochs.h @@ -76,7 +76,6 @@ struct bochs_device { /* ttm */ struct { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; bool initialized; } ttm; diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c index c697d456656f..3bd773ef78dd 100644 --- a/drivers/gpu/drm/bochs/bochs_mm.c +++ b/drivers/gpu/drm/bochs/bochs_mm.c @@ -16,35 +16,6 @@ static inline struct bochs_device *bochs_bdev(struct ttm_bo_device *bd) return container_of(bd, struct bochs_device, ttm.bdev); } -static int bochs_ttm_global_init(struct bochs_device *bochs) -{ - struct drm_global_reference *global_ref; - int r; - - global_ref = &bochs->ttm.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - - return 0; -} - -static void bochs_ttm_global_release(struct bochs_device *bochs) -{ - if (bochs->ttm.bo_global_ref.ref.release == NULL) - return; - - drm_global_item_unref(&bochs->ttm.bo_global_ref.ref); - bochs->ttm.bo_global_ref.ref.release = NULL; -} - - static void bochs_bo_ttm_destroy(struct ttm_buffer_object *tbo) { struct bochs_bo *bo; @@ -182,12 +153,7 @@ int bochs_mm_init(struct bochs_device *bochs) struct ttm_bo_device *bdev = &bochs->ttm.bdev; int ret; - ret = bochs_ttm_global_init(bochs); - if (ret) - return ret; - ret = ttm_bo_device_init(&bochs->ttm.bdev, - bochs->ttm.bo_global_ref.ref.object, &bochs_bo_driver, bochs->dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -214,7 +180,6 @@ void bochs_mm_fini(struct bochs_device *bochs) return; ttm_bo_device_release(&bochs->ttm.bdev); - bochs_ttm_global_release(bochs); bochs->ttm.initialized = false; } diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h index 01852fbda9da..f2b2e0d169fa 100644 --- a/drivers/gpu/drm/cirrus/cirrus_drv.h +++ b/drivers/gpu/drm/cirrus/cirrus_drv.h @@ -136,7 +136,6 @@ struct cirrus_device { int fb_mtrr; struct { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; bool mm_inited; diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c index 7801c56c3c9b..e075810b4bd4 100644 --- a/drivers/gpu/drm/cirrus/cirrus_ttm.c +++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c @@ -36,35 +36,6 @@ cirrus_bdev(struct ttm_bo_device *bd) return container_of(bd, struct cirrus_device, ttm.bdev); } -static int cirrus_ttm_global_init(struct cirrus_device *cirrus) -{ - struct drm_global_reference *global_ref; - int r; - - global_ref = &cirrus->ttm.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - return 0; -} - -static void -cirrus_ttm_global_release(struct cirrus_device *cirrus) -{ - if (cirrus->ttm.bo_global_ref.ref.release == NULL) - return; - - drm_global_item_unref(&cirrus->ttm.bo_global_ref.ref); - cirrus->ttm.bo_global_ref.ref.release = NULL; -} - - static void cirrus_bo_ttm_destroy(struct ttm_buffer_object *tbo) { struct cirrus_bo *bo; @@ -204,12 +175,7 @@ int cirrus_mm_init(struct cirrus_device *cirrus) struct drm_device *dev = cirrus->dev; struct ttm_bo_device *bdev = &cirrus->ttm.bdev; - ret = cirrus_ttm_global_init(cirrus); - if (ret) - return ret; - ret = ttm_bo_device_init(&cirrus->ttm.bdev, - cirrus->ttm.bo_global_ref.ref.object, &cirrus_bo_driver, dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -245,8 +211,6 @@ void cirrus_mm_fini(struct cirrus_device *cirrus) ttm_bo_device_release(&cirrus->ttm.bdev); - cirrus_ttm_global_release(cirrus); - arch_phys_wc_del(cirrus->fb_mtrr); cirrus->fb_mtrr = 0; arch_io_free_memtype_wc(pci_resource_start(dev->pdev, 0), diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h index 60479502e277..3c168ae77b0c 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h @@ -49,7 +49,6 @@ struct hibmc_drm_private { bool mode_config_initialized; /* ttm */ - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; bool initialized; diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c index 14071c849121..dd383267884c 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c @@ -29,32 +29,6 @@ hibmc_bdev(struct ttm_bo_device *bd) return container_of(bd, struct hibmc_drm_private, bdev); } -static int hibmc_ttm_global_init(struct hibmc_drm_private *hibmc) -{ - int ret; - - hibmc->bo_global_ref.ref.global_type = DRM_GLOBAL_TTM_BO; - hibmc->bo_global_ref.ref.size = sizeof(struct ttm_bo_global); - hibmc->bo_global_ref.ref.init = &ttm_bo_global_ref_init; - hibmc->bo_global_ref.ref.release = &ttm_bo_global_ref_release; - ret = drm_global_item_ref(&hibmc->bo_global_ref.ref); - if (ret) { - DRM_ERROR("failed setting up TTM BO subsystem: %d\n", ret); - return ret; - } - return 0; -} - -static void -hibmc_ttm_global_release(struct hibmc_drm_private *hibmc) -{ - if (hibmc->bo_global_ref.ref.release == NULL) - return; - - drm_global_item_unref(&hibmc->bo_global_ref.ref); - hibmc->bo_global_ref.ref.release = NULL; -} - static void hibmc_bo_ttm_destroy(struct ttm_buffer_object *tbo) { struct hibmc_bo *bo = container_of(tbo, struct hibmc_bo, bo); @@ -214,18 +188,12 @@ int hibmc_mm_init(struct hibmc_drm_private *hibmc) struct drm_device *dev = hibmc->dev; struct ttm_bo_device *bdev = &hibmc->bdev; - ret = hibmc_ttm_global_init(hibmc); - if (ret) - return ret; - ret = ttm_bo_device_init(&hibmc->bdev, - hibmc->bo_global_ref.ref.object, &hibmc_bo_driver, dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, true); if (ret) { - hibmc_ttm_global_release(hibmc); DRM_ERROR("error initializing bo driver: %d\n", ret); return ret; } @@ -233,7 +201,6 @@ int hibmc_mm_init(struct hibmc_drm_private *hibmc) ret = ttm_bo_init_mm(bdev, TTM_PL_VRAM, hibmc->fb_size >> PAGE_SHIFT); if (ret) { - hibmc_ttm_global_release(hibmc); DRM_ERROR("failed ttm VRAM init: %d\n", ret); return ret; } @@ -248,7 +215,6 @@ void hibmc_mm_fini(struct hibmc_drm_private *hibmc) return; ttm_bo_device_release(&hibmc->bdev); - hibmc_ttm_global_release(hibmc); hibmc->mm_inited = false; } diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.h b/drivers/gpu/drm/mgag200/mgag200_drv.h index e5348955a3cc..0aaedc554879 100644 --- a/drivers/gpu/drm/mgag200/mgag200_drv.h +++ b/drivers/gpu/drm/mgag200/mgag200_drv.h @@ -212,7 +212,6 @@ struct mga_device { int fb_mtrr; struct { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c index 11bdc8121a7d..d96a9b32455e 100644 --- a/drivers/gpu/drm/mgag200/mgag200_ttm.c +++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c @@ -36,35 +36,6 @@ mgag200_bdev(struct ttm_bo_device *bd) return container_of(bd, struct mga_device, ttm.bdev); } -static int mgag200_ttm_global_init(struct mga_device *ast) -{ - struct drm_global_reference *global_ref; - int r; - - global_ref = &ast->ttm.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - return 0; -} - -static void -mgag200_ttm_global_release(struct mga_device *ast) -{ - if (ast->ttm.bo_global_ref.ref.release == NULL) - return; - - drm_global_item_unref(&ast->ttm.bo_global_ref.ref); - ast->ttm.bo_global_ref.ref.release = NULL; -} - - static void mgag200_bo_ttm_destroy(struct ttm_buffer_object *tbo) { struct mgag200_bo *bo; @@ -204,12 +175,7 @@ int mgag200_mm_init(struct mga_device *mdev) struct drm_device *dev = mdev->dev; struct ttm_bo_device *bdev = &mdev->ttm.bdev; - ret = mgag200_ttm_global_init(mdev); - if (ret) - return ret; - ret = ttm_bo_device_init(&mdev->ttm.bdev, - mdev->ttm.bo_global_ref.ref.object, &mgag200_bo_driver, dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -240,8 +206,6 @@ void mgag200_mm_fini(struct mga_device *mdev) ttm_bo_device_release(&mdev->ttm.bdev); - mgag200_ttm_global_release(mdev); - arch_io_free_memtype_wc(pci_resource_start(dev->pdev, 0), pci_resource_len(dev->pdev, 0)); arch_phys_wc_del(mdev->fb_mtrr); diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index 0f3bb1a11fc9..d20b9ba4b1c1 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -146,7 +146,6 @@ struct nouveau_drm { /* TTM interface support */ struct { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; atomic_t validate_sequence; int (*move)(struct nouveau_channel *, diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index 69448b02649e..1543c2f8d3d3 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -174,38 +174,6 @@ nouveau_ttm_mmap(struct file *filp, struct vm_area_struct *vma) return ttm_bo_mmap(filp, vma, &drm->ttm.bdev); } -int -nouveau_ttm_global_init(struct nouveau_drm *drm) -{ - struct drm_global_reference *global_ref; - int ret; - - global_ref = &drm->ttm.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - - ret = drm_global_item_ref(global_ref); - if (unlikely(ret != 0)) { - DRM_ERROR("Failed setting up TTM BO subsystem\n"); - drm->ttm.bo_global_ref.ref.release = NULL; - return ret; - } - - return 0; -} - -void -nouveau_ttm_global_release(struct nouveau_drm *drm) -{ - if (drm->ttm.bo_global_ref.ref.release == NULL) - return; - - drm_global_item_unref(&drm->ttm.bo_global_ref.ref); - drm->ttm.bo_global_ref.ref.release = NULL; -} - static int nouveau_ttm_init_host(struct nouveau_drm *drm, u8 kind) { @@ -268,12 +236,7 @@ nouveau_ttm_init(struct nouveau_drm *drm) drm->agp.cma = pci->agp.cma; } - ret = nouveau_ttm_global_init(drm); - if (ret) - return ret; - ret = ttm_bo_device_init(&drm->ttm.bdev, - drm->ttm.bo_global_ref.ref.object, &nouveau_bo_driver, dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -328,8 +291,6 @@ nouveau_ttm_fini(struct nouveau_drm *drm) ttm_bo_device_release(&drm->ttm.bdev); - nouveau_ttm_global_release(drm); - arch_phys_wc_del(drm->ttm.mtrr); drm->ttm.mtrr = 0; arch_io_free_memtype_wc(device->func->resource_addr(device, 1), diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h index 52912e54e990..2310d6e9ff1f 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.h +++ b/drivers/gpu/drm/qxl/qxl_drv.h @@ -127,8 +127,6 @@ struct qxl_output { #define drm_encoder_to_qxl_output(x) container_of(x, struct qxl_output, enc) struct qxl_mman { - struct ttm_bo_global_ref bo_global_ref; - bool mem_global_referenced; struct ttm_bo_device bdev; }; diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index bb8cc9b16780..1468fddc19d0 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -46,34 +46,6 @@ static struct qxl_device *qxl_get_qdev(struct ttm_bo_device *bdev) return qdev; } -static int qxl_ttm_global_init(struct qxl_device *qdev) -{ - struct drm_global_reference *global_ref; - int r; - - global_ref = &qdev->mman.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - - qdev->mman.mem_global_referenced = true; - return 0; -} - -static void qxl_ttm_global_fini(struct qxl_device *qdev) -{ - if (qdev->mman.mem_global_referenced) { - drm_global_item_unref(&qdev->mman.bo_global_ref.ref); - qdev->mman.mem_global_referenced = false; - } -} - static struct vm_operations_struct qxl_ttm_vm_ops; static const struct vm_operations_struct *ttm_vm_ops; @@ -345,12 +317,8 @@ int qxl_ttm_init(struct qxl_device *qdev) int r; int num_io_pages; /* != rom->num_io_pages, we include surface0 */ - r = qxl_ttm_global_init(qdev); - if (r) - return r; /* No others user of address space so set it to 0 */ r = ttm_bo_device_init(&qdev->mman.bdev, - qdev->mman.bo_global_ref.ref.object, &qxl_bo_driver, qdev->ddev.anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, 0); @@ -386,7 +354,6 @@ void qxl_ttm_fini(struct qxl_device *qdev) ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_VRAM); ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_PRIV); ttm_bo_device_release(&qdev->mman.bdev); - qxl_ttm_global_fini(qdev); DRM_INFO("qxl: ttm finalized\n"); } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 06fb952b6290..32808e50be12 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -448,9 +448,7 @@ struct radeon_surface_reg { * TTM. */ struct radeon_mman { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; - bool mem_global_referenced; bool initialized; #if defined(CONFIG_DEBUG_FS) diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 2104429ddbac..9920a6fc11bf 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -60,39 +60,6 @@ static struct radeon_device *radeon_get_rdev(struct ttm_bo_device *bdev) return rdev; } - -/* - * Global memory. - */ -static int radeon_ttm_global_init(struct radeon_device *rdev) -{ - struct drm_global_reference *global_ref; - int r; - - rdev->mman.mem_global_referenced = false; - global_ref = &rdev->mman.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - - rdev->mman.mem_global_referenced = true; - return 0; -} - -static void radeon_ttm_global_fini(struct radeon_device *rdev) -{ - if (rdev->mman.mem_global_referenced) { - drm_global_item_unref(&rdev->mman.bo_global_ref.ref); - rdev->mman.mem_global_referenced = false; - } -} - static int radeon_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags) { return 0; @@ -821,13 +788,8 @@ int radeon_ttm_init(struct radeon_device *rdev) { int r; - r = radeon_ttm_global_init(rdev); - if (r) { - return r; - } /* No others user of address space so set it to 0 */ r = ttm_bo_device_init(&rdev->mman.bdev, - rdev->mman.bo_global_ref.ref.object, &radeon_bo_driver, rdev->ddev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, @@ -899,7 +861,6 @@ void radeon_ttm_fini(struct radeon_device *rdev) ttm_bo_clean_mm(&rdev->mman.bdev, TTM_PL_TT); ttm_bo_device_release(&rdev->mman.bdev); radeon_gart_fini(rdev); - radeon_ttm_global_fini(rdev); rdev->mman.initialized = false; DRM_INFO("radeon: ttm finalized\n"); } diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index d89183f95570..df028805b7e2 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1530,7 +1530,7 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj) kfree(glob); } -void ttm_bo_global_release(void) +static void ttm_bo_global_release(void) { struct ttm_bo_global *glob = &ttm_bo_glob; @@ -1544,9 +1544,8 @@ void ttm_bo_global_release(void) out: mutex_unlock(&ttm_global_mutex); } -EXPORT_SYMBOL(ttm_bo_global_release); -int ttm_bo_global_init(void) +static int ttm_bo_global_init(void) { struct ttm_bo_global *glob = &ttm_bo_glob; int ret = 0; @@ -1583,8 +1582,6 @@ out: mutex_unlock(&ttm_global_mutex); return ret; } -EXPORT_SYMBOL(ttm_bo_global_init); - int ttm_bo_device_release(struct ttm_bo_device *bdev) { @@ -1623,18 +1620,25 @@ int ttm_bo_device_release(struct ttm_bo_device *bdev) drm_vma_offset_manager_destroy(&bdev->vma_manager); + if (!ret) + ttm_bo_global_release(); + return ret; } EXPORT_SYMBOL(ttm_bo_device_release); int ttm_bo_device_init(struct ttm_bo_device *bdev, - struct ttm_bo_global *glob, struct ttm_bo_driver *driver, struct address_space *mapping, uint64_t file_page_offset, bool need_dma32) { - int ret = -EINVAL; + struct ttm_bo_global *glob = &ttm_bo_glob; + int ret; + + ret = ttm_bo_global_init(); + if (ret) + return ret; bdev->driver = driver; @@ -1661,6 +1665,7 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev, return 0; out_no_sys: + ttm_bo_global_release(); return ret; } EXPORT_SYMBOL(ttm_bo_device_init); diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h index d9756b59e329..1acbf182536f 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.h +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h @@ -142,8 +142,6 @@ struct virtio_gpu_fbdev { }; struct virtio_gpu_mman { - struct ttm_bo_global_ref bo_global_ref; - bool mem_global_referenced; struct ttm_bo_device bdev; }; diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c index 8510109e09da..8fc088d5ef0d 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ttm.c +++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c @@ -50,35 +50,6 @@ virtio_gpu_device *virtio_gpu_get_vgdev(struct ttm_bo_device *bdev) return vgdev; } -static int virtio_gpu_ttm_global_init(struct virtio_gpu_device *vgdev) -{ - struct drm_global_reference *global_ref; - int r; - - vgdev->mman.mem_global_referenced = false; - global_ref = &vgdev->mman.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - r = drm_global_item_ref(global_ref); - if (r != 0) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return r; - } - - vgdev->mman.mem_global_referenced = true; - return 0; -} - -static void virtio_gpu_ttm_global_fini(struct virtio_gpu_device *vgdev) -{ - if (vgdev->mman.mem_global_referenced) { - drm_global_item_unref(&vgdev->mman.bo_global_ref.ref); - vgdev->mman.mem_global_referenced = false; - } -} - int virtio_gpu_mmap(struct file *filp, struct vm_area_struct *vma) { struct drm_file *file_priv; @@ -356,12 +327,8 @@ int virtio_gpu_ttm_init(struct virtio_gpu_device *vgdev) { int r; - r = virtio_gpu_ttm_global_init(vgdev); - if (r) - return r; /* No others user of address space so set it to 0 */ r = ttm_bo_device_init(&vgdev->mman.bdev, - vgdev->mman.bo_global_ref.ref.object, &virtio_gpu_bo_driver, vgdev->ddev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, 0); @@ -380,13 +347,11 @@ int virtio_gpu_ttm_init(struct virtio_gpu_device *vgdev) err_mm_init: ttm_bo_device_release(&vgdev->mman.bdev); err_dev_init: - virtio_gpu_ttm_global_fini(vgdev); return r; } void virtio_gpu_ttm_fini(struct virtio_gpu_device *vgdev) { ttm_bo_device_release(&vgdev->mman.bdev); - virtio_gpu_ttm_global_fini(vgdev); DRM_INFO("virtio_gpu: ttm finalized\n"); } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c index 67494148accd..b9c078860a7c 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c @@ -801,11 +801,6 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset) DRM_INFO("MMIO at 0x%08x size is %u kiB\n", dev_priv->mmio_start, dev_priv->mmio_size / 1024); - ret = vmw_ttm_global_init(dev_priv); - if (unlikely(ret != 0)) - goto out_err0; - - vmw_master_init(&dev_priv->fbdev_master); ttm_lock_set_kill(&dev_priv->fbdev_master.lock, false, SIGTERM); dev_priv->active_master = &dev_priv->fbdev_master; @@ -816,7 +811,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset) if (unlikely(dev_priv->mmio_virt == NULL)) { ret = -ENOMEM; DRM_ERROR("Failed mapping MMIO.\n"); - goto out_err3; + goto out_err0; } /* Need mmio memory to check for fifo pitchlock cap. */ @@ -870,7 +865,6 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset) } ret = ttm_bo_device_init(&dev_priv->bdev, - dev_priv->bo_global_ref.ref.object, &vmw_bo_driver, dev->anon_inode->i_mapping, VMWGFX_FILE_PAGE_OFFSET, @@ -992,8 +986,6 @@ out_no_device: ttm_object_device_release(&dev_priv->tdev); out_err4: memunmap(dev_priv->mmio_virt); -out_err3: - vmw_ttm_global_release(dev_priv); out_err0: for (i = vmw_res_context; i < vmw_res_max; ++i) idr_destroy(&dev_priv->res_idr[i]); @@ -1045,7 +1037,6 @@ static void vmw_driver_unload(struct drm_device *dev) memunmap(dev_priv->mmio_virt); if (dev_priv->ctx.staged_bindings) vmw_binding_state_free(dev_priv->ctx.staged_bindings); - vmw_ttm_global_release(dev_priv); for (i = vmw_res_context; i < vmw_res_max; ++i) idr_destroy(&dev_priv->res_idr[i]); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 252f202ae897..28df788da44e 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -417,7 +417,6 @@ enum { struct vmw_private { struct ttm_bo_device bdev; - struct ttm_bo_global_ref bo_global_ref; struct vmw_fifo_state fifo; @@ -841,8 +840,6 @@ extern int vmw_fifo_flush(struct vmw_private *dev_priv, * TTM glue - vmwgfx_ttm_glue.c */ -extern int vmw_ttm_global_init(struct vmw_private *dev_priv); -extern void vmw_ttm_global_release(struct vmw_private *dev_priv); extern int vmw_mmap(struct file *filp, struct vm_area_struct *vma); /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c index 0ac473cd5136..154eb09aa91e 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c @@ -42,30 +42,3 @@ int vmw_mmap(struct file *filp, struct vm_area_struct *vma) dev_priv = vmw_priv(file_priv->minor->dev); return ttm_bo_mmap(filp, vma, &dev_priv->bdev); } - -int vmw_ttm_global_init(struct vmw_private *dev_priv) -{ - struct drm_global_reference *global_ref; - int ret; - - global_ref = &dev_priv->bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - ret = drm_global_item_ref(global_ref); - - if (unlikely(ret != 0)) { - DRM_ERROR("Failed setting up TTM buffer objects.\n"); - goto out_no_bo; - } - - return 0; -out_no_bo: - return ret; -} - -void vmw_ttm_global_release(struct vmw_private *dev_priv) -{ - drm_global_item_unref(&dev_priv->bo_global_ref.ref); -} diff --git a/drivers/staging/vboxvideo/vbox_drv.h b/drivers/staging/vboxvideo/vbox_drv.h index dc257a892ecb..fa933d422951 100644 --- a/drivers/staging/vboxvideo/vbox_drv.h +++ b/drivers/staging/vboxvideo/vbox_drv.h @@ -99,7 +99,6 @@ struct vbox_private { int fb_mtrr; struct { - struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; } ttm; diff --git a/drivers/staging/vboxvideo/vbox_ttm.c b/drivers/staging/vboxvideo/vbox_ttm.c index 73b895328059..b36ec019c332 100644 --- a/drivers/staging/vboxvideo/vbox_ttm.c +++ b/drivers/staging/vboxvideo/vbox_ttm.c @@ -35,37 +35,6 @@ static inline struct vbox_private *vbox_bdev(struct ttm_bo_device *bd) return container_of(bd, struct vbox_private, ttm.bdev); } -/** - * Adds the vbox memory manager object/structures to the global memory manager. - */ -static int vbox_ttm_global_init(struct vbox_private *vbox) -{ - struct drm_global_reference *global_ref; - int ret; - - global_ref = &vbox->ttm.bo_global_ref.ref; - global_ref->global_type = DRM_GLOBAL_TTM_BO; - global_ref->size = sizeof(struct ttm_bo_global); - global_ref->init = &ttm_bo_global_ref_init; - global_ref->release = &ttm_bo_global_ref_release; - - ret = drm_global_item_ref(global_ref); - if (ret) { - DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - return ret; - } - - return 0; -} - -/** - * Removes the vbox memory manager object from the global memory manager. - */ -static void vbox_ttm_global_release(struct vbox_private *vbox) -{ - drm_global_item_unref(&vbox->ttm.bo_global_ref.ref); -} - static void vbox_bo_ttm_destroy(struct ttm_buffer_object *tbo) { struct vbox_bo *bo; @@ -203,18 +172,13 @@ int vbox_mm_init(struct vbox_private *vbox) struct drm_device *dev = &vbox->ddev; struct ttm_bo_device *bdev = &vbox->ttm.bdev; - ret = vbox_ttm_global_init(vbox); - if (ret) - return ret; - ret = ttm_bo_device_init(&vbox->ttm.bdev, - vbox->ttm.bo_global_ref.ref.object, &vbox_bo_driver, dev->anon_inode->i_mapping, DRM_FILE_PAGE_OFFSET, true); if (ret) { DRM_ERROR("Error initialising bo driver; %d\n", ret); - goto err_ttm_global_release; + return ret; } ret = ttm_bo_init_mm(bdev, TTM_PL_VRAM, @@ -236,8 +200,6 @@ int vbox_mm_init(struct vbox_private *vbox) err_device_release: ttm_bo_device_release(&vbox->ttm.bdev); -err_ttm_global_release: - vbox_ttm_global_release(vbox); return ret; } @@ -251,7 +213,6 @@ void vbox_mm_fini(struct vbox_private *vbox) arch_phys_wc_del(vbox->fb_mtrr); #endif ttm_bo_device_release(&vbox->ttm.bdev); - vbox_ttm_global_release(vbox); } void vbox_ttm_placement(struct vbox_bo *bo, int domain) diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 26be74939f10..6fb589f64633 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -569,9 +569,6 @@ void ttm_bo_mem_put(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); void ttm_bo_mem_put_locked(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); -void ttm_bo_global_release(void); -int ttm_bo_global_init(void); - int ttm_bo_device_release(struct ttm_bo_device *bdev); /** @@ -589,7 +586,7 @@ int ttm_bo_device_release(struct ttm_bo_device *bdev); * Returns: * !0: Failure. */ -int ttm_bo_device_init(struct ttm_bo_device *bdev, struct ttm_bo_global *glob, +int ttm_bo_device_init(struct ttm_bo_device *bdev, struct ttm_bo_driver *driver, struct address_space *mapping, uint64_t file_page_offset, bool need_dma32); @@ -888,40 +885,4 @@ pgprot_t ttm_io_prot(uint32_t caching_flags, pgprot_t tmp); extern const struct ttm_mem_type_manager_func ttm_bo_manager_func; -/** - * struct ttm_bo_global_ref - Argument to initialize a struct ttm_bo_global. - */ - -struct ttm_bo_global_ref { - struct drm_global_reference ref; -}; - -/** - * ttm_bo_global_ref_init - * - * @ref: DRM global reference - * - * Helper function that initializes a struct ttm_bo_global. This function - * is used as init call-back function for DRM global references of type - * DRM_GLOBAL_TTM_BO_REF. - */ -static inline int ttm_bo_global_ref_init(struct drm_global_reference *ref) -{ - return ttm_bo_global_init(); -} - -/** - * ttm_bo_global_ref_release - * - * @ref: DRM global reference - * - * Helper function that releases a struct ttm_bo_global. This function - * is used as release call-back function for DRM global references of type - * DRM_GLOBAL_TTM_BO_REF. - */ -static inline void ttm_bo_global_ref_release(struct drm_global_reference *ref) -{ - ttm_bo_global_release(); -} - #endif -- cgit v1.2.3 From 2bb42410b1bd324912389c6ac748df1c1befd69f Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Fri, 19 Oct 2018 10:54:23 +0200 Subject: drm: Remove drm_global.{c,h} v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The content of drm_global.{c,h} is obsolete. v2: rebase on dropping TTM functionality Signed-off-by: Thomas Zimmermann Signed-off-by: Christian König Reviewed-by: Junwei Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/Makefile | 2 +- drivers/gpu/drm/drm_drv.c | 2 - drivers/gpu/drm/drm_global.c | 137 ---------------------------------------- include/drm/drmP.h | 1 - include/drm/drm_global.h | 53 ---------------- include/drm/ttm/ttm_bo_driver.h | 1 - 6 files changed, 1 insertion(+), 195 deletions(-) delete mode 100644 drivers/gpu/drm/drm_global.c delete mode 100644 include/drm/drm_global.h (limited to 'include') diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index bc6a16a3c36e..32a837b72765 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -11,7 +11,7 @@ drm-y := drm_auth.o drm_bufs.o drm_cache.o \ drm_sysfs.o drm_hashtab.o drm_mm.o \ drm_crtc.o drm_fourcc.o drm_modes.o drm_edid.o \ drm_info.o drm_encoder_slave.o \ - drm_trace_points.o drm_global.o drm_prime.o \ + drm_trace_points.o drm_prime.o \ drm_rect.o drm_vma_manager.o drm_flip_work.o \ drm_modeset_lock.o drm_atomic.o drm_bridge.o \ drm_framebuffer.o drm_connector.o drm_blend.o \ diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c index 36e8e9cbec52..fc29f46b7c32 100644 --- a/drivers/gpu/drm/drm_drv.c +++ b/drivers/gpu/drm/drm_drv.c @@ -975,14 +975,12 @@ static void drm_core_exit(void) drm_sysfs_destroy(); idr_destroy(&drm_minors_idr); drm_connector_ida_destroy(); - drm_global_release(); } static int __init drm_core_init(void) { int ret; - drm_global_init(); drm_connector_ida_init(); idr_init(&drm_minors_idr); diff --git a/drivers/gpu/drm/drm_global.c b/drivers/gpu/drm/drm_global.c deleted file mode 100644 index 5799e2782dd1..000000000000 --- a/drivers/gpu/drm/drm_global.c +++ /dev/null @@ -1,137 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 OR MIT -/************************************************************************** - * - * Copyright 2008-2009 VMware, Inc., Palo Alto, CA., USA - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the - * "Software"), to deal in the Software without restriction, including - * without limitation the rights to use, copy, modify, merge, publish, - * distribute, sub license, and/or sell copies of the Software, and to - * permit persons to whom the Software is furnished to do so, subject to - * the following conditions: - * - * The above copyright notice and this permission notice (including the - * next paragraph) shall be included in all copies or substantial portions - * of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, - * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR - * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE - * USE OR OTHER DEALINGS IN THE SOFTWARE. - * - **************************************************************************/ -/* - * Authors: Thomas Hellstrom - */ - -#include -#include -#include -#include - -struct drm_global_item { - struct mutex mutex; - void *object; - int refcount; -}; - -static struct drm_global_item glob[DRM_GLOBAL_NUM]; - -void drm_global_init(void) -{ - int i; - - for (i = 0; i < DRM_GLOBAL_NUM; ++i) { - struct drm_global_item *item = &glob[i]; - mutex_init(&item->mutex); - item->object = NULL; - item->refcount = 0; - } -} - -void drm_global_release(void) -{ - int i; - for (i = 0; i < DRM_GLOBAL_NUM; ++i) { - struct drm_global_item *item = &glob[i]; - BUG_ON(item->object != NULL); - BUG_ON(item->refcount != 0); - } -} - -/** - * drm_global_item_ref - Initialize and acquire reference to memory - * object - * @ref: Object for initialization - * - * This initializes a memory object, allocating memory and calling the - * .init() hook. Further calls will increase the reference count for - * that item. - * - * Returns: - * Zero on success, non-zero otherwise. - */ -int drm_global_item_ref(struct drm_global_reference *ref) -{ - int ret = 0; - struct drm_global_item *item = &glob[ref->global_type]; - - mutex_lock(&item->mutex); - if (item->refcount == 0) { - ref->object = kzalloc(ref->size, GFP_KERNEL); - if (unlikely(ref->object == NULL)) { - ret = -ENOMEM; - goto error_unlock; - } - ret = ref->init(ref); - if (unlikely(ret != 0)) - goto error_free; - - item->object = ref->object; - } else { - ref->object = item->object; - } - - ++item->refcount; - mutex_unlock(&item->mutex); - return 0; - -error_free: - kfree(ref->object); - ref->object = NULL; -error_unlock: - mutex_unlock(&item->mutex); - return ret; -} -EXPORT_SYMBOL(drm_global_item_ref); - -/** - * drm_global_item_unref - Drop reference to memory - * object - * @ref: Object being removed - * - * Drop a reference to the memory object and eventually call the - * release() hook. The allocated object should be dropped in the - * release() hook or before calling this function - * - */ - -void drm_global_item_unref(struct drm_global_reference *ref) -{ - struct drm_global_item *item = &glob[ref->global_type]; - - mutex_lock(&item->mutex); - BUG_ON(item->refcount == 0); - BUG_ON(ref->object != item->object); - if (--item->refcount == 0) { - ref->release(ref); - item->object = NULL; - } - mutex_unlock(&item->mutex); -} -EXPORT_SYMBOL(drm_global_item_unref); - diff --git a/include/drm/drmP.h b/include/drm/drmP.h index 05350424a4d3..2557001d1b21 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -68,7 +68,6 @@ #include #include #include -#include #include #include #include diff --git a/include/drm/drm_global.h b/include/drm/drm_global.h deleted file mode 100644 index 3a830602a2e4..000000000000 --- a/include/drm/drm_global.h +++ /dev/null @@ -1,53 +0,0 @@ -/************************************************************************** - * - * Copyright 2008-2009 VMware, Inc., Palo Alto, CA., USA - * All Rights Reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the - * "Software"), to deal in the Software without restriction, including - * without limitation the rights to use, copy, modify, merge, publish, - * distribute, sub license, and/or sell copies of the Software, and to - * permit persons to whom the Software is furnished to do so, subject to - * the following conditions: - * - * The above copyright notice and this permission notice (including the - * next paragraph) shall be included in all copies or substantial portions - * of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, - * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR - * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE - * USE OR OTHER DEALINGS IN THE SOFTWARE. - * - **************************************************************************/ -/* - * Authors: Thomas Hellstrom - */ - -#ifndef _DRM_GLOBAL_H_ -#define _DRM_GLOBAL_H_ -enum drm_global_types { - DRM_GLOBAL_TTM_MEM = 0, - DRM_GLOBAL_TTM_BO, - DRM_GLOBAL_TTM_OBJECT, - DRM_GLOBAL_NUM -}; - -struct drm_global_reference { - enum drm_global_types global_type; - size_t size; - void *object; - int (*init) (struct drm_global_reference *); - void (*release) (struct drm_global_reference *); -}; - -void drm_global_init(void); -void drm_global_release(void); -int drm_global_item_ref(struct drm_global_reference *ref); -void drm_global_item_unref(struct drm_global_reference *ref); - -#endif diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 6fb589f64633..1021106438b2 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -31,7 +31,6 @@ #define _TTM_BO_DRIVER_H_ #include -#include #include #include #include -- cgit v1.2.3 From faf6e1a87e07423a729e04fb2e8188742e89ea4c Mon Sep 17 00:00:00 2001 From: Andrey Grodzovsky Date: Thu, 18 Oct 2018 12:32:46 -0400 Subject: drm/sched: Add boolean to mark if sched is ready to work v5 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Problem: A particular scheduler may become unsuable (underlying HW) after some event (e.g. GPU reset). If it's later chosen by the get free sched. policy a command will fail to be submitted. Fix: Add a driver specific callback to report the sched status so rq with bad sched can be avoided in favor of working one or none in which case job init will fail. v2: Switch from driver callback to flag in scheduler. v3: rebase v4: Remove ready paramter from drm_sched_init, set uncoditionally to true once init done. v5: fix missed change in v3d in v4 (Alex) Signed-off-by: Andrey Grodzovsky Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/scheduler/sched_entity.c | 9 ++++++++- drivers/gpu/drm/scheduler/sched_main.c | 6 ++++++ include/drm/gpu_scheduler.h | 3 +++ 3 files changed, 17 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c index 3e22a54a99c2..ba54c30a466e 100644 --- a/drivers/gpu/drm/scheduler/sched_entity.c +++ b/drivers/gpu/drm/scheduler/sched_entity.c @@ -130,7 +130,14 @@ drm_sched_entity_get_free_sched(struct drm_sched_entity *entity) int i; for (i = 0; i < entity->num_rq_list; ++i) { - num_jobs = atomic_read(&entity->rq_list[i]->sched->num_jobs); + struct drm_gpu_scheduler *sched = entity->rq_list[i]->sched; + + if (!entity->rq_list[i]->sched->ready) { + DRM_WARN("sched%s is not ready, skipping", sched->name); + continue; + } + + num_jobs = atomic_read(&sched->num_jobs); if (num_jobs < min_jobs) { min_jobs = num_jobs; rq = entity->rq_list[i]; diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c index 63b997d9c562..6b2fd49334f7 100644 --- a/drivers/gpu/drm/scheduler/sched_main.c +++ b/drivers/gpu/drm/scheduler/sched_main.c @@ -420,6 +420,9 @@ int drm_sched_job_init(struct drm_sched_job *job, struct drm_gpu_scheduler *sched; drm_sched_entity_select_rq(entity); + if (!entity->rq) + return -ENOENT; + sched = entity->rq->sched; job->sched = sched; @@ -633,6 +636,7 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, return PTR_ERR(sched->thread); } + sched->ready = true; return 0; } EXPORT_SYMBOL(drm_sched_init); @@ -648,5 +652,7 @@ void drm_sched_fini(struct drm_gpu_scheduler *sched) { if (sched->thread) kthread_stop(sched->thread); + + sched->ready = false; } EXPORT_SYMBOL(drm_sched_fini); diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index 0684dcd99c0f..4ae192a21c3f 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -264,6 +264,7 @@ struct drm_sched_backend_ops { * @hang_limit: once the hangs by a job crosses this limit then it is marked * guilty and it will be considered for scheduling further. * @num_jobs: the number of jobs in queue in the scheduler + * @ready: marks if the underlying HW is ready to work * * One scheduler is implemented for each hardware ring. */ @@ -283,12 +284,14 @@ struct drm_gpu_scheduler { spinlock_t job_list_lock; int hang_limit; atomic_t num_jobs; + bool ready; }; int drm_sched_init(struct drm_gpu_scheduler *sched, const struct drm_sched_backend_ops *ops, uint32_t hw_submission, unsigned hang_limit, long timeout, const char *name); + void drm_sched_fini(struct drm_gpu_scheduler *sched); int drm_sched_job_init(struct drm_sched_job *job, struct drm_sched_entity *entity, -- cgit v1.2.3 From 26efecf9558895a89c2920d258601b4afba10fd0 Mon Sep 17 00:00:00 2001 From: Sharat Masetty Date: Mon, 29 Oct 2018 15:02:28 +0530 Subject: drm/scheduler: Add drm_sched_job_cleanup MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch adds a new API to clean up the scheduler job resources. This is primarliy needed in cases the job was created but was not queued to the scheduler queue. Additionally with this change, the layer which creates the scheduler job also gets to free up the job's resources and this entails moving the dma_fence_put(finished_fence) to the drivers ops free handler routines. Signed-off-by: Sharat Masetty Reviewed-by: Christian König Acked-by: Andrey Grodzovsky Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 3 +-- drivers/gpu/drm/amd/amdgpu/amdgpu_job.c | 2 ++ drivers/gpu/drm/etnaviv/etnaviv_sched.c | 3 +++ drivers/gpu/drm/scheduler/sched_entity.c | 1 - drivers/gpu/drm/scheduler/sched_main.c | 13 ++++++++++++- drivers/gpu/drm/v3d/v3d_sched.c | 2 ++ include/drm/gpu_scheduler.h | 1 + 7 files changed, 21 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 663043c8f0f5..5d768f95de24 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -1260,8 +1260,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, return 0; error_abort: - dma_fence_put(&job->base.s_fence->finished); - job->base.s_fence = NULL; + drm_sched_job_cleanup(&job->base); amdgpu_mn_unlock(p->mn); error_unlock: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 755f733bf0d9..e0af44fd6a0c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -112,6 +112,8 @@ static void amdgpu_job_free_cb(struct drm_sched_job *s_job) struct amdgpu_ring *ring = to_amdgpu_ring(s_job->sched); struct amdgpu_job *job = to_amdgpu_job(s_job); + drm_sched_job_cleanup(s_job); + amdgpu_ring_priority_put(ring, s_job->s_priority); dma_fence_put(job->fence); amdgpu_sync_free(&job->sync); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c index f8c5f1e6c5e2..88b2768d91c9 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c @@ -125,6 +125,8 @@ static void etnaviv_sched_free_job(struct drm_sched_job *sched_job) { struct etnaviv_gem_submit *submit = to_etnaviv_submit(sched_job); + drm_sched_job_cleanup(sched_job); + etnaviv_submit_put(submit); } @@ -157,6 +159,7 @@ int etnaviv_sched_push_job(struct drm_sched_entity *sched_entity, submit->out_fence, 0, INT_MAX, GFP_KERNEL); if (submit->out_fence_id < 0) { + drm_sched_job_cleanup(&submit->sched_job); ret = -ENOMEM; goto out_unlock; } diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c index ba54c30a466e..4463d3826ecb 100644 --- a/drivers/gpu/drm/scheduler/sched_entity.c +++ b/drivers/gpu/drm/scheduler/sched_entity.c @@ -211,7 +211,6 @@ static void drm_sched_entity_kill_jobs_cb(struct dma_fence *f, drm_sched_fence_finished(job->s_fence); WARN_ON(job->s_fence->parent); - dma_fence_put(&job->s_fence->finished); job->sched->ops->free_job(job); } diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c index 6b2fd49334f7..18ebbb05762e 100644 --- a/drivers/gpu/drm/scheduler/sched_main.c +++ b/drivers/gpu/drm/scheduler/sched_main.c @@ -233,7 +233,6 @@ static void drm_sched_job_finish(struct work_struct *work) drm_sched_start_timeout(sched); spin_unlock(&sched->job_list_lock); - dma_fence_put(&s_job->s_fence->finished); sched->ops->free_job(s_job); } @@ -440,6 +439,18 @@ int drm_sched_job_init(struct drm_sched_job *job, } EXPORT_SYMBOL(drm_sched_job_init); +/** + * drm_sched_job_cleanup - clean up scheduler job resources + * + * @job: scheduler job to clean up + */ +void drm_sched_job_cleanup(struct drm_sched_job *job) +{ + dma_fence_put(&job->s_fence->finished); + job->s_fence = NULL; +} +EXPORT_SYMBOL(drm_sched_job_cleanup); + /** * drm_sched_ready - is the scheduler ready * diff --git a/drivers/gpu/drm/v3d/v3d_sched.c b/drivers/gpu/drm/v3d/v3d_sched.c index 80b641ffc3be..445b2ef03303 100644 --- a/drivers/gpu/drm/v3d/v3d_sched.c +++ b/drivers/gpu/drm/v3d/v3d_sched.c @@ -35,6 +35,8 @@ v3d_job_free(struct drm_sched_job *sched_job) { struct v3d_job *job = to_v3d_job(sched_job); + drm_sched_job_cleanup(sched_job); + v3d_exec_put(job->exec); } diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index 4ae192a21c3f..926379d53484 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -296,6 +296,7 @@ void drm_sched_fini(struct drm_gpu_scheduler *sched); int drm_sched_job_init(struct drm_sched_job *job, struct drm_sched_entity *entity, void *owner); +void drm_sched_job_cleanup(struct drm_sched_job *job); void drm_sched_wakeup(struct drm_gpu_scheduler *sched); void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, struct drm_sched_job *job); -- cgit v1.2.3 From aba118389a6fb2ad7958de0f37b5869852bd38cf Mon Sep 17 00:00:00 2001 From: "Dmitry V. Levin" Date: Thu, 1 Nov 2018 14:03:08 +0300 Subject: uapi: fix linux/kfd_ioctl.h userspace compilation errors Consistently use types provided by via to fix the following linux/kfd_ioctl.h userspace compilation errors: /usr/include/linux/kfd_ioctl.h:250:2: error: unknown type name 'uint32_t' uint32_t reset_type; /usr/include/linux/kfd_ioctl.h:251:2: error: unknown type name 'uint32_t' uint32_t reset_cause; /usr/include/linux/kfd_ioctl.h:252:2: error: unknown type name 'uint32_t' uint32_t memory_lost; /usr/include/linux/kfd_ioctl.h:253:2: error: unknown type name 'uint32_t' uint32_t gpu_id; Fixes: 0c119abad7f0d ("drm/amd: Add kfd ioctl defines for hw_exception event") Cc: # v4.19 Signed-off-by: Dmitry V. Levin Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- include/uapi/linux/kfd_ioctl.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h index f5ff8a76e208..dae897f38e59 100644 --- a/include/uapi/linux/kfd_ioctl.h +++ b/include/uapi/linux/kfd_ioctl.h @@ -255,10 +255,10 @@ struct kfd_hsa_memory_exception_data { /* hw exception data */ struct kfd_hsa_hw_exception_data { - uint32_t reset_type; - uint32_t reset_cause; - uint32_t memory_lost; - uint32_t gpu_id; + __u32 reset_type; + __u32 reset_cause; + __u32 memory_lost; + __u32 gpu_id; }; /* Event data */ -- cgit v1.2.3 From 8e7f91719db36440d63de37331367be9700ca0c7 Mon Sep 17 00:00:00 2001 From: "Dmitry V. Levin" Date: Thu, 1 Nov 2018 14:03:28 +0300 Subject: uapi: fix more linux/kfd_ioctl.h userspace compilation errors Consistently use types provided by via to fix struct kfd_ioctl_get_queue_wave_state_args userspace compilation errors. Fixes: 5df099e8bc83f ("drm/amdkfd: Add wavefront context save state retrieval ioctl") Signed-off-by: Dmitry V. Levin Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- include/uapi/linux/kfd_ioctl.h | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h index dae897f38e59..b01eb502d49c 100644 --- a/include/uapi/linux/kfd_ioctl.h +++ b/include/uapi/linux/kfd_ioctl.h @@ -83,11 +83,11 @@ struct kfd_ioctl_set_cu_mask_args { }; struct kfd_ioctl_get_queue_wave_state_args { - uint64_t ctl_stack_address; /* to KFD */ - uint32_t ctl_stack_used_size; /* from KFD */ - uint32_t save_area_used_size; /* from KFD */ - uint32_t queue_id; /* to KFD */ - uint32_t pad; + __u64 ctl_stack_address; /* to KFD */ + __u32 ctl_stack_used_size; /* from KFD */ + __u32 save_area_used_size; /* from KFD */ + __u32 queue_id; /* to KFD */ + __u32 pad; }; /* For kfd_ioctl_set_memory_policy_args.default_policy and alternate_policy */ -- cgit v1.2.3 From 9349e23907be1954ccdf6d771d640e2788da1643 Mon Sep 17 00:00:00 2001 From: "Dmitry V. Levin" Date: Thu, 1 Nov 2018 14:03:08 +0300 Subject: uapi: fix linux/kfd_ioctl.h userspace compilation errors Consistently use types provided by via to fix the following linux/kfd_ioctl.h userspace compilation errors: /usr/include/linux/kfd_ioctl.h:250:2: error: unknown type name 'uint32_t' uint32_t reset_type; /usr/include/linux/kfd_ioctl.h:251:2: error: unknown type name 'uint32_t' uint32_t reset_cause; /usr/include/linux/kfd_ioctl.h:252:2: error: unknown type name 'uint32_t' uint32_t memory_lost; /usr/include/linux/kfd_ioctl.h:253:2: error: unknown type name 'uint32_t' uint32_t gpu_id; Fixes: 0c119abad7f0d ("drm/amd: Add kfd ioctl defines for hw_exception event") Cc: # v4.19 Signed-off-by: Dmitry V. Levin Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- include/uapi/linux/kfd_ioctl.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h index f5ff8a76e208..dae897f38e59 100644 --- a/include/uapi/linux/kfd_ioctl.h +++ b/include/uapi/linux/kfd_ioctl.h @@ -255,10 +255,10 @@ struct kfd_hsa_memory_exception_data { /* hw exception data */ struct kfd_hsa_hw_exception_data { - uint32_t reset_type; - uint32_t reset_cause; - uint32_t memory_lost; - uint32_t gpu_id; + __u32 reset_type; + __u32 reset_cause; + __u32 memory_lost; + __u32 gpu_id; }; /* Event data */ -- cgit v1.2.3 From 788012e33cb4ba7eedae6b621f0fa27f9196f2d0 Mon Sep 17 00:00:00 2001 From: "Dmitry V. Levin" Date: Thu, 1 Nov 2018 14:03:28 +0300 Subject: uapi: fix more linux/kfd_ioctl.h userspace compilation errors Consistently use types provided by via to fix struct kfd_ioctl_get_queue_wave_state_args userspace compilation errors. Fixes: 5df099e8bc83f ("drm/amdkfd: Add wavefront context save state retrieval ioctl") Signed-off-by: Dmitry V. Levin Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- include/uapi/linux/kfd_ioctl.h | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h index dae897f38e59..b01eb502d49c 100644 --- a/include/uapi/linux/kfd_ioctl.h +++ b/include/uapi/linux/kfd_ioctl.h @@ -83,11 +83,11 @@ struct kfd_ioctl_set_cu_mask_args { }; struct kfd_ioctl_get_queue_wave_state_args { - uint64_t ctl_stack_address; /* to KFD */ - uint32_t ctl_stack_used_size; /* from KFD */ - uint32_t save_area_used_size; /* from KFD */ - uint32_t queue_id; /* to KFD */ - uint32_t pad; + __u64 ctl_stack_address; /* to KFD */ + __u32 ctl_stack_used_size; /* from KFD */ + __u32 save_area_used_size; /* from KFD */ + __u32 queue_id; /* to KFD */ + __u32 pad; }; /* For kfd_ioctl_set_memory_policy_args.default_policy and alternate_policy */ -- cgit v1.2.3 From 4c0608f4a0e76dfb82d3accd20081f4bf47ed143 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 30 Oct 2018 09:45:55 -0400 Subject: XArray: Regularise xa_reserve The xa_reserve() function was a little unusual in that it attempted to be callable for all kinds of locking scenarios. Make it look like the other APIs with __xa_reserve, xa_reserve_bh and xa_reserve_irq variants. Signed-off-by: Matthew Wilcox --- Documentation/core-api/xarray.rst | 13 +++++++ include/linux/xarray.h | 80 ++++++++++++++++++++++++++++++++++++++- lib/test_xarray.c | 6 +++ lib/xarray.c | 18 ++++----- 4 files changed, 105 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/Documentation/core-api/xarray.rst b/Documentation/core-api/xarray.rst index a4e705108f42..65c77a81b689 100644 --- a/Documentation/core-api/xarray.rst +++ b/Documentation/core-api/xarray.rst @@ -105,6 +105,15 @@ may result in the entry being marked at some, but not all of the other indices. Storing into one index may result in the entry retrieved by some, but not all of the other indices changing. +Sometimes you need to ensure that a subsequent call to :c:func:`xa_store` +will not need to allocate memory. The :c:func:`xa_reserve` function +will store a reserved entry at the indicated index. Users of the normal +API will see this entry as containing ``NULL``. If you do not need to +use the reserved entry, you can call :c:func:`xa_release` to remove the +unused entry. If another user has stored to the entry in the meantime, +:c:func:`xa_release` will do nothing; if instead you want the entry to +become ``NULL``, you should use :c:func:`xa_erase`. + Finally, you can remove all entries from an XArray by calling :c:func:`xa_destroy`. If the XArray entries are pointers, you may wish to free the entries first. You can do this by iterating over all present @@ -167,6 +176,9 @@ Takes xa_lock internally: * :c:func:`xa_alloc` * :c:func:`xa_alloc_bh` * :c:func:`xa_alloc_irq` + * :c:func:`xa_reserve` + * :c:func:`xa_reserve_bh` + * :c:func:`xa_reserve_irq` * :c:func:`xa_destroy` * :c:func:`xa_set_mark` * :c:func:`xa_clear_mark` @@ -177,6 +189,7 @@ Assumes xa_lock held on entry: * :c:func:`__xa_erase` * :c:func:`__xa_cmpxchg` * :c:func:`__xa_alloc` + * :c:func:`__xa_reserve` * :c:func:`__xa_set_mark` * :c:func:`__xa_clear_mark` diff --git a/include/linux/xarray.h b/include/linux/xarray.h index d9514928ddac..c2cb0426c60c 100644 --- a/include/linux/xarray.h +++ b/include/linux/xarray.h @@ -291,7 +291,6 @@ void *xa_load(struct xarray *, unsigned long index); void *xa_store(struct xarray *, unsigned long index, void *entry, gfp_t); void *xa_cmpxchg(struct xarray *, unsigned long index, void *old, void *entry, gfp_t); -int xa_reserve(struct xarray *, unsigned long index, gfp_t); void *xa_store_range(struct xarray *, unsigned long first, unsigned long last, void *entry, gfp_t); bool xa_get_mark(struct xarray *, unsigned long index, xa_mark_t); @@ -455,6 +454,7 @@ void *__xa_store(struct xarray *, unsigned long index, void *entry, gfp_t); void *__xa_cmpxchg(struct xarray *, unsigned long index, void *old, void *entry, gfp_t); int __xa_alloc(struct xarray *, u32 *id, u32 max, void *entry, gfp_t); +int __xa_reserve(struct xarray *, unsigned long index, gfp_t); void __xa_set_mark(struct xarray *, unsigned long index, xa_mark_t); void __xa_clear_mark(struct xarray *, unsigned long index, xa_mark_t); @@ -621,6 +621,84 @@ static inline int xa_alloc_irq(struct xarray *xa, u32 *id, u32 max, void *entry, return err; } +/** + * xa_reserve() - Reserve this index in the XArray. + * @xa: XArray. + * @index: Index into array. + * @gfp: Memory allocation flags. + * + * Ensures there is somewhere to store an entry at @index in the array. + * If there is already something stored at @index, this function does + * nothing. If there was nothing there, the entry is marked as reserved. + * Loading from a reserved entry returns a %NULL pointer. + * + * If you do not use the entry that you have reserved, call xa_release() + * or xa_erase() to free any unnecessary memory. + * + * Context: Any context. Takes and releases the xa_lock. + * May sleep if the @gfp flags permit. + * Return: 0 if the reservation succeeded or -ENOMEM if it failed. + */ +static inline +int xa_reserve(struct xarray *xa, unsigned long index, gfp_t gfp) +{ + int ret; + + xa_lock(xa); + ret = __xa_reserve(xa, index, gfp); + xa_unlock(xa); + + return ret; +} + +/** + * xa_reserve_bh() - Reserve this index in the XArray. + * @xa: XArray. + * @index: Index into array. + * @gfp: Memory allocation flags. + * + * A softirq-disabling version of xa_reserve(). + * + * Context: Any context. Takes and releases the xa_lock while + * disabling softirqs. + * Return: 0 if the reservation succeeded or -ENOMEM if it failed. + */ +static inline +int xa_reserve_bh(struct xarray *xa, unsigned long index, gfp_t gfp) +{ + int ret; + + xa_lock_bh(xa); + ret = __xa_reserve(xa, index, gfp); + xa_unlock_bh(xa); + + return ret; +} + +/** + * xa_reserve_irq() - Reserve this index in the XArray. + * @xa: XArray. + * @index: Index into array. + * @gfp: Memory allocation flags. + * + * An interrupt-disabling version of xa_reserve(). + * + * Context: Process context. Takes and releases the xa_lock while + * disabling interrupts. + * Return: 0 if the reservation succeeded or -ENOMEM if it failed. + */ +static inline +int xa_reserve_irq(struct xarray *xa, unsigned long index, gfp_t gfp) +{ + int ret; + + xa_lock_irq(xa); + ret = __xa_reserve(xa, index, gfp); + xa_unlock_irq(xa); + + return ret; +} + /* Everything below here is the Advanced API. Proceed with caution. */ /* diff --git a/lib/test_xarray.c b/lib/test_xarray.c index 126127658b49..e5294b20b52f 100644 --- a/lib/test_xarray.c +++ b/lib/test_xarray.c @@ -373,6 +373,12 @@ static noinline void check_reserve(struct xarray *xa) xa_erase_index(xa, 12345678); XA_BUG_ON(xa, !xa_empty(xa)); + /* And so does xa_insert */ + xa_reserve(xa, 12345678, GFP_KERNEL); + XA_BUG_ON(xa, xa_insert(xa, 12345678, xa_mk_value(12345678), 0) != 0); + xa_erase_index(xa, 12345678); + XA_BUG_ON(xa, !xa_empty(xa)); + /* Can iterate through a reserved entry */ xa_store_index(xa, 5, GFP_KERNEL); xa_reserve(xa, 6, GFP_KERNEL); diff --git a/lib/xarray.c b/lib/xarray.c index e7be4e47c6a9..9cab8cfef8a8 100644 --- a/lib/xarray.c +++ b/lib/xarray.c @@ -1488,7 +1488,7 @@ void *__xa_cmpxchg(struct xarray *xa, unsigned long index, EXPORT_SYMBOL(__xa_cmpxchg); /** - * xa_reserve() - Reserve this index in the XArray. + * __xa_reserve() - Reserve this index in the XArray. * @xa: XArray. * @index: Index into array. * @gfp: Memory allocation flags. @@ -1496,33 +1496,29 @@ EXPORT_SYMBOL(__xa_cmpxchg); * Ensures there is somewhere to store an entry at @index in the array. * If there is already something stored at @index, this function does * nothing. If there was nothing there, the entry is marked as reserved. - * Loads from @index will continue to see a %NULL pointer until a - * subsequent store to @index. + * Loading from a reserved entry returns a %NULL pointer. * * If you do not use the entry that you have reserved, call xa_release() * or xa_erase() to free any unnecessary memory. * - * Context: Process context. Takes and releases the xa_lock, IRQ or BH safe - * if specified in XArray flags. May sleep if the @gfp flags permit. + * Context: Any context. Expects the xa_lock to be held on entry. May + * release the lock, sleep and reacquire the lock if the @gfp flags permit. * Return: 0 if the reservation succeeded or -ENOMEM if it failed. */ -int xa_reserve(struct xarray *xa, unsigned long index, gfp_t gfp) +int __xa_reserve(struct xarray *xa, unsigned long index, gfp_t gfp) { XA_STATE(xas, xa, index); - unsigned int lock_type = xa_lock_type(xa); void *curr; do { - xas_lock_type(&xas, lock_type); curr = xas_load(&xas); if (!curr) xas_store(&xas, XA_ZERO_ENTRY); - xas_unlock_type(&xas, lock_type); - } while (xas_nomem(&xas, gfp)); + } while (__xas_nomem(&xas, gfp)); return xas_error(&xas); } -EXPORT_SYMBOL(xa_reserve); +EXPORT_SYMBOL(__xa_reserve); #ifdef CONFIG_XARRAY_MULTI static void xas_set_range(struct xa_state *xas, unsigned long first, -- cgit v1.2.3 From c5beb07e7a06b24f4f27304f6282b5dbd929543b Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 31 Oct 2018 14:39:28 -0400 Subject: XArray: Unify xa_cmpxchg and __xa_cmpxchg xa_cmpxchg() was one of the largest functions in the xarray implementation. By turning it into a wrapper and having the callers take the lock (like several other functions), we save 160 bytes on a tinyconfig build and reduce the duplication in xarray.c. Signed-off-by: Matthew Wilcox --- include/linux/xarray.h | 113 ++++++++++++++++++++++++++++++------------------- lib/xarray.c | 41 ------------------ 2 files changed, 69 insertions(+), 85 deletions(-) (limited to 'include') diff --git a/include/linux/xarray.h b/include/linux/xarray.h index c2cb0426c60c..8e59d4fbd55e 100644 --- a/include/linux/xarray.h +++ b/include/linux/xarray.h @@ -289,8 +289,6 @@ struct xarray { void xa_init_flags(struct xarray *, gfp_t flags); void *xa_load(struct xarray *, unsigned long index); void *xa_store(struct xarray *, unsigned long index, void *entry, gfp_t); -void *xa_cmpxchg(struct xarray *, unsigned long index, - void *old, void *entry, gfp_t); void *xa_store_range(struct xarray *, unsigned long first, unsigned long last, void *entry, gfp_t); bool xa_get_mark(struct xarray *, unsigned long index, xa_mark_t); @@ -359,48 +357,6 @@ static inline void *xa_erase(struct xarray *xa, unsigned long index) return xa_store(xa, index, NULL, 0); } -/** - * xa_insert() - Store this entry in the XArray unless another entry is - * already present. - * @xa: XArray. - * @index: Index into array. - * @entry: New entry. - * @gfp: Memory allocation flags. - * - * If you would rather see the existing entry in the array, use xa_cmpxchg(). - * This function is for users who don't care what the entry is, only that - * one is present. - * - * Context: Process context. Takes and releases the xa_lock. - * May sleep if the @gfp flags permit. - * Return: 0 if the store succeeded. -EEXIST if another entry was present. - * -ENOMEM if memory could not be allocated. - */ -static inline int xa_insert(struct xarray *xa, unsigned long index, - void *entry, gfp_t gfp) -{ - void *curr = xa_cmpxchg(xa, index, NULL, entry, gfp); - if (!curr) - return 0; - if (xa_is_err(curr)) - return xa_err(curr); - return -EEXIST; -} - -/** - * xa_release() - Release a reserved entry. - * @xa: XArray. - * @index: Index of entry. - * - * After calling xa_reserve(), you can call this function to release the - * reservation. If the entry at @index has been stored to, this function - * will do nothing. - */ -static inline void xa_release(struct xarray *xa, unsigned long index) -{ - xa_cmpxchg(xa, index, NULL, NULL, 0); -} - /** * xa_for_each() - Iterate over a portion of an XArray. * @xa: XArray. @@ -534,6 +490,61 @@ static inline void *xa_erase_irq(struct xarray *xa, unsigned long index) return entry; } +/** + * xa_cmpxchg() - Conditionally replace an entry in the XArray. + * @xa: XArray. + * @index: Index into array. + * @old: Old value to test against. + * @entry: New value to place in array. + * @gfp: Memory allocation flags. + * + * If the entry at @index is the same as @old, replace it with @entry. + * If the return value is equal to @old, then the exchange was successful. + * + * Context: Any context. Takes and releases the xa_lock. May sleep + * if the @gfp flags permit. + * Return: The old value at this index or xa_err() if an error happened. + */ +static inline void *xa_cmpxchg(struct xarray *xa, unsigned long index, + void *old, void *entry, gfp_t gfp) +{ + void *curr; + + xa_lock(xa); + curr = __xa_cmpxchg(xa, index, old, entry, gfp); + xa_unlock(xa); + + return curr; +} + +/** + * xa_insert() - Store this entry in the XArray unless another entry is + * already present. + * @xa: XArray. + * @index: Index into array. + * @entry: New entry. + * @gfp: Memory allocation flags. + * + * If you would rather see the existing entry in the array, use xa_cmpxchg(). + * This function is for users who don't care what the entry is, only that + * one is present. + * + * Context: Process context. Takes and releases the xa_lock. + * May sleep if the @gfp flags permit. + * Return: 0 if the store succeeded. -EEXIST if another entry was present. + * -ENOMEM if memory could not be allocated. + */ +static inline int xa_insert(struct xarray *xa, unsigned long index, + void *entry, gfp_t gfp) +{ + void *curr = xa_cmpxchg(xa, index, NULL, entry, gfp); + if (!curr) + return 0; + if (xa_is_err(curr)) + return xa_err(curr); + return -EEXIST; +} + /** * xa_alloc() - Find somewhere to store this entry in the XArray. * @xa: XArray. @@ -699,6 +710,20 @@ int xa_reserve_irq(struct xarray *xa, unsigned long index, gfp_t gfp) return ret; } +/** + * xa_release() - Release a reserved entry. + * @xa: XArray. + * @index: Index of entry. + * + * After calling xa_reserve(), you can call this function to release the + * reservation. If the entry at @index has been stored to, this function + * will do nothing. + */ +static inline void xa_release(struct xarray *xa, unsigned long index) +{ + xa_cmpxchg(xa, index, NULL, NULL, 0); +} + /* Everything below here is the Advanced API. Proceed with caution. */ /* diff --git a/lib/xarray.c b/lib/xarray.c index 9cab8cfef8a8..77671d4a7910 100644 --- a/lib/xarray.c +++ b/lib/xarray.c @@ -1406,47 +1406,6 @@ void *__xa_store(struct xarray *xa, unsigned long index, void *entry, gfp_t gfp) } EXPORT_SYMBOL(__xa_store); -/** - * xa_cmpxchg() - Conditionally replace an entry in the XArray. - * @xa: XArray. - * @index: Index into array. - * @old: Old value to test against. - * @entry: New value to place in array. - * @gfp: Memory allocation flags. - * - * If the entry at @index is the same as @old, replace it with @entry. - * If the return value is equal to @old, then the exchange was successful. - * - * Context: Process context. Takes and releases the xa_lock. May sleep - * if the @gfp flags permit. - * Return: The old value at this index or xa_err() if an error happened. - */ -void *xa_cmpxchg(struct xarray *xa, unsigned long index, - void *old, void *entry, gfp_t gfp) -{ - XA_STATE(xas, xa, index); - void *curr; - - if (WARN_ON_ONCE(xa_is_internal(entry))) - return XA_ERROR(-EINVAL); - - do { - xas_lock(&xas); - curr = xas_load(&xas); - if (curr == XA_ZERO_ENTRY) - curr = NULL; - if (curr == old) { - xas_store(&xas, entry); - if (xa_track_free(xa) && entry) - xas_clear_mark(&xas, XA_FREE_MARK); - } - xas_unlock(&xas); - } while (xas_nomem(&xas, gfp)); - - return xas_result(&xas, curr); -} -EXPORT_SYMBOL(xa_cmpxchg); - /** * __xa_cmpxchg() - Store this entry in the XArray. * @xa: XArray. -- cgit v1.2.3 From 9c16bb88905456a9b1299338041f05fa7699971b Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 5 Nov 2018 15:48:49 -0500 Subject: XArray: Turn xa_erase into an exported function Make xa_erase() take the spinlock and then call __xa_erase(), but make it out of line since it's such a common function. Signed-off-by: Matthew Wilcox --- include/linux/xarray.h | 18 +----------------- lib/xarray.c | 24 ++++++++++++++++++++++++ 2 files changed, 25 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/include/linux/xarray.h b/include/linux/xarray.h index 8e59d4fbd55e..4c839c17a99b 100644 --- a/include/linux/xarray.h +++ b/include/linux/xarray.h @@ -289,6 +289,7 @@ struct xarray { void xa_init_flags(struct xarray *, gfp_t flags); void *xa_load(struct xarray *, unsigned long index); void *xa_store(struct xarray *, unsigned long index, void *entry, gfp_t); +void *xa_erase(struct xarray *, unsigned long index); void *xa_store_range(struct xarray *, unsigned long first, unsigned long last, void *entry, gfp_t); bool xa_get_mark(struct xarray *, unsigned long index, xa_mark_t); @@ -340,23 +341,6 @@ static inline bool xa_marked(const struct xarray *xa, xa_mark_t mark) return xa->xa_flags & XA_FLAGS_MARK(mark); } -/** - * xa_erase() - Erase this entry from the XArray. - * @xa: XArray. - * @index: Index of entry. - * - * This function is the equivalent of calling xa_store() with %NULL as - * the third argument. The XArray does not need to allocate memory, so - * the user does not need to provide GFP flags. - * - * Context: Process context. Takes and releases the xa_lock. - * Return: The entry which used to be at this index. - */ -static inline void *xa_erase(struct xarray *xa, unsigned long index) -{ - return xa_store(xa, index, NULL, 0); -} - /** * xa_for_each() - Iterate over a portion of an XArray. * @xa: XArray. diff --git a/lib/xarray.c b/lib/xarray.c index 77671d4a7910..b55aa8c1c20f 100644 --- a/lib/xarray.c +++ b/lib/xarray.c @@ -1336,6 +1336,30 @@ void *__xa_erase(struct xarray *xa, unsigned long index) } EXPORT_SYMBOL(__xa_erase); +/** + * xa_erase() - Erase this entry from the XArray. + * @xa: XArray. + * @index: Index of entry. + * + * This function is the equivalent of calling xa_store() with %NULL as + * the third argument. The XArray does not need to allocate memory, so + * the user does not need to provide GFP flags. + * + * Context: Any context. Takes and releases the xa_lock. + * Return: The entry which used to be at this index. + */ +void *xa_erase(struct xarray *xa, unsigned long index) +{ + void *entry; + + xa_lock(xa); + entry = __xa_erase(xa, index); + xa_unlock(xa); + + return entry; +} +EXPORT_SYMBOL(xa_erase); + /** * xa_store() - Store this entry in the XArray. * @xa: XArray. -- cgit v1.2.3 From 84e5acb76dacb8ebd648a86a53907ce0dd616534 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 26 Oct 2018 14:41:29 -0400 Subject: XArray: Add xa_store_bh() and xa_store_irq() These convenience wrappers disable interrupts while taking the spinlock. A number of drivers would otherwise have to open-code these functions. Signed-off-by: Matthew Wilcox --- Documentation/core-api/xarray.rst | 5 +++- include/linux/xarray.h | 52 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 56 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/Documentation/core-api/xarray.rst b/Documentation/core-api/xarray.rst index 65c77a81b689..8a6e2087de77 100644 --- a/Documentation/core-api/xarray.rst +++ b/Documentation/core-api/xarray.rst @@ -167,6 +167,8 @@ Takes RCU read lock: Takes xa_lock internally: * :c:func:`xa_store` + * :c:func:`xa_store_bh` + * :c:func:`xa_store_irq` * :c:func:`xa_insert` * :c:func:`xa_erase` * :c:func:`xa_erase_bh` @@ -247,7 +249,8 @@ Sharing the XArray with interrupt context is also possible, either using :c:func:`xa_lock_irqsave` in both the interrupt handler and process context, or :c:func:`xa_lock_irq` in process context and :c:func:`xa_lock` in the interrupt handler. Some of the more common patterns have helper -functions such as :c:func:`xa_erase_bh` and :c:func:`xa_erase_irq`. +functions such as :c:func:`xa_store_bh`, :c:func:`xa_store_irq`, +:c:func:`xa_erase_bh` and :c:func:`xa_erase_irq`. Sometimes you need to protect access to the XArray with a mutex because that lock sits above another mutex in the locking hierarchy. That does diff --git a/include/linux/xarray.h b/include/linux/xarray.h index 4c839c17a99b..52d9732e4ec4 100644 --- a/include/linux/xarray.h +++ b/include/linux/xarray.h @@ -426,6 +426,58 @@ static inline int __xa_insert(struct xarray *xa, unsigned long index, return -EEXIST; } +/** + * xa_store_bh() - Store this entry in the XArray. + * @xa: XArray. + * @index: Index into array. + * @entry: New entry. + * @gfp: Memory allocation flags. + * + * This function is like calling xa_store() except it disables softirqs + * while holding the array lock. + * + * Context: Any context. Takes and releases the xa_lock while + * disabling softirqs. + * Return: The entry which used to be at this index. + */ +static inline void *xa_store_bh(struct xarray *xa, unsigned long index, + void *entry, gfp_t gfp) +{ + void *curr; + + xa_lock_bh(xa); + curr = __xa_store(xa, index, entry, gfp); + xa_unlock_bh(xa); + + return curr; +} + +/** + * xa_store_irq() - Erase this entry from the XArray. + * @xa: XArray. + * @index: Index into array. + * @entry: New entry. + * @gfp: Memory allocation flags. + * + * This function is like calling xa_store() except it disables interrupts + * while holding the array lock. + * + * Context: Process context. Takes and releases the xa_lock while + * disabling interrupts. + * Return: The entry which used to be at this index. + */ +static inline void *xa_store_irq(struct xarray *xa, unsigned long index, + void *entry, gfp_t gfp) +{ + void *curr; + + xa_lock_irq(xa); + curr = __xa_store(xa, index, entry, gfp); + xa_unlock_irq(xa); + + return curr; +} + /** * xa_erase_bh() - Erase this entry from the XArray. * @xa: XArray. -- cgit v1.2.3 From 804dfaf01bcc9daa4298c608ba9018abf616ec48 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 5 Nov 2018 16:37:15 -0500 Subject: XArray: Fix Documentation Minor fixes. Signed-off-by: Matthew Wilcox --- Documentation/core-api/xarray.rst | 6 +++++- include/linux/xarray.h | 4 ++-- lib/xarray.c | 10 +++++----- 3 files changed, 12 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/Documentation/core-api/xarray.rst b/Documentation/core-api/xarray.rst index 616ac406bf86..dbe96cb5558e 100644 --- a/Documentation/core-api/xarray.rst +++ b/Documentation/core-api/xarray.rst @@ -74,7 +74,8 @@ using :c:func:`xa_load`. xa_store will overwrite any entry with the new entry and return the previous entry stored at that index. You can use :c:func:`xa_erase` instead of calling :c:func:`xa_store` with a ``NULL`` entry. There is no difference between an entry that has never -been stored to and one that has most recently had ``NULL`` stored to it. +been stored to, one that has been erased and one that has most recently +had ``NULL`` stored to it. You can conditionally replace an entry at an index by using :c:func:`xa_cmpxchg`. Like :c:func:`cmpxchg`, it will only succeed if @@ -114,6 +115,9 @@ unused entry. If another user has stored to the entry in the meantime, :c:func:`xa_release` will do nothing; if instead you want the entry to become ``NULL``, you should use :c:func:`xa_erase`. +If all entries in the array are ``NULL``, the :c:func:`xa_empty` function +will return ``true``. + Finally, you can remove all entries from an XArray by calling :c:func:`xa_destroy`. If the XArray entries are pointers, you may wish to free the entries first. You can do this by iterating over all present diff --git a/include/linux/xarray.h b/include/linux/xarray.h index 52d9732e4ec4..564892e19f8c 100644 --- a/include/linux/xarray.h +++ b/include/linux/xarray.h @@ -487,7 +487,7 @@ static inline void *xa_store_irq(struct xarray *xa, unsigned long index, * the third argument. The XArray does not need to allocate memory, so * the user does not need to provide GFP flags. * - * Context: Process context. Takes and releases the xa_lock while + * Context: Any context. Takes and releases the xa_lock while * disabling softirqs. * Return: The entry which used to be at this index. */ @@ -622,7 +622,7 @@ static inline int xa_alloc(struct xarray *xa, u32 *id, u32 max, void *entry, * Updates the @id pointer with the index, then stores the entry at that * index. A concurrent lookup will not see an uninitialised @id. * - * Context: Process context. Takes and releases the xa_lock while + * Context: Any context. Takes and releases the xa_lock while * disabling softirqs. May sleep if the @gfp flags permit. * Return: 0 on success, -ENOMEM if memory allocation fails or -ENOSPC if * there is no more space in the XArray. diff --git a/lib/xarray.c b/lib/xarray.c index c3e2084aa313..7946380cd6c9 100644 --- a/lib/xarray.c +++ b/lib/xarray.c @@ -610,8 +610,8 @@ static int xas_expand(struct xa_state *xas, void *head) * (see the xa_cmpxchg() implementation for an example). * * Return: If the slot already existed, returns the contents of this slot. - * If the slot was newly created, returns NULL. If it failed to create the - * slot, returns NULL and indicates the error in @xas. + * If the slot was newly created, returns %NULL. If it failed to create the + * slot, returns %NULL and indicates the error in @xas. */ static void *xas_create(struct xa_state *xas) { @@ -1640,7 +1640,7 @@ EXPORT_SYMBOL(__xa_alloc); * @index: Index of entry. * @mark: Mark number. * - * Attempting to set a mark on a NULL entry does not succeed. + * Attempting to set a mark on a %NULL entry does not succeed. * * Context: Any context. Expects xa_lock to be held on entry. */ @@ -1710,7 +1710,7 @@ EXPORT_SYMBOL(xa_get_mark); * @index: Index of entry. * @mark: Mark number. * - * Attempting to set a mark on a NULL entry does not succeed. + * Attempting to set a mark on a %NULL entry does not succeed. * * Context: Process context. Takes and releases the xa_lock. */ @@ -1879,7 +1879,7 @@ static unsigned int xas_extract_marked(struct xa_state *xas, void **dst, * * The @filter may be an XArray mark value, in which case entries which are * marked with that mark will be copied. It may also be %XA_PRESENT, in - * which case all entries which are not NULL will be copied. + * which case all entries which are not %NULL will be copied. * * The entries returned may not represent a snapshot of the XArray at a * moment in time. For example, if another thread stores to index 5, then -- cgit v1.2.3 From df18bfd35bbf7cb1a420b5beede1de29343793b3 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 2 Nov 2018 07:09:07 -0400 Subject: media: v4l: fix uapi mpeg slice params definition We get a headers_check warning about the newly defined ioctl command structures: ./usr/include/linux/v4l2-controls.h:1105: found __[us]{8,16,32,64} type without #include This is resolved by including linux/types.h, as suggested by the warning, but there is another problem: Three of the four structures have an odd number of __u8 headers, but are aligned to 32 bit in the v4l2_ctrl_mpeg2_slice_params, so we get an implicit padding byte for each one. To solve that, let's add explicit padding that can be set to zero and verified in the kernel. Fixes: c27bb30e7b6d ("media: v4l: Add definitions for MPEG-2 slice format and metadata") Signed-off-by: Arnd Bergmann Signed-off-by: Mauro Carvalho Chehab --- drivers/media/v4l2-core/v4l2-ctrls.c | 5 +++++ include/uapi/linux/v4l2-controls.h | 7 +++++-- 2 files changed, 10 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/media/v4l2-core/v4l2-ctrls.c b/drivers/media/v4l2-core/v4l2-ctrls.c index 6e37950292cd..5f2b033a7a42 100644 --- a/drivers/media/v4l2-core/v4l2-ctrls.c +++ b/drivers/media/v4l2-core/v4l2-ctrls.c @@ -1664,6 +1664,11 @@ static int std_validate(const struct v4l2_ctrl *ctrl, u32 idx, p_mpeg2_slice_params->forward_ref_index >= VIDEO_MAX_FRAME) return -EINVAL; + if (p_mpeg2_slice_params->pad || + p_mpeg2_slice_params->picture.pad || + p_mpeg2_slice_params->sequence.pad) + return -EINVAL; + return 0; case V4L2_CTRL_TYPE_MPEG2_QUANTIZATION: diff --git a/include/uapi/linux/v4l2-controls.h b/include/uapi/linux/v4l2-controls.h index 86a54916206f..998983a6e6b7 100644 --- a/include/uapi/linux/v4l2-controls.h +++ b/include/uapi/linux/v4l2-controls.h @@ -47,11 +47,11 @@ * videodev2.h. */ -#include - #ifndef __LINUX_V4L2_CONTROLS_H #define __LINUX_V4L2_CONTROLS_H +#include + /* Control classes */ #define V4L2_CTRL_CLASS_USER 0x00980000 /* Old-style 'user' controls */ #define V4L2_CTRL_CLASS_MPEG 0x00990000 /* MPEG-compression controls */ @@ -1112,6 +1112,7 @@ struct v4l2_mpeg2_sequence { __u8 profile_and_level_indication; __u8 progressive_sequence; __u8 chroma_format; + __u8 pad; }; struct v4l2_mpeg2_picture { @@ -1130,6 +1131,7 @@ struct v4l2_mpeg2_picture { __u8 alternate_scan; __u8 repeat_first_field; __u8 progressive_frame; + __u8 pad; }; struct v4l2_ctrl_mpeg2_slice_params { @@ -1144,6 +1146,7 @@ struct v4l2_ctrl_mpeg2_slice_params { __u8 backward_ref_index; __u8 forward_ref_index; + __u8 pad; }; struct v4l2_ctrl_mpeg2_quantization { -- cgit v1.2.3 From ef86eaf97acd6d82cd3fd40f997b1c8c4895a443 Mon Sep 17 00:00:00 2001 From: Ezequiel Garcia Date: Thu, 18 Oct 2018 14:54:29 -0400 Subject: media: Rename vb2_m2m_request_queue -> v4l2_m2m_request_queue To be consistent with the rest of the mem2mem helpers, rename vb2_m2m_request_queue to v4l2_m2m_request_queue. This is just a cosmetic change. Signed-off-by: Ezequiel Garcia Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- drivers/media/platform/vim2m.c | 2 +- drivers/media/v4l2-core/v4l2-mem2mem.c | 4 ++-- drivers/staging/media/sunxi/cedrus/cedrus.c | 2 +- include/media/v4l2-mem2mem.h | 2 +- 4 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/media/platform/vim2m.c b/drivers/media/platform/vim2m.c index af150a0395df..d82db738f174 100644 --- a/drivers/media/platform/vim2m.c +++ b/drivers/media/platform/vim2m.c @@ -1009,7 +1009,7 @@ static const struct v4l2_m2m_ops m2m_ops = { static const struct media_device_ops m2m_media_ops = { .req_validate = vb2_request_validate, - .req_queue = vb2_m2m_request_queue, + .req_queue = v4l2_m2m_request_queue, }; static int vim2m_probe(struct platform_device *pdev) diff --git a/drivers/media/v4l2-core/v4l2-mem2mem.c b/drivers/media/v4l2-core/v4l2-mem2mem.c index d7806db222d8..1ed2465972ac 100644 --- a/drivers/media/v4l2-core/v4l2-mem2mem.c +++ b/drivers/media/v4l2-core/v4l2-mem2mem.c @@ -953,7 +953,7 @@ void v4l2_m2m_buf_queue(struct v4l2_m2m_ctx *m2m_ctx, } EXPORT_SYMBOL_GPL(v4l2_m2m_buf_queue); -void vb2_m2m_request_queue(struct media_request *req) +void v4l2_m2m_request_queue(struct media_request *req) { struct media_request_object *obj, *obj_safe; struct v4l2_m2m_ctx *m2m_ctx = NULL; @@ -997,7 +997,7 @@ void vb2_m2m_request_queue(struct media_request *req) if (m2m_ctx) v4l2_m2m_try_schedule(m2m_ctx); } -EXPORT_SYMBOL_GPL(vb2_m2m_request_queue); +EXPORT_SYMBOL_GPL(v4l2_m2m_request_queue); /* Videobuf2 ioctl helpers */ diff --git a/drivers/staging/media/sunxi/cedrus/cedrus.c b/drivers/staging/media/sunxi/cedrus/cedrus.c index 82558455384a..dd121f66fa2d 100644 --- a/drivers/staging/media/sunxi/cedrus/cedrus.c +++ b/drivers/staging/media/sunxi/cedrus/cedrus.c @@ -253,7 +253,7 @@ static const struct v4l2_m2m_ops cedrus_m2m_ops = { static const struct media_device_ops cedrus_m2m_media_ops = { .req_validate = cedrus_request_validate, - .req_queue = vb2_m2m_request_queue, + .req_queue = v4l2_m2m_request_queue, }; static int cedrus_probe(struct platform_device *pdev) diff --git a/include/media/v4l2-mem2mem.h b/include/media/v4l2-mem2mem.h index 58c1ecf3d648..5467264771ec 100644 --- a/include/media/v4l2-mem2mem.h +++ b/include/media/v4l2-mem2mem.h @@ -624,7 +624,7 @@ v4l2_m2m_dst_buf_remove_by_idx(struct v4l2_m2m_ctx *m2m_ctx, unsigned int idx) /* v4l2 request helper */ -void vb2_m2m_request_queue(struct media_request *req); +void v4l2_m2m_request_queue(struct media_request *req); /* v4l2 ioctl helpers */ -- cgit v1.2.3 From aa9b760cec2385ad408bb2e346c7f6dc1be69a79 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Sun, 4 Nov 2018 11:32:47 +0100 Subject: HID: fix up .raw_event() documentation The documentation for the .raw_event() callback says that if the driver return 1, there will be no further processing of the event, but this is not true, the actual code in hid-core.c looks like this: if (hdrv && hdrv->raw_event && hid_match_report(hid, report)) { ret = hdrv->raw_event(hid, report, data, size); if (ret < 0) goto unlock; } ret = hid_report_raw_event(hid, type, data, size, interrupt); The only return value that has any effect on the processing is a negative error. Correct this as it seems to confuse people: I found bogus code in the Razer out-of-tree driver attempting to return 1 here. Signed-off-by: Linus Walleij Signed-off-by: Jiri Kosina --- include/linux/hid.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/hid.h b/include/linux/hid.h index 2827b87590d8..387c70df6f29 100644 --- a/include/linux/hid.h +++ b/include/linux/hid.h @@ -722,8 +722,8 @@ struct hid_usage_id { * input will not be passed to raw_event unless hid_device_io_start is * called. * - * raw_event and event should return 0 on no action performed, 1 when no - * further processing should be done and negative on error + * raw_event and event should return negative on error, any other value will + * pass the event on to .event() typically return 0 for success. * * input_mapping shall return a negative value to completely ignore this usage * (e.g. doubled or invalid usage), zero to continue with parsing of this -- cgit v1.2.3 From 8bd66d147c88bd441178c7b4c774ae5a185f19b8 Mon Sep 17 00:00:00 2001 From: "ndesaulniers@google.com" Date: Wed, 31 Oct 2018 12:39:01 -0700 Subject: include/linux/compiler*.h: define asm_volatile_goto asm_volatile_goto should also be defined for other compilers that support asm goto. Fixes commit 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive"). Signed-off-by: Nick Desaulniers Signed-off-by: Miguel Ojeda --- include/linux/compiler_types.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include') diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h index 3439d7d0249a..4a3f9c09c92d 100644 --- a/include/linux/compiler_types.h +++ b/include/linux/compiler_types.h @@ -130,6 +130,10 @@ struct ftrace_likely_data { # define randomized_struct_fields_end #endif +#ifndef asm_volatile_goto +#define asm_volatile_goto(x...) asm goto(x) +#endif + /* Are two types/vars the same type (ignoring qualifiers)? */ #define __same_type(a, b) __builtin_types_compatible_p(typeof(a), typeof(b)) -- cgit v1.2.3 From 98ee3fc7ef8395f8b7a379e6608aee91efc66d48 Mon Sep 17 00:00:00 2001 From: Boris Brezillon Date: Tue, 6 Nov 2018 17:25:37 +0100 Subject: mtd: nand: Fix nanddev_pos_next_page() kernel-doc header Function name is wrong in the kernel-doc header. Fixes: 9c3736a3de21 ("mtd: nand: Add core infrastructure to deal with NAND devices") Signed-off-by: Boris Brezillon Reviewed-by: Miquel Raynal --- include/linux/mtd/nand.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/mtd/nand.h b/include/linux/mtd/nand.h index 78b86dea2f29..7f53ece2c039 100644 --- a/include/linux/mtd/nand.h +++ b/include/linux/mtd/nand.h @@ -568,7 +568,7 @@ static inline void nanddev_pos_next_eraseblock(struct nand_device *nand, } /** - * nanddev_pos_next_eraseblock() - Move a position to the next page + * nanddev_pos_next_page() - Move a position to the next page * @nand: NAND device * @pos: the position to update * -- cgit v1.2.3 From 81bd415c91eb966118d773dddf254aebf3022411 Mon Sep 17 00:00:00 2001 From: Mathieu Malaterre Date: Wed, 6 Jun 2018 21:42:32 +0200 Subject: watchdog/core: Add missing prototypes for weak functions MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The split out of the hard lockup detector exposed two new weak functions, but no prototypes for them, which triggers the build warning: kernel/watchdog.c:109:12: warning: no previous prototype for ‘watchdog_nmi_enable’ [-Wmissing-prototypes] kernel/watchdog.c:115:13: warning: no previous prototype for ‘watchdog_nmi_disable’ [-Wmissing-prototypes] Add the prototypes. Fixes: 73ce0511c436 ("kernel/watchdog.c: move hardlockup detector to separate file") Signed-off-by: Mathieu Malaterre Signed-off-by: Thomas Gleixner Cc: Babu Moger Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20180606194232.17653-1-malat@debian.org --- include/linux/nmi.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/nmi.h b/include/linux/nmi.h index 08f9247e9827..9003e29cde46 100644 --- a/include/linux/nmi.h +++ b/include/linux/nmi.h @@ -119,6 +119,8 @@ static inline int hardlockup_detector_perf_init(void) { return 0; } void watchdog_nmi_stop(void); void watchdog_nmi_start(void); int watchdog_nmi_probe(void); +int watchdog_nmi_enable(unsigned int cpu); +void watchdog_nmi_disable(unsigned int cpu); /** * touch_nmi_watchdog - restart NMI watchdog timeout. -- cgit v1.2.3 From 781f0766cc41a9dd2e5d118ef4b1d5d89430257b Mon Sep 17 00:00:00 2001 From: Kai-Heng Feng Date: Fri, 19 Oct 2018 16:14:50 +0800 Subject: USB: Wait for extra delay time after USB_PORT_FEAT_RESET for quirky hub Devices connected under Terminus Technology Inc. Hub (1a40:0101) may fail to work after the system resumes from suspend: [ 206.063325] usb 3-2.4: reset full-speed USB device number 4 using xhci_hcd [ 206.143691] usb 3-2.4: device descriptor read/64, error -32 [ 206.351671] usb 3-2.4: device descriptor read/64, error -32 Info for this hub: T: Bus=03 Lev=01 Prnt=01 Port=01 Cnt=01 Dev#= 2 Spd=480 MxCh= 4 D: Ver= 2.00 Cls=09(hub ) Sub=00 Prot=01 MxPS=64 #Cfgs= 1 P: Vendor=1a40 ProdID=0101 Rev=01.11 S: Product=USB 2.0 Hub C: #Ifs= 1 Cfg#= 1 Atr=e0 MxPwr=100mA I: If#= 0 Alt= 0 #EPs= 1 Cls=09(hub ) Sub=00 Prot=00 Driver=hub Some expirements indicate that the USB devices connected to the hub are innocent, it's the hub itself is to blame. The hub needs extra delay time after it resets its port. Hence wait for extra delay, if the device is connected to this quirky hub. Signed-off-by: Kai-Heng Feng Cc: stable Acked-by: Alan Stern Signed-off-by: Greg Kroah-Hartman --- Documentation/admin-guide/kernel-parameters.txt | 2 ++ drivers/usb/core/hub.c | 14 +++++++++++--- drivers/usb/core/quirks.c | 6 ++++++ include/linux/usb/quirks.h | 3 +++ 4 files changed, 22 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 81d1d5a74728..19f4423e70d9 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -4713,6 +4713,8 @@ prevent spurious wakeup); n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a pause after every control message); + o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra + delay after resetting its port); Example: quirks=0781:5580:bk,0a5c:5834:gij usbhid.mousepoll= diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index c6077d582d29..d9bd7576786a 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -2794,6 +2794,7 @@ static int hub_port_reset(struct usb_hub *hub, int port1, int i, status; u16 portchange, portstatus; struct usb_port *port_dev = hub->ports[port1 - 1]; + int reset_recovery_time; if (!hub_is_superspeed(hub->hdev)) { if (warm) { @@ -2885,11 +2886,18 @@ static int hub_port_reset(struct usb_hub *hub, int port1, done: if (status == 0) { - /* TRSTRCY = 10 ms; plus some extra */ if (port_dev->quirks & USB_PORT_QUIRK_FAST_ENUM) usleep_range(10000, 12000); - else - msleep(10 + 40); + else { + /* TRSTRCY = 10 ms; plus some extra */ + reset_recovery_time = 10 + 40; + + /* Hub needs extra delay after resetting its port. */ + if (hub->hdev->quirks & USB_QUIRK_HUB_SLOW_RESET) + reset_recovery_time += 100; + + msleep(reset_recovery_time); + } if (udev) { struct usb_hcd *hcd = bus_to_hcd(udev->bus); diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c index 178d6c6063c0..4d7d948eae63 100644 --- a/drivers/usb/core/quirks.c +++ b/drivers/usb/core/quirks.c @@ -128,6 +128,9 @@ static int quirks_param_set(const char *val, const struct kernel_param *kp) case 'n': flags |= USB_QUIRK_DELAY_CTRL_MSG; break; + case 'o': + flags |= USB_QUIRK_HUB_SLOW_RESET; + break; /* Ignore unrecognized flag characters */ } } @@ -380,6 +383,9 @@ static const struct usb_device_id usb_quirk_list[] = { { USB_DEVICE(0x1a0a, 0x0200), .driver_info = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL }, + /* Terminus Technology Inc. Hub */ + { USB_DEVICE(0x1a40, 0x0101), .driver_info = USB_QUIRK_HUB_SLOW_RESET }, + /* Corsair K70 RGB */ { USB_DEVICE(0x1b1c, 0x1b13), .driver_info = USB_QUIRK_DELAY_INIT }, diff --git a/include/linux/usb/quirks.h b/include/linux/usb/quirks.h index b7a99ce56bc9..a1be64c9940f 100644 --- a/include/linux/usb/quirks.h +++ b/include/linux/usb/quirks.h @@ -66,4 +66,7 @@ /* Device needs a pause after every control message. */ #define USB_QUIRK_DELAY_CTRL_MSG BIT(13) +/* Hub needs extra delay after resetting its port. */ +#define USB_QUIRK_HUB_SLOW_RESET BIT(14) + #endif /* __LINUX_USB_QUIRKS_H */ -- cgit v1.2.3 From 64e3d12f769d60eaee6d2e53a9b7f0b3814f32ed Mon Sep 17 00:00:00 2001 From: Kuo-Hsin Yang Date: Tue, 6 Nov 2018 13:23:24 +0000 Subject: mm, drm/i915: mark pinned shmemfs pages as unevictable The i915 driver uses shmemfs to allocate backing storage for gem objects. These shmemfs pages can be pinned (increased ref count) by shmem_read_mapping_page_gfp(). When a lot of pages are pinned, vmscan wastes a lot of time scanning these pinned pages. In some extreme case, all pages in the inactive anon lru are pinned, and only the inactive anon lru is scanned due to inactive_ratio, the system cannot swap and invokes the oom-killer. Mark these pinned pages as unevictable to speed up vmscan. Export pagevec API check_move_unevictable_pages(). This patch was inspired by Chris Wilson's change [1]. [1]: https://patchwork.kernel.org/patch/9768741/ Cc: Chris Wilson Cc: Joonas Lahtinen Cc: Peter Zijlstra Cc: Andrew Morton Cc: Dave Hansen Signed-off-by: Kuo-Hsin Yang Acked-by: Michal Hocko # mm part Reviewed-by: Chris Wilson Acked-by: Dave Hansen Acked-by: Andrew Morton Link: https://patchwork.freedesktop.org/patch/msgid/20181106132324.17390-1-chris@chris-wilson.co.uk Signed-off-by: Chris Wilson --- Documentation/vm/unevictable-lru.rst | 6 +++++- drivers/gpu/drm/i915/i915_gem.c | 33 +++++++++++++++++++++++++++++---- include/linux/swap.h | 4 +++- mm/shmem.c | 2 +- mm/vmscan.c | 22 +++++++++++----------- 5 files changed, 49 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/Documentation/vm/unevictable-lru.rst b/Documentation/vm/unevictable-lru.rst index fdd84cb8d511..b8e29f977f2d 100644 --- a/Documentation/vm/unevictable-lru.rst +++ b/Documentation/vm/unevictable-lru.rst @@ -143,7 +143,7 @@ using a number of wrapper functions: Query the address space, and return true if it is completely unevictable. -These are currently used in two places in the kernel: +These are currently used in three places in the kernel: (1) By ramfs to mark the address spaces of its inodes when they are created, and this mark remains for the life of the inode. @@ -154,6 +154,10 @@ These are currently used in two places in the kernel: swapped out; the application must touch the pages manually if it wants to ensure they're in memory. + (3) By the i915 driver to mark pinned address space until it's unpinned. The + amount of unevictable memory marked by i915 driver is roughly the bounded + object size in debugfs/dri/0/i915_gem_objects. + Detecting Unevictable Pages --------------------------- diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 347b3836c809..5b80b0c14aed 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2382,11 +2382,23 @@ void __i915_gem_object_invalidate(struct drm_i915_gem_object *obj) invalidate_mapping_pages(mapping, 0, (loff_t)-1); } +/* + * Move pages to appropriate lru and release the pagevec, decrementing the + * ref count of those pages. + */ +static void check_release_pagevec(struct pagevec *pvec) +{ + check_move_unevictable_pages(pvec); + __pagevec_release(pvec); + cond_resched(); +} + static void i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj, struct sg_table *pages) { struct sgt_iter sgt_iter; + struct pagevec pvec; struct page *page; __i915_gem_object_release_shmem(obj, pages, true); @@ -2396,6 +2408,9 @@ i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj, if (i915_gem_object_needs_bit17_swizzle(obj)) i915_gem_object_save_bit_17_swizzle(obj, pages); + mapping_clear_unevictable(file_inode(obj->base.filp)->i_mapping); + + pagevec_init(&pvec); for_each_sgt_page(page, sgt_iter, pages) { if (obj->mm.dirty) set_page_dirty(page); @@ -2403,9 +2418,11 @@ i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj, if (obj->mm.madv == I915_MADV_WILLNEED) mark_page_accessed(page); - put_page(page); - cond_resched(); + if (!pagevec_add(&pvec, page)) + check_release_pagevec(&pvec); } + if (pagevec_count(&pvec)) + check_release_pagevec(&pvec); obj->mm.dirty = false; sg_free_table(pages); @@ -2526,6 +2543,7 @@ static int i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj) unsigned long last_pfn = 0; /* suppress gcc warning */ unsigned int max_segment = i915_sg_segment_size(); unsigned int sg_page_sizes; + struct pagevec pvec; gfp_t noreclaim; int ret; @@ -2561,6 +2579,7 @@ rebuild_st: * Fail silently without starting the shrinker */ mapping = obj->base.filp->f_mapping; + mapping_set_unevictable(mapping); noreclaim = mapping_gfp_constraint(mapping, ~__GFP_RECLAIM); noreclaim |= __GFP_NORETRY | __GFP_NOWARN; @@ -2675,8 +2694,14 @@ rebuild_st: err_sg: sg_mark_end(sg); err_pages: - for_each_sgt_page(page, sgt_iter, st) - put_page(page); + mapping_clear_unevictable(mapping); + pagevec_init(&pvec); + for_each_sgt_page(page, sgt_iter, st) { + if (!pagevec_add(&pvec, page)) + check_release_pagevec(&pvec); + } + if (pagevec_count(&pvec)) + check_release_pagevec(&pvec); sg_free_table(st); kfree(st); diff --git a/include/linux/swap.h b/include/linux/swap.h index 8e2c11e692ba..6c95df96c9aa 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -18,6 +18,8 @@ struct notifier_block; struct bio; +struct pagevec; + #define SWAP_FLAG_PREFER 0x8000 /* set if swap priority specified */ #define SWAP_FLAG_PRIO_MASK 0x7fff #define SWAP_FLAG_PRIO_SHIFT 0 @@ -373,7 +375,7 @@ static inline int node_reclaim(struct pglist_data *pgdat, gfp_t mask, #endif extern int page_evictable(struct page *page); -extern void check_move_unevictable_pages(struct page **, int nr_pages); +extern void check_move_unevictable_pages(struct pagevec *pvec); extern int kswapd_run(int nid); extern void kswapd_stop(int nid); diff --git a/mm/shmem.c b/mm/shmem.c index 446942677cd4..0c3b005a59eb 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -781,7 +781,7 @@ void shmem_unlock_mapping(struct address_space *mapping) break; index = indices[pvec.nr - 1] + 1; pagevec_remove_exceptionals(&pvec); - check_move_unevictable_pages(pvec.pages, pvec.nr); + check_move_unevictable_pages(&pvec); pagevec_release(&pvec); cond_resched(); } diff --git a/mm/vmscan.c b/mm/vmscan.c index c7ce2c161225..0dbc493026a2 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -46,6 +46,7 @@ #include #include #include +#include #include #include #include @@ -4162,17 +4163,16 @@ int page_evictable(struct page *page) return ret; } -#ifdef CONFIG_SHMEM /** - * check_move_unevictable_pages - check pages for evictability and move to appropriate zone lru list - * @pages: array of pages to check - * @nr_pages: number of pages to check + * check_move_unevictable_pages - check pages for evictability and move to + * appropriate zone lru list + * @pvec: pagevec with lru pages to check * - * Checks pages for evictability and moves them to the appropriate lru list. - * - * This function is only used for SysV IPC SHM_UNLOCK. + * Checks pages for evictability, if an evictable page is in the unevictable + * lru list, moves it to the appropriate evictable lru list. This function + * should be only used for lru pages. */ -void check_move_unevictable_pages(struct page **pages, int nr_pages) +void check_move_unevictable_pages(struct pagevec *pvec) { struct lruvec *lruvec; struct pglist_data *pgdat = NULL; @@ -4180,8 +4180,8 @@ void check_move_unevictable_pages(struct page **pages, int nr_pages) int pgrescued = 0; int i; - for (i = 0; i < nr_pages; i++) { - struct page *page = pages[i]; + for (i = 0; i < pvec->nr; i++) { + struct page *page = pvec->pages[i]; struct pglist_data *pagepgdat = page_pgdat(page); pgscanned++; @@ -4213,4 +4213,4 @@ void check_move_unevictable_pages(struct page **pages, int nr_pages) spin_unlock_irq(&pgdat->lru_lock); } } -#endif /* CONFIG_SHMEM */ +EXPORT_SYMBOL_GPL(check_move_unevictable_pages); -- cgit v1.2.3 From 24efee412c75843755da0ddd7bbc2db2ce9129f5 Mon Sep 17 00:00:00 2001 From: Miguel Ojeda Date: Tue, 6 Nov 2018 22:52:11 +0100 Subject: Compiler Attributes: improve explanation of header Explain better what "optional" attributes are, and avoid calling them so to avoid confusion. Simply retain "Optional" as a word to look for in the comments. Moreover, add a couple sentences to explain a bit more the intention and the documentation links. Signed-off-by: Miguel Ojeda --- include/linux/compiler_attributes.h | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/linux/compiler_attributes.h b/include/linux/compiler_attributes.h index 6b28c1b7310c..f8c400ba1929 100644 --- a/include/linux/compiler_attributes.h +++ b/include/linux/compiler_attributes.h @@ -4,22 +4,26 @@ /* * The attributes in this file are unconditionally defined and they directly - * map to compiler attribute(s) -- except those that are optional. + * map to compiler attribute(s), unless one of the compilers does not support + * the attribute. In that case, __has_attribute is used to check for support + * and the reason is stated in its comment ("Optional: ..."). * * Any other "attributes" (i.e. those that depend on a configuration option, * on a compiler, on an architecture, on plugins, on other attributes...) * should be defined elsewhere (e.g. compiler_types.h or compiler-*.h). + * The intention is to keep this file as simple as possible, as well as + * compiler- and version-agnostic (e.g. avoiding GCC_VERSION checks). * * This file is meant to be sorted (by actual attribute name, * not by #define identifier). Use the __attribute__((__name__)) syntax * (i.e. with underscores) to avoid future collisions with other macros. - * If an attribute is optional, state the reason in the comment. + * Provide links to the documentation of each supported compiler, if it exists. */ /* - * To check for optional attributes, we use __has_attribute, which is supported - * on gcc >= 5, clang >= 2.9 and icc >= 17. In the meantime, to support - * 4.6 <= gcc < 5, we implement __has_attribute by hand. + * __has_attribute is supported on gcc >= 5, clang >= 2.9 and icc >= 17. + * In the meantime, to support 4.6 <= gcc < 5, we implement __has_attribute + * by hand. * * sparse does not support __has_attribute (yet) and defines __GNUC_MINOR__ * depending on the compiler used to build it; however, these attributes have -- cgit v1.2.3 From 23c625ce3065e40c933a4239efb9b11f1194a343 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Thu, 8 Nov 2018 14:55:21 +0100 Subject: libceph: assume argonaut on the server side No one is running pre-argonaut. In addition one of the argonaut features (NOSRCADDR) has been required since day one (and a half, 2.6.34 vs 2.6.35) of the kernel client. Allow for the possibility of reusing these feature bits later. Signed-off-by: Ilya Dryomov Reviewed-by: Sage Weil --- fs/ceph/mds_client.c | 12 +++--------- include/linux/ceph/ceph_features.h | 8 +------- 2 files changed, 4 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 67a9aeb2f4ec..bd13a3267ae0 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -80,12 +80,8 @@ static int parse_reply_info_in(void **p, void *end, info->symlink = *p; *p += info->symlink_len; - if (features & CEPH_FEATURE_DIRLAYOUTHASH) - ceph_decode_copy_safe(p, end, &info->dir_layout, - sizeof(info->dir_layout), bad); - else - memset(&info->dir_layout, 0, sizeof(info->dir_layout)); - + ceph_decode_copy_safe(p, end, &info->dir_layout, + sizeof(info->dir_layout), bad); ceph_decode_32_safe(p, end, info->xattr_len, bad); ceph_decode_need(p, end, info->xattr_len, bad); info->xattr_data = *p; @@ -3182,10 +3178,8 @@ static void send_mds_reconnect(struct ceph_mds_client *mdsc, recon_state.pagelist = pagelist; if (session->s_con.peer_features & CEPH_FEATURE_MDSENC) recon_state.msg_version = 3; - else if (session->s_con.peer_features & CEPH_FEATURE_FLOCK) - recon_state.msg_version = 2; else - recon_state.msg_version = 1; + recon_state.msg_version = 2; err = iterate_session_caps(session, encode_caps_cb, &recon_state); if (err < 0) goto fail; diff --git a/include/linux/ceph/ceph_features.h b/include/linux/ceph/ceph_features.h index 6b92b3395fa9..65a38c4a02a1 100644 --- a/include/linux/ceph/ceph_features.h +++ b/include/linux/ceph/ceph_features.h @@ -213,12 +213,6 @@ DEFINE_CEPH_FEATURE_DEPRECATED(63, 1, RESERVED_BROKEN, LUMINOUS) // client-facin CEPH_FEATURE_NEW_OSDOPREPLY_ENCODING | \ CEPH_FEATURE_CEPHX_V2) -#define CEPH_FEATURES_REQUIRED_DEFAULT \ - (CEPH_FEATURE_NOSRCADDR | \ - CEPH_FEATURE_SUBSCRIBE2 | \ - CEPH_FEATURE_RECONNECT_SEQ | \ - CEPH_FEATURE_PGID64 | \ - CEPH_FEATURE_PGPOOL3 | \ - CEPH_FEATURE_OSDENC) +#define CEPH_FEATURES_REQUIRED_DEFAULT 0 #endif -- cgit v1.2.3 From a4310fa2f24687888ce80fdb0e88583561a23700 Mon Sep 17 00:00:00 2001 From: Marc Kleine-Budde Date: Wed, 31 Oct 2018 10:37:46 +0100 Subject: can: dev: can_get_echo_skb(): factor out non sending code to __can_get_echo_skb() This patch factors out all non sending parts of can_get_echo_skb() into a seperate function __can_get_echo_skb(), so that it can be re-used in an upcoming patch. Cc: linux-stable Signed-off-by: Marc Kleine-Budde --- drivers/net/can/dev.c | 36 +++++++++++++++++++++++++----------- include/linux/can/dev.h | 1 + 2 files changed, 26 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/drivers/net/can/dev.c b/drivers/net/can/dev.c index 49163570a63a..80530ab37b1e 100644 --- a/drivers/net/can/dev.c +++ b/drivers/net/can/dev.c @@ -477,14 +477,7 @@ void can_put_echo_skb(struct sk_buff *skb, struct net_device *dev, } EXPORT_SYMBOL_GPL(can_put_echo_skb); -/* - * Get the skb from the stack and loop it back locally - * - * The function is typically called when the TX done interrupt - * is handled in the device driver. The driver must protect - * access to priv->echo_skb, if necessary. - */ -unsigned int can_get_echo_skb(struct net_device *dev, unsigned int idx) +struct sk_buff *__can_get_echo_skb(struct net_device *dev, unsigned int idx, u8 *len_ptr) { struct can_priv *priv = netdev_priv(dev); @@ -495,13 +488,34 @@ unsigned int can_get_echo_skb(struct net_device *dev, unsigned int idx) struct can_frame *cf = (struct can_frame *)skb->data; u8 dlc = cf->can_dlc; - netif_rx(priv->echo_skb[idx]); + *len_ptr = dlc; priv->echo_skb[idx] = NULL; - return dlc; + return skb; } - return 0; + return NULL; +} + +/* + * Get the skb from the stack and loop it back locally + * + * The function is typically called when the TX done interrupt + * is handled in the device driver. The driver must protect + * access to priv->echo_skb, if necessary. + */ +unsigned int can_get_echo_skb(struct net_device *dev, unsigned int idx) +{ + struct sk_buff *skb; + u8 len; + + skb = __can_get_echo_skb(dev, idx, &len); + if (!skb) + return 0; + + netif_rx(skb); + + return len; } EXPORT_SYMBOL_GPL(can_get_echo_skb); diff --git a/include/linux/can/dev.h b/include/linux/can/dev.h index a83e1f632eb7..f01623aef2f7 100644 --- a/include/linux/can/dev.h +++ b/include/linux/can/dev.h @@ -169,6 +169,7 @@ void can_change_state(struct net_device *dev, struct can_frame *cf, void can_put_echo_skb(struct sk_buff *skb, struct net_device *dev, unsigned int idx); +struct sk_buff *__can_get_echo_skb(struct net_device *dev, unsigned int idx, u8 *len_ptr); unsigned int can_get_echo_skb(struct net_device *dev, unsigned int idx); void can_free_echo_skb(struct net_device *dev, unsigned int idx); -- cgit v1.2.3 From 55059f2b7f868cd43b3ad30e28e18347e1b46ace Mon Sep 17 00:00:00 2001 From: Oleksij Rempel Date: Tue, 18 Sep 2018 11:40:38 +0200 Subject: can: rx-offload: introduce can_rx_offload_get_echo_skb() and can_rx_offload_queue_sorted() functions Current CAN framework can't guarantee proper/chronological order of RX and TX-ECHO messages. To make this possible, drivers should use this functions instead of can_get_echo_skb(). Signed-off-by: Oleksij Rempel Cc: linux-stable Signed-off-by: Marc Kleine-Budde --- drivers/net/can/rx-offload.c | 46 ++++++++++++++++++++++++++++++++++++++++++ include/linux/can/rx-offload.h | 4 ++++ 2 files changed, 50 insertions(+) (limited to 'include') diff --git a/drivers/net/can/rx-offload.c b/drivers/net/can/rx-offload.c index c7d05027a7a0..c368686e2164 100644 --- a/drivers/net/can/rx-offload.c +++ b/drivers/net/can/rx-offload.c @@ -211,6 +211,52 @@ int can_rx_offload_irq_offload_fifo(struct can_rx_offload *offload) } EXPORT_SYMBOL_GPL(can_rx_offload_irq_offload_fifo); +int can_rx_offload_queue_sorted(struct can_rx_offload *offload, + struct sk_buff *skb, u32 timestamp) +{ + struct can_rx_offload_cb *cb; + unsigned long flags; + + if (skb_queue_len(&offload->skb_queue) > + offload->skb_queue_len_max) + return -ENOMEM; + + cb = can_rx_offload_get_cb(skb); + cb->timestamp = timestamp; + + spin_lock_irqsave(&offload->skb_queue.lock, flags); + __skb_queue_add_sort(&offload->skb_queue, skb, can_rx_offload_compare); + spin_unlock_irqrestore(&offload->skb_queue.lock, flags); + + can_rx_offload_schedule(offload); + + return 0; +} +EXPORT_SYMBOL_GPL(can_rx_offload_queue_sorted); + +unsigned int can_rx_offload_get_echo_skb(struct can_rx_offload *offload, + unsigned int idx, u32 timestamp) +{ + struct net_device *dev = offload->dev; + struct net_device_stats *stats = &dev->stats; + struct sk_buff *skb; + u8 len; + int err; + + skb = __can_get_echo_skb(dev, idx, &len); + if (!skb) + return 0; + + err = can_rx_offload_queue_sorted(offload, skb, timestamp); + if (err) { + stats->rx_errors++; + stats->tx_fifo_errors++; + } + + return len; +} +EXPORT_SYMBOL_GPL(can_rx_offload_get_echo_skb); + int can_rx_offload_irq_queue_err_skb(struct can_rx_offload *offload, struct sk_buff *skb) { if (skb_queue_len(&offload->skb_queue) > diff --git a/include/linux/can/rx-offload.h b/include/linux/can/rx-offload.h index cb31683bbe15..01a7c9e5d8d8 100644 --- a/include/linux/can/rx-offload.h +++ b/include/linux/can/rx-offload.h @@ -41,6 +41,10 @@ int can_rx_offload_add_timestamp(struct net_device *dev, struct can_rx_offload * int can_rx_offload_add_fifo(struct net_device *dev, struct can_rx_offload *offload, unsigned int weight); int can_rx_offload_irq_offload_timestamp(struct can_rx_offload *offload, u64 reg); int can_rx_offload_irq_offload_fifo(struct can_rx_offload *offload); +int can_rx_offload_queue_sorted(struct can_rx_offload *offload, + struct sk_buff *skb, u32 timestamp); +unsigned int can_rx_offload_get_echo_skb(struct can_rx_offload *offload, + unsigned int idx, u32 timestamp); int can_rx_offload_irq_queue_err_skb(struct can_rx_offload *offload, struct sk_buff *skb); void can_rx_offload_reset(struct can_rx_offload *offload); void can_rx_offload_del(struct can_rx_offload *offload); -- cgit v1.2.3 From 4530ec36bb1e0d24f41c33229694adacda3d5d89 Mon Sep 17 00:00:00 2001 From: Oleksij Rempel Date: Tue, 18 Sep 2018 11:40:40 +0200 Subject: can: rx-offload: rename can_rx_offload_irq_queue_err_skb() to can_rx_offload_queue_tail() This function has nothing todo with error. Signed-off-by: Oleksij Rempel Cc: linux-stable Signed-off-by: Marc Kleine-Budde --- drivers/net/can/flexcan.c | 4 ++-- drivers/net/can/rx-offload.c | 5 +++-- include/linux/can/rx-offload.h | 3 ++- 3 files changed, 7 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/net/can/flexcan.c b/drivers/net/can/flexcan.c index 41a175f80c4b..5923bd0ec118 100644 --- a/drivers/net/can/flexcan.c +++ b/drivers/net/can/flexcan.c @@ -610,7 +610,7 @@ static void flexcan_irq_bus_err(struct net_device *dev, u32 reg_esr) if (tx_errors) dev->stats.tx_errors++; - can_rx_offload_irq_queue_err_skb(&priv->offload, skb); + can_rx_offload_queue_tail(&priv->offload, skb); } static void flexcan_irq_state(struct net_device *dev, u32 reg_esr) @@ -650,7 +650,7 @@ static void flexcan_irq_state(struct net_device *dev, u32 reg_esr) if (unlikely(new_state == CAN_STATE_BUS_OFF)) can_bus_off(dev); - can_rx_offload_irq_queue_err_skb(&priv->offload, skb); + can_rx_offload_queue_tail(&priv->offload, skb); } static inline struct flexcan_priv *rx_offload_to_priv(struct can_rx_offload *offload) diff --git a/drivers/net/can/rx-offload.c b/drivers/net/can/rx-offload.c index c368686e2164..2ce4fa8698c7 100644 --- a/drivers/net/can/rx-offload.c +++ b/drivers/net/can/rx-offload.c @@ -257,7 +257,8 @@ unsigned int can_rx_offload_get_echo_skb(struct can_rx_offload *offload, } EXPORT_SYMBOL_GPL(can_rx_offload_get_echo_skb); -int can_rx_offload_irq_queue_err_skb(struct can_rx_offload *offload, struct sk_buff *skb) +int can_rx_offload_queue_tail(struct can_rx_offload *offload, + struct sk_buff *skb) { if (skb_queue_len(&offload->skb_queue) > offload->skb_queue_len_max) @@ -268,7 +269,7 @@ int can_rx_offload_irq_queue_err_skb(struct can_rx_offload *offload, struct sk_b return 0; } -EXPORT_SYMBOL_GPL(can_rx_offload_irq_queue_err_skb); +EXPORT_SYMBOL_GPL(can_rx_offload_queue_tail); static int can_rx_offload_init_queue(struct net_device *dev, struct can_rx_offload *offload, unsigned int weight) { diff --git a/include/linux/can/rx-offload.h b/include/linux/can/rx-offload.h index 01a7c9e5d8d8..8268811a697e 100644 --- a/include/linux/can/rx-offload.h +++ b/include/linux/can/rx-offload.h @@ -45,7 +45,8 @@ int can_rx_offload_queue_sorted(struct can_rx_offload *offload, struct sk_buff *skb, u32 timestamp); unsigned int can_rx_offload_get_echo_skb(struct can_rx_offload *offload, unsigned int idx, u32 timestamp); -int can_rx_offload_irq_queue_err_skb(struct can_rx_offload *offload, struct sk_buff *skb); +int can_rx_offload_queue_tail(struct can_rx_offload *offload, + struct sk_buff *skb); void can_rx_offload_reset(struct can_rx_offload *offload); void can_rx_offload_del(struct can_rx_offload *offload); void can_rx_offload_enable(struct can_rx_offload *offload); -- cgit v1.2.3 From 18e962ac0781bcb70d433de3b2a325ff792b4288 Mon Sep 17 00:00:00 2001 From: Omar Sandoval Date: Mon, 12 Nov 2018 00:08:46 -0800 Subject: kyber: fix wrong strlcpy() size in trace_kyber_latency() When copying to the latency type, we should be passing LATENCY_TYPE_LEN, not DOMAIN_LEN (this isn't a problem in practice because we only pass "total" or "I/O"). Fix it by changing all of the strlcpy() calls to use sizeof(). Fixes: 6c3b7af1c975 ("kyber: add tracepoints") Reported-by: Jordan Glover Tested-by: Jordan Glover Signed-off-by: Omar Sandoval Signed-off-by: Jens Axboe --- include/trace/events/kyber.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/trace/events/kyber.h b/include/trace/events/kyber.h index a9834c37ac40..c0e7d24ca256 100644 --- a/include/trace/events/kyber.h +++ b/include/trace/events/kyber.h @@ -31,8 +31,8 @@ TRACE_EVENT(kyber_latency, TP_fast_assign( __entry->dev = disk_devt(dev_to_disk(kobj_to_dev(q->kobj.parent))); - strlcpy(__entry->domain, domain, DOMAIN_LEN); - strlcpy(__entry->type, type, DOMAIN_LEN); + strlcpy(__entry->domain, domain, sizeof(__entry->domain)); + strlcpy(__entry->type, type, sizeof(__entry->type)); __entry->percentile = percentile; __entry->numerator = numerator; __entry->denominator = denominator; @@ -60,7 +60,7 @@ TRACE_EVENT(kyber_adjust, TP_fast_assign( __entry->dev = disk_devt(dev_to_disk(kobj_to_dev(q->kobj.parent))); - strlcpy(__entry->domain, domain, DOMAIN_LEN); + strlcpy(__entry->domain, domain, sizeof(__entry->domain)); __entry->depth = depth; ), @@ -82,7 +82,7 @@ TRACE_EVENT(kyber_throttled, TP_fast_assign( __entry->dev = disk_devt(dev_to_disk(kobj_to_dev(q->kobj.parent))); - strlcpy(__entry->domain, domain, DOMAIN_LEN); + strlcpy(__entry->domain, domain, sizeof(__entry->domain)); ), TP_printk("%d,%d %s", MAJOR(__entry->dev), MINOR(__entry->dev), -- cgit v1.2.3 From eff896288872d687d9662000ec9ae11b6d61766f Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Wed, 14 Nov 2018 09:55:43 -0800 Subject: efi/arm: Defer persistent reservations until after paging_init() The new memory EFI reservation feature we introduced to allow memory reservations to persist across kexec may trigger an unbounded number of calls to memblock_reserve(). The memblock subsystem can deal with this fine, but not before memblock resizing is enabled, which we can only do after paging_init(), when the memory we reallocate the array into is actually mapped. So break out the memreserve table processing into a separate routine and call it after paging_init() on arm64. On ARM, because of limited reviewing bandwidth of the maintainer, we cannot currently fix this, so instead, disable the EFI persistent memreserve entirely on ARM so we can fix it later. Tested-by: Marc Zyngier Signed-off-by: Ard Biesheuvel Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-efi@vger.kernel.org Link: http://lkml.kernel.org/r/20181114175544.12860-5-ard.biesheuvel@linaro.org Signed-off-by: Ingo Molnar --- arch/arm64/kernel/setup.c | 1 + drivers/firmware/efi/efi.c | 4 ++++ drivers/firmware/efi/libstub/arm-stub.c | 3 +++ include/linux/efi.h | 7 +++++++ 4 files changed, 15 insertions(+) (limited to 'include') diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 953e316521fc..f4fc1e0544b7 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -313,6 +313,7 @@ void __init setup_arch(char **cmdline_p) arm64_memblock_init(); paging_init(); + efi_apply_persistent_mem_reservations(); acpi_table_upgrade(); diff --git a/drivers/firmware/efi/efi.c b/drivers/firmware/efi/efi.c index 249eb70691b0..72a4da76d274 100644 --- a/drivers/firmware/efi/efi.c +++ b/drivers/firmware/efi/efi.c @@ -592,7 +592,11 @@ int __init efi_config_parse_tables(void *config_tables, int count, int sz, early_memunmap(tbl, sizeof(*tbl)); } + return 0; +} +int __init efi_apply_persistent_mem_reservations(void) +{ if (efi.mem_reserve != EFI_INVALID_TABLE_ADDR) { unsigned long prsv = efi.mem_reserve; diff --git a/drivers/firmware/efi/libstub/arm-stub.c b/drivers/firmware/efi/libstub/arm-stub.c index 30ac0c975f8a..3d36142cf812 100644 --- a/drivers/firmware/efi/libstub/arm-stub.c +++ b/drivers/firmware/efi/libstub/arm-stub.c @@ -75,6 +75,9 @@ void install_memreserve_table(efi_system_table_t *sys_table_arg) efi_guid_t memreserve_table_guid = LINUX_EFI_MEMRESERVE_TABLE_GUID; efi_status_t status; + if (IS_ENABLED(CONFIG_ARM)) + return; + status = efi_call_early(allocate_pool, EFI_LOADER_DATA, sizeof(*rsv), (void **)&rsv); if (status != EFI_SUCCESS) { diff --git a/include/linux/efi.h b/include/linux/efi.h index 845174e113ce..100ce4a4aff6 100644 --- a/include/linux/efi.h +++ b/include/linux/efi.h @@ -1167,6 +1167,8 @@ static inline bool efi_enabled(int feature) extern void efi_reboot(enum reboot_mode reboot_mode, const char *__unused); extern bool efi_is_table_address(unsigned long phys_addr); + +extern int efi_apply_persistent_mem_reservations(void); #else static inline bool efi_enabled(int feature) { @@ -1185,6 +1187,11 @@ static inline bool efi_is_table_address(unsigned long phys_addr) { return false; } + +static inline int efi_apply_persistent_mem_reservations(void) +{ + return 0; +} #endif extern int efi_status_to_err(efi_status_t status); -- cgit v1.2.3 From 7150ceaacb27f7b3bf494e72cd4be4e11612dfff Mon Sep 17 00:00:00 2001 From: David Howells Date: Mon, 12 Nov 2018 22:33:22 +0000 Subject: rxrpc: Fix life check The life-checking function, which is used by kAFS to make sure that a call is still live in the event of a pending signal, only samples the received packet serial number counter; it doesn't actually provoke a change in the counter, rather relying on the server to happen to give us a packet in the time window. Fix this by adding a function to force a ping to be transmitted. kAFS then keeps track of whether there's been a stall, and if so, uses the new function to ping the server, resetting the timeout to allow the reply to come back. If there's a stall, a ping and the call is *still* stalled in the same place after another period, then the call will be aborted. Fixes: bc5e3a546d55 ("rxrpc: Use MSG_WAITALL to tell sendmsg() to temporarily ignore signals") Fixes: f4d15fb6f99a ("rxrpc: Provide functions for allowing cleaner handling of signals") Signed-off-by: David Howells Signed-off-by: David S. Miller --- Documentation/networking/rxrpc.txt | 17 +++++++++++------ fs/afs/rxrpc.c | 11 ++++++++++- include/net/af_rxrpc.h | 3 ++- include/trace/events/rxrpc.h | 2 ++ net/rxrpc/af_rxrpc.c | 27 +++++++++++++++++++++++---- 5 files changed, 48 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/Documentation/networking/rxrpc.txt b/Documentation/networking/rxrpc.txt index 605e00cdd6be..89f1302d593a 100644 --- a/Documentation/networking/rxrpc.txt +++ b/Documentation/networking/rxrpc.txt @@ -1056,18 +1056,23 @@ The kernel interface functions are as follows: u32 rxrpc_kernel_check_life(struct socket *sock, struct rxrpc_call *call); + void rxrpc_kernel_probe_life(struct socket *sock, + struct rxrpc_call *call); - This returns a number that is updated when ACKs are received from the peer - (notably including PING RESPONSE ACKs which we can elicit by sending PING - ACKs to see if the call still exists on the server). The caller should - compare the numbers of two calls to see if the call is still alive after - waiting for a suitable interval. + The first function returns a number that is updated when ACKs are received + from the peer (notably including PING RESPONSE ACKs which we can elicit by + sending PING ACKs to see if the call still exists on the server). The + caller should compare the numbers of two calls to see if the call is still + alive after waiting for a suitable interval. This allows the caller to work out if the server is still contactable and if the call is still alive on the server whilst waiting for the server to process a client operation. - This function may transmit a PING ACK. + The second function causes a ping ACK to be transmitted to try to provoke + the peer into responding, which would then cause the value returned by the + first function to change. Note that this must be called in TASK_RUNNING + state. (*) Get reply timestamp. diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c index 59970886690f..a7b44863d502 100644 --- a/fs/afs/rxrpc.c +++ b/fs/afs/rxrpc.c @@ -576,6 +576,7 @@ static long afs_wait_for_call_to_complete(struct afs_call *call, { signed long rtt2, timeout; long ret; + bool stalled = false; u64 rtt; u32 life, last_life; @@ -609,12 +610,20 @@ static long afs_wait_for_call_to_complete(struct afs_call *call, life = rxrpc_kernel_check_life(call->net->socket, call->rxcall); if (timeout == 0 && - life == last_life && signal_pending(current)) + life == last_life && signal_pending(current)) { + if (stalled) break; + __set_current_state(TASK_RUNNING); + rxrpc_kernel_probe_life(call->net->socket, call->rxcall); + timeout = rtt2; + stalled = true; + continue; + } if (life != last_life) { timeout = rtt2; last_life = life; + stalled = false; } timeout = schedule_timeout(timeout); diff --git a/include/net/af_rxrpc.h b/include/net/af_rxrpc.h index de587948042a..1adefe42c0a6 100644 --- a/include/net/af_rxrpc.h +++ b/include/net/af_rxrpc.h @@ -77,7 +77,8 @@ int rxrpc_kernel_retry_call(struct socket *, struct rxrpc_call *, struct sockaddr_rxrpc *, struct key *); int rxrpc_kernel_check_call(struct socket *, struct rxrpc_call *, enum rxrpc_call_completion *, u32 *); -u32 rxrpc_kernel_check_life(struct socket *, struct rxrpc_call *); +u32 rxrpc_kernel_check_life(const struct socket *, const struct rxrpc_call *); +void rxrpc_kernel_probe_life(struct socket *, struct rxrpc_call *); u32 rxrpc_kernel_get_epoch(struct socket *, struct rxrpc_call *); bool rxrpc_kernel_get_reply_time(struct socket *, struct rxrpc_call *, ktime_t *); diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 573d5b901fb1..5b50fe4906d2 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -181,6 +181,7 @@ enum rxrpc_timer_trace { enum rxrpc_propose_ack_trace { rxrpc_propose_ack_client_tx_end, rxrpc_propose_ack_input_data, + rxrpc_propose_ack_ping_for_check_life, rxrpc_propose_ack_ping_for_keepalive, rxrpc_propose_ack_ping_for_lost_ack, rxrpc_propose_ack_ping_for_lost_reply, @@ -380,6 +381,7 @@ enum rxrpc_tx_point { #define rxrpc_propose_ack_traces \ EM(rxrpc_propose_ack_client_tx_end, "ClTxEnd") \ EM(rxrpc_propose_ack_input_data, "DataIn ") \ + EM(rxrpc_propose_ack_ping_for_check_life, "ChkLife") \ EM(rxrpc_propose_ack_ping_for_keepalive, "KeepAlv") \ EM(rxrpc_propose_ack_ping_for_lost_ack, "LostAck") \ EM(rxrpc_propose_ack_ping_for_lost_reply, "LostRpl") \ diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index 64362d078da8..a2522f9d71e2 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -375,16 +375,35 @@ EXPORT_SYMBOL(rxrpc_kernel_end_call); * getting ACKs from the server. Returns a number representing the life state * which can be compared to that returned by a previous call. * - * If this is a client call, ping ACKs will be sent to the server to find out - * whether it's still responsive and whether the call is still alive on the - * server. + * If the life state stalls, rxrpc_kernel_probe_life() should be called and + * then 2RTT waited. */ -u32 rxrpc_kernel_check_life(struct socket *sock, struct rxrpc_call *call) +u32 rxrpc_kernel_check_life(const struct socket *sock, + const struct rxrpc_call *call) { return call->acks_latest; } EXPORT_SYMBOL(rxrpc_kernel_check_life); +/** + * rxrpc_kernel_probe_life - Poke the peer to see if it's still alive + * @sock: The socket the call is on + * @call: The call to check + * + * In conjunction with rxrpc_kernel_check_life(), allow a kernel service to + * find out whether a call is still alive by pinging it. This should cause the + * life state to be bumped in about 2*RTT. + * + * The must be called in TASK_RUNNING state on pain of might_sleep() objecting. + */ +void rxrpc_kernel_probe_life(struct socket *sock, struct rxrpc_call *call) +{ + rxrpc_propose_ACK(call, RXRPC_ACK_PING, 0, 0, true, false, + rxrpc_propose_ack_ping_for_check_life); + rxrpc_send_ack_packet(call, true, NULL); +} +EXPORT_SYMBOL(rxrpc_kernel_probe_life); + /** * rxrpc_kernel_get_epoch - Retrieve the epoch value from a call. * @sock: The socket the call is on -- cgit v1.2.3 From fe841686470d383e33b606d0704ef4295141c582 Mon Sep 17 00:00:00 2001 From: Joonas Lahtinen Date: Fri, 16 Nov 2018 15:55:09 +0200 Subject: Revert "drm/i915/perf: add a parameter to control the size of OA buffer" Userspace portion is still missing. This reverts commit cd956bfcd0f58d20485ac0a785415f7d9327a95f. Cc: Lionel Landwerlin Cc: Matthew Auld Signed-off-by: Joonas Lahtinen Acked-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20181116135510.13807-1-joonas.lahtinen@linux.intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 - drivers/gpu/drm/i915/i915_perf.c | 99 ++++++++++++++-------------------------- drivers/gpu/drm/i915/i915_reg.h | 2 - include/uapi/drm/i915_drm.h | 7 --- 4 files changed, 33 insertions(+), 76 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index d69b71d368d3..017f851a586a 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2004,7 +2004,6 @@ struct drm_i915_private { u32 last_ctx_id; int format; int format_size; - int size_exponent; /** * Locks reads and writes to all head/tail state diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 2c2b63be7a6c..c762418d3b01 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -212,7 +212,13 @@ #include "i915_oa_icl.h" #include "intel_lrc_reg.h" -#define OA_TAKEN(tail, head) (((tail) - (head)) & (dev_priv->perf.oa.oa_buffer.vma->size - 1)) +/* HW requires this to be a power of two, between 128k and 16M, though driver + * is currently generally designed assuming the largest 16M size is used such + * that the overflow cases are unlikely in normal operation. + */ +#define OA_BUFFER_SIZE SZ_16M + +#define OA_TAKEN(tail, head) ((tail - head) & (OA_BUFFER_SIZE - 1)) /** * DOC: OA Tail Pointer Race @@ -356,7 +362,6 @@ struct perf_open_properties { int oa_format; bool oa_periodic; int oa_period_exponent; - u32 oa_buffer_size_exponent; }; static void free_oa_config(struct drm_i915_private *dev_priv, @@ -519,7 +524,7 @@ static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) * could put the tail out of bounds... */ if (hw_tail >= gtt_offset && - hw_tail < (gtt_offset + dev_priv->perf.oa.oa_buffer.vma->size)) { + hw_tail < (gtt_offset + OA_BUFFER_SIZE)) { dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = aging_tail = hw_tail; dev_priv->perf.oa.oa_buffer.aging_timestamp = now; @@ -648,7 +653,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, int report_size = dev_priv->perf.oa.oa_buffer.format_size; u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr; u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); - u32 mask = (dev_priv->perf.oa.oa_buffer.vma->size - 1); + u32 mask = (OA_BUFFER_SIZE - 1); size_t start_offset = *offset; unsigned long flags; unsigned int aged_tail_idx; @@ -688,8 +693,8 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, * only be incremented by multiples of the report size (notably also * all a power of two). */ - if (WARN_ONCE(head > dev_priv->perf.oa.oa_buffer.vma->size || head % report_size || - tail > dev_priv->perf.oa.oa_buffer.vma->size || tail % report_size, + if (WARN_ONCE(head > OA_BUFFER_SIZE || head % report_size || + tail > OA_BUFFER_SIZE || tail % report_size, "Inconsistent OA buffer pointers: head = %u, tail = %u\n", head, tail)) return -EIO; @@ -712,7 +717,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, * here would imply a driver bug that would result * in an overrun. */ - if (WARN_ON((dev_priv->perf.oa.oa_buffer.vma->size - head) < report_size)) { + if (WARN_ON((OA_BUFFER_SIZE - head) < report_size)) { DRM_ERROR("Spurious OA head ptr: non-integral report offset\n"); break; } @@ -871,6 +876,11 @@ static int gen8_oa_read(struct i915_perf_stream *stream, * automatically triggered reports in this condition and so we * have to assume that old reports are now being trampled * over. + * + * Considering how we don't currently give userspace control + * over the OA buffer size and always configure a large 16MB + * buffer, then a buffer overflow does anyway likely indicate + * that something has gone quite badly wrong. */ if (oastatus & GEN8_OASTATUS_OABUFFER_OVERFLOW) { ret = append_oa_status(stream, buf, count, offset, @@ -932,7 +942,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, int report_size = dev_priv->perf.oa.oa_buffer.format_size; u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr; u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); - u32 mask = (dev_priv->perf.oa.oa_buffer.vma->size - 1); + u32 mask = (OA_BUFFER_SIZE - 1); size_t start_offset = *offset; unsigned long flags; unsigned int aged_tail_idx; @@ -969,8 +979,8 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, * only be incremented by multiples of the report size (notably also * all a power of two). */ - if (WARN_ONCE(head > dev_priv->perf.oa.oa_buffer.vma->size || head % report_size || - tail > dev_priv->perf.oa.oa_buffer.vma->size || tail % report_size, + if (WARN_ONCE(head > OA_BUFFER_SIZE || head % report_size || + tail > OA_BUFFER_SIZE || tail % report_size, "Inconsistent OA buffer pointers: head = %u, tail = %u\n", head, tail)) return -EIO; @@ -990,7 +1000,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, * here would imply a driver bug that would result * in an overrun. */ - if (WARN_ON((dev_priv->perf.oa.oa_buffer.vma->size - head) < report_size)) { + if (WARN_ON((OA_BUFFER_SIZE - head) < report_size)) { DRM_ERROR("Spurious OA head ptr: non-integral report offset\n"); break; } @@ -1385,9 +1395,7 @@ static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv) I915_WRITE(GEN7_OABUFFER, gtt_offset); - I915_WRITE(GEN7_OASTATUS1, gtt_offset | - ((dev_priv->perf.oa.oa_buffer.size_exponent - 17) << - GEN7_OASTATUS1_BUFFER_SIZE_SHIFT)); /* tail */ + I915_WRITE(GEN7_OASTATUS1, gtt_offset | OABUFFER_SIZE_16M); /* tail */ /* Mark that we need updated tail pointers to read from... */ dev_priv->perf.oa.oa_buffer.tails[0].offset = INVALID_TAIL_PTR; @@ -1412,8 +1420,7 @@ static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv) * the assumption that new reports are being written to zeroed * memory... */ - memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, - dev_priv->perf.oa.oa_buffer.vma->size); + memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, OA_BUFFER_SIZE); /* Maybe make ->pollin per-stream state if we support multiple * concurrent streams in the future. @@ -1443,9 +1450,7 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) * bit." */ I915_WRITE(GEN8_OABUFFER, gtt_offset | - ((dev_priv->perf.oa.oa_buffer.size_exponent - 17) << - GEN8_OABUFFER_BUFFER_SIZE_SHIFT) | - GEN8_OABUFFER_MEM_SELECT_GGTT); + OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT); I915_WRITE(GEN8_OATAILPTR, gtt_offset & GEN8_OATAILPTR_MASK); /* Mark that we need updated tail pointers to read from... */ @@ -1473,8 +1478,7 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) * the assumption that new reports are being written to zeroed * memory... */ - memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, - dev_priv->perf.oa.oa_buffer.vma->size); + memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, OA_BUFFER_SIZE); /* * Maybe make ->pollin per-stream state if we support multiple @@ -1483,24 +1487,23 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) dev_priv->perf.oa.pollin = false; } -static int alloc_oa_buffer(struct drm_i915_private *dev_priv, int size_exponent) +static int alloc_oa_buffer(struct drm_i915_private *dev_priv) { struct drm_i915_gem_object *bo; struct i915_vma *vma; - size_t size = 1U << size_exponent; int ret; if (WARN_ON(dev_priv->perf.oa.oa_buffer.vma)) return -ENODEV; - if (WARN_ON(size < SZ_128K || size > SZ_16M)) - return -EINVAL; - ret = i915_mutex_lock_interruptible(&dev_priv->drm); if (ret) return ret; - bo = i915_gem_object_create(dev_priv, size); + BUILD_BUG_ON_NOT_POWER_OF_2(OA_BUFFER_SIZE); + BUILD_BUG_ON(OA_BUFFER_SIZE < SZ_128K || OA_BUFFER_SIZE > SZ_16M); + + bo = i915_gem_object_create(dev_priv, OA_BUFFER_SIZE); if (IS_ERR(bo)) { DRM_ERROR("Failed to allocate OA buffer\n"); ret = PTR_ERR(bo); @@ -1518,7 +1521,6 @@ static int alloc_oa_buffer(struct drm_i915_private *dev_priv, int size_exponent) goto err_unref; } dev_priv->perf.oa.oa_buffer.vma = vma; - dev_priv->perf.oa.oa_buffer.size_exponent = size_exponent; dev_priv->perf.oa.oa_buffer.vaddr = i915_gem_object_pin_map(bo, I915_MAP_WB); @@ -1527,10 +1529,9 @@ static int alloc_oa_buffer(struct drm_i915_private *dev_priv, int size_exponent) goto err_unpin; } - DRM_DEBUG_DRIVER("OA Buffer initialized, gtt offset = 0x%x, vaddr = %p, size = %llu\n", + DRM_DEBUG_DRIVER("OA Buffer initialized, gtt offset = 0x%x, vaddr = %p\n", i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma), - dev_priv->perf.oa.oa_buffer.vaddr, - dev_priv->perf.oa.oa_buffer.vma->size); + dev_priv->perf.oa.oa_buffer.vaddr); goto unlock; @@ -2090,7 +2091,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, intel_runtime_pm_get(dev_priv); intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); - ret = alloc_oa_buffer(dev_priv, props->oa_buffer_size_exponent); + ret = alloc_oa_buffer(dev_priv); if (ret) goto err_oa_buf_alloc; @@ -2649,26 +2650,6 @@ static u64 oa_exponent_to_ns(struct drm_i915_private *dev_priv, int exponent) 1000ULL * INTEL_INFO(dev_priv)->cs_timestamp_frequency_khz); } -static int -select_oa_buffer_exponent(struct drm_i915_private *i915, - u64 requested_size) -{ - int order; - - /* - * When no size is specified, use the largest size supported by all - * generations. - */ - if (!requested_size) - return order_base_2(SZ_16M); - - order = order_base_2(clamp_t(u64, requested_size, SZ_128K, SZ_16M)); - if (requested_size != (1UL << order)) - return -EINVAL; - - return order; -} - /** * read_properties_unlocked - validate + copy userspace stream open properties * @dev_priv: i915 device instance @@ -2796,14 +2777,6 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, props->oa_periodic = true; props->oa_period_exponent = value; break; - case DRM_I915_PERF_PROP_OA_BUFFER_SIZE: - ret = select_oa_buffer_exponent(dev_priv, value); - if (ret < 0) { - DRM_DEBUG("OA buffer size invalid %llu\n", value); - return ret; - } - props->oa_buffer_size_exponent = ret; - break; case DRM_I915_PERF_PROP_MAX: MISSING_CASE(id); return -EINVAL; @@ -2812,12 +2785,6 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, uprop += 2; } - /* If no buffer size was requested, select the default one. */ - if (!props->oa_buffer_size_exponent) { - props->oa_buffer_size_exponent = - select_oa_buffer_exponent(dev_priv, 0); - } - return 0; } diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 94ba86018a4f..edb58af1e903 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -586,14 +586,12 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg) #define GEN8_OABUFFER_UDW _MMIO(0x23b4) #define GEN8_OABUFFER _MMIO(0x2b14) #define GEN8_OABUFFER_MEM_SELECT_GGTT (1 << 0) /* 0: PPGTT, 1: GGTT */ -#define GEN8_OABUFFER_BUFFER_SIZE_SHIFT 3 #define GEN7_OASTATUS1 _MMIO(0x2364) #define GEN7_OASTATUS1_TAIL_MASK 0xffffffc0 #define GEN7_OASTATUS1_COUNTER_OVERFLOW (1 << 2) #define GEN7_OASTATUS1_OABUFFER_OVERFLOW (1 << 1) #define GEN7_OASTATUS1_REPORT_LOST (1 << 0) -#define GEN7_OASTATUS1_BUFFER_SIZE_SHIFT 3 #define GEN7_OASTATUS2 _MMIO(0x2368) #define GEN7_OASTATUS2_HEAD_MASK 0xffffffc0 diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h index e477ef8c644e..298b2e197744 100644 --- a/include/uapi/drm/i915_drm.h +++ b/include/uapi/drm/i915_drm.h @@ -1540,13 +1540,6 @@ enum drm_i915_perf_property_id { */ DRM_I915_PERF_PROP_OA_EXPONENT, - /** - * Specify a global OA buffer size to be allocated in bytes. The size - * specified must be supported by HW (currently supported sizes are - * powers of 2 ranging from 128Kb to 16Mb). - */ - DRM_I915_PERF_PROP_OA_BUFFER_SIZE, - DRM_I915_PERF_PROP_MAX /* non-ABI */ }; -- cgit v1.2.3 From 69fec325a64383667b8a35df5d48d6ce52fb2782 Mon Sep 17 00:00:00 2001 From: Xin Long Date: Sun, 18 Nov 2018 16:14:47 +0800 Subject: Revert "sctp: remove sctp_transport_pmtu_check" This reverts commit 22d7be267eaa8114dcc28d66c1c347f667d7878a. The dst's mtu in transport can be updated by a non sctp place like in xfrm where the MTU information didn't get synced between asoc, transport and dst, so it is still needed to do the pmtu check in sctp_packet_config. Acked-by: Neil Horman Signed-off-by: David S. Miller --- include/net/sctp/sctp.h | 12 ++++++++++++ net/sctp/output.c | 3 +++ 2 files changed, 15 insertions(+) (limited to 'include') diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h index 8c2caa370e0f..ab9242e51d9e 100644 --- a/include/net/sctp/sctp.h +++ b/include/net/sctp/sctp.h @@ -608,4 +608,16 @@ static inline __u32 sctp_dst_mtu(const struct dst_entry *dst) SCTP_DEFAULT_MINSEGMENT)); } +static inline bool sctp_transport_pmtu_check(struct sctp_transport *t) +{ + __u32 pmtu = sctp_dst_mtu(t->dst); + + if (t->pathmtu == pmtu) + return true; + + t->pathmtu = pmtu; + + return false; +} + #endif /* __net_sctp_h__ */ diff --git a/net/sctp/output.c b/net/sctp/output.c index 88dfa6ae1fb4..b0e74a3e77ec 100644 --- a/net/sctp/output.c +++ b/net/sctp/output.c @@ -118,6 +118,9 @@ void sctp_packet_config(struct sctp_packet *packet, __u32 vtag, sctp_transport_route(tp, NULL, sp); if (asoc->param_flags & SPP_PMTUD_ENABLE) sctp_assoc_sync_pmtu(asoc); + } else if (!sctp_transport_pmtu_check(tp)) { + if (asoc->param_flags & SPP_PMTUD_ENABLE) + sctp_assoc_sync_pmtu(asoc); } if (asoc->pmtu_pending) { -- cgit v1.2.3 From b34087157dd76e8d96e5e52808134a791ac61e57 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Wed, 21 Nov 2018 16:00:50 +0000 Subject: dma-direct: Make DIRECT_MAPPING_ERROR viable for SWIOTLB With the overflow buffer removed, we no longer have a unique address which is guaranteed not to be a valid DMA target to use as an error token. The DIRECT_MAPPING_ERROR value of 0 tries to at least represent an unlikely DMA target, but unfortunately there are already SWIOTLB users with DMA-able memory at physical address 0 which now gets falsely treated as a mapping failure and leads to all manner of misbehaviour. The best we can do to mitigate that is flip DIRECT_MAPPING_ERROR to the other commonly-used error value of all-bits-set, since the last single byte of memory is by far the least-likely-valid DMA target. Fixes: dff8d6c1ed58 ("swiotlb: remove the overflow buffer") Reported-by: John Stultz Tested-by: John Stultz Acked-by: Konrad Rzeszutek Wilk Signed-off-by: Robin Murphy Signed-off-by: Christoph Hellwig --- include/linux/dma-direct.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/dma-direct.h b/include/linux/dma-direct.h index bd73e7a91410..9e66bfe369aa 100644 --- a/include/linux/dma-direct.h +++ b/include/linux/dma-direct.h @@ -5,7 +5,7 @@ #include #include -#define DIRECT_MAPPING_ERROR 0 +#define DIRECT_MAPPING_ERROR (~(dma_addr_t)0) #ifdef CONFIG_ARCH_HAS_PHYS_TO_DMA #include -- cgit v1.2.3 From 86de5921a3d5dd246df661e09bdd0a6131b39ae3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 20 Nov 2018 05:53:59 -0800 Subject: tcp: defer SACK compression after DupThresh Jean-Louis reported a TCP regression and bisected to recent SACK compression. After a loss episode (receiver not able to keep up and dropping packets because its backlog is full), linux TCP stack is sending a single SACK (DUPACK). Sender waits a full RTO timer before recovering losses. While RFC 6675 says in section 5, "Algorithm Details", (2) If DupAcks < DupThresh but IsLost (HighACK + 1) returns true -- indicating at least three segments have arrived above the current cumulative acknowledgment point, which is taken to indicate loss -- go to step (4). ... (4) Invoke fast retransmit and enter loss recovery as follows: there are old TCP stacks not implementing this strategy, and still counting the dupacks before starting fast retransmit. While these stacks probably perform poorly when receivers implement LRO/GRO, we should be a little more gentle to them. This patch makes sure we do not enable SACK compression unless 3 dupacks have been sent since last rcv_nxt update. Ideally we should even rearm the timer to send one or two more DUPACK if no more packets are coming, but that will be work aiming for linux-4.21. Many thanks to Jean-Louis for bisecting the issue, providing packet captures and testing this patch. Fixes: 5d9f4262b7ea ("tcp: add SACK compression") Reported-by: Jean-Louis Dupond Tested-by: Jean-Louis Dupond Signed-off-by: Eric Dumazet Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- include/linux/tcp.h | 1 + net/ipv4/tcp_input.c | 14 ++++++++++++-- net/ipv4/tcp_output.c | 6 +++--- net/ipv4/tcp_timer.c | 2 +- 4 files changed, 17 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/tcp.h b/include/linux/tcp.h index 8ed77bb4ed86..a9b0280687d5 100644 --- a/include/linux/tcp.h +++ b/include/linux/tcp.h @@ -196,6 +196,7 @@ struct tcp_sock { u32 rcv_tstamp; /* timestamp of last received ACK (for keepalives) */ u32 lsndtime; /* timestamp of last sent data packet (for restart window) */ u32 last_oow_ack_time; /* timestamp of last out-of-window ACK */ + u32 compressed_ack_rcv_nxt; u32 tsoffset; /* timestamp offset */ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index e695584bb33f..1e37c1388189 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4268,7 +4268,7 @@ static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq) * If the sack array is full, forget about the last one. */ if (this_sack >= TCP_NUM_SACKS) { - if (tp->compressed_ack) + if (tp->compressed_ack > TCP_FASTRETRANS_THRESH) tcp_send_ack(sk); this_sack--; tp->rx_opt.num_sacks--; @@ -5189,7 +5189,17 @@ send_now: if (!tcp_is_sack(tp) || tp->compressed_ack >= sock_net(sk)->ipv4.sysctl_tcp_comp_sack_nr) goto send_now; - tp->compressed_ack++; + + if (tp->compressed_ack_rcv_nxt != tp->rcv_nxt) { + tp->compressed_ack_rcv_nxt = tp->rcv_nxt; + if (tp->compressed_ack > TCP_FASTRETRANS_THRESH) + NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED, + tp->compressed_ack - TCP_FASTRETRANS_THRESH); + tp->compressed_ack = 0; + } + + if (++tp->compressed_ack <= TCP_FASTRETRANS_THRESH) + goto send_now; if (hrtimer_is_queued(&tp->compressed_ack_timer)) return; diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 9c34b97d365d..3f510cad0b3e 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -180,10 +180,10 @@ static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts, { struct tcp_sock *tp = tcp_sk(sk); - if (unlikely(tp->compressed_ack)) { + if (unlikely(tp->compressed_ack > TCP_FASTRETRANS_THRESH)) { NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED, - tp->compressed_ack); - tp->compressed_ack = 0; + tp->compressed_ack - TCP_FASTRETRANS_THRESH); + tp->compressed_ack = TCP_FASTRETRANS_THRESH; if (hrtimer_try_to_cancel(&tp->compressed_ack_timer) == 1) __sock_put(sk); } diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index 676020663ce8..5f8b6d3cd855 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -740,7 +740,7 @@ static enum hrtimer_restart tcp_compressed_ack_kick(struct hrtimer *timer) bh_lock_sock(sk); if (!sock_owned_by_user(sk)) { - if (tp->compressed_ack) + if (tp->compressed_ack > TCP_FASTRETRANS_THRESH) tcp_send_ack(sk); } else { if (!test_and_set_bit(TCP_DELACK_TIMER_DEFERRED, -- cgit v1.2.3 From f1539a0c2545d7bd82e451bd1464f2a820f55de4 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Wed, 21 Nov 2018 16:27:11 +0100 Subject: Revert "HID: input: Create a utility class for counting scroll events" This reverts commit 1ff2e1a44e02d4bdbb9be67c7d9acc240a67141f. It turns out the current API is not that compatible with some Microsoft mice, so better start again from scratch. Signed-off-by: Benjamin Tissoires Acked-by: Harry Cutts Acked-by: Dmitry Torokhov Acked-by: Jiri Kosina --- drivers/hid/hid-input.c | 45 --------------------------------------------- include/linux/hid.h | 28 ---------------------------- 2 files changed, 73 deletions(-) (limited to 'include') diff --git a/drivers/hid/hid-input.c b/drivers/hid/hid-input.c index 28ee2ed88a1a..d6fab5798487 100644 --- a/drivers/hid/hid-input.c +++ b/drivers/hid/hid-input.c @@ -1841,48 +1841,3 @@ void hidinput_disconnect(struct hid_device *hid) } EXPORT_SYMBOL_GPL(hidinput_disconnect); -/** - * hid_scroll_counter_handle_scroll() - Send high- and low-resolution scroll - * events given a high-resolution wheel - * movement. - * @counter: a hid_scroll_counter struct describing the wheel. - * @hi_res_value: the movement of the wheel, in the mouse's high-resolution - * units. - * - * Given a high-resolution movement, this function converts the movement into - * microns and emits high-resolution scroll events for the input device. It also - * uses the multiplier from &struct hid_scroll_counter to emit low-resolution - * scroll events when appropriate for backwards-compatibility with userspace - * input libraries. - */ -void hid_scroll_counter_handle_scroll(struct hid_scroll_counter *counter, - int hi_res_value) -{ - int low_res_scroll_amount; - /* Some wheels will rest 7/8ths of a notch from the previous notch - * after slow movement, so we want the threshold for low-res events to - * be in the middle of the notches (e.g. after 4/8ths) as opposed to on - * the notches themselves (8/8ths). - */ - int threshold = counter->resolution_multiplier / 2; - - input_report_rel(counter->dev, REL_WHEEL_HI_RES, - hi_res_value * counter->microns_per_hi_res_unit); - - counter->remainder += hi_res_value; - if (abs(counter->remainder) >= threshold) { - /* Add (or subtract) 1 because we want to trigger when the wheel - * is half-way to the next notch (i.e. scroll 1 notch after a - * 1/2 notch movement, 2 notches after a 1 1/2 notch movement, - * etc.). - */ - low_res_scroll_amount = - counter->remainder / counter->resolution_multiplier - + (hi_res_value > 0 ? 1 : -1); - input_report_rel(counter->dev, REL_WHEEL, - low_res_scroll_amount); - counter->remainder -= - low_res_scroll_amount * counter->resolution_multiplier; - } -} -EXPORT_SYMBOL_GPL(hid_scroll_counter_handle_scroll); diff --git a/include/linux/hid.h b/include/linux/hid.h index 387c70df6f29..a355d61940f2 100644 --- a/include/linux/hid.h +++ b/include/linux/hid.h @@ -1139,34 +1139,6 @@ static inline u32 hid_report_len(struct hid_report *report) int hid_report_raw_event(struct hid_device *hid, int type, u8 *data, u32 size, int interrupt); - -/** - * struct hid_scroll_counter - Utility class for processing high-resolution - * scroll events. - * @dev: the input device for which events should be reported. - * @microns_per_hi_res_unit: the amount moved by the user's finger for each - * high-resolution unit reported by the mouse, in - * microns. - * @resolution_multiplier: the wheel's resolution in high-resolution mode as a - * multiple of its lower resolution. For example, if - * moving the wheel by one "notch" would result in a - * value of 1 in low-resolution mode but 8 in - * high-resolution, the multiplier is 8. - * @remainder: counts the number of high-resolution units moved since the last - * low-resolution event (REL_WHEEL or REL_HWHEEL) was sent. Should - * only be used by class methods. - */ -struct hid_scroll_counter { - struct input_dev *dev; - int microns_per_hi_res_unit; - int resolution_multiplier; - - int remainder; -}; - -void hid_scroll_counter_handle_scroll(struct hid_scroll_counter *counter, - int hi_res_value); - /* HID quirks API */ unsigned long hid_lookup_quirk(const struct hid_device *hdev); int hid_quirks_init(char **quirks_param, __u16 bus, int count); -- cgit v1.2.3 From ffe0e7cf290f5c9d1392134b4ef8da2a3761a4cd Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Wed, 21 Nov 2018 16:27:12 +0100 Subject: Revert "Input: Add the `REL_WHEEL_HI_RES` event code" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit aaf9978c3c0291ef3beaa97610bc9c3084656a85. Quoting Peter: There is a HID feature report called "Resolution Multiplier" Described in the "Enhanced Wheel Support in Windows" doc and the "USB HID Usage Tables" page 30. http://download.microsoft.com/download/b/d/1/bd1f7ef4-7d72-419e-bc5c-9f79ad7bb66e/wheel.docx https://www.usb.org/sites/default/files/documents/hut1_12v2.pdf This was new for Windows Vista, so we're only a decade behind here. I only accidentally found this a few days ago while debugging a stuck button on a Microsoft mouse. The docs above describe it like this: a wheel control by default sends value 1 per notch. If the resolution multiplier is active, the wheel is expected to send a value of $multiplier per notch (e.g. MS Sculpt mouse) or just send events more often, i.e. for less physical motion (e.g. MS Comfort mouse). For the latter, you need the right HW of course. The Sculpt mouse has tactile wheel clicks, so nothing really changes. The Comfort mouse has continuous motion with no tactile clicks. Similar to the free-wheeling Logitech mice but without any inertia. Note that the doc also says that Vista and onwards *always* enable this feature where available. An example HID definition looks like this: Usage Page Generic Desktop (0x01) Usage Resolution Multiplier (0x48) Logical Minimum 0 Logical Maximum 1 Physical Minimum 1 Physical Maximum 16 Report Size 2 # in bits Report Count 1 Feature (Data, Var, Abs) So the actual bits have values 0 or 1 and that reflects real values 1 or 16. We've only seen single-bits so far, so there's low-res and hi-res, but nothing in between. The multiplier is available for HID usages "Wheel" and "AC Pan" (horiz wheel). Microsoft suggests that > Vendors should ship their devices with smooth scrolling disabled and allow > Windows to enable it. This ensures that the device works like a regular HID > device on legacy operating systems that do not support smooth scrolling. (see the wheel doc linked above) The mice that we tested so far do reset on unplug. Device Support looks to be all (?) Microsoft mice but nothing else Not supported: - Logitech G500s, G303 - Roccat Kone XTD - all the cheap Lenovo, HP, Dell, Logitech USB mice that come with a workstation that I could find don't have it. - Etekcity something something - Razer Imperator Supported: - Microsoft Comfort Optical Mouse 3000 - yes, physical: 1:4 - Microsoft Sculpt Ergonomic Mouse - yes, physical: 1:12 - Microsoft Surface mouse - yes, physical: 1:4 So again, I think this is really just available on Microsoft mice, but probably all decent MS mice released over the last decade. Looking at the hardware itself: - no noticeable notches in the weel - low-res: 18 events per 360deg rotation (click angle 20 deg) - high-res: 72 events per 360deg → matches multiplier of 4 - I can feel the notches during wheel turns - low-res: 24 events per 360 deg rotation (click angle 15 deg) - horiz wheel is tilt-based, continuous output value 1 - high-res: 24 events per 360deg with value 12 → matches multiplier of 12 - horiz wheel output rate doubles/triples?, values is 3 - It's a touch strip, not a wheel so no notches - high-res: events have value 4 instead of 1 a bit strange given that it doesn't actually have notches. Ok, why is this an issue for the current API? First, because the logitech multiplier used in Harry's patches looks suspiciously like the Resolution Multiplier so I think we should assume it's the same thing. Nestor, can you shed some light on that? - `REL_WHEEL` is defined as the number of notches, emulated where needed. - `REL_WHEEL_HI_RES` is the movement of the user's finger in microns. - `WM_MOUSEWHEEL` (Windows) is is a multiple of 120, defined as "the threshold for action to be taken and one such action" https://docs.microsoft.com/en-us/windows/desktop/inputdev/wm-mousewheel If the multiplier is set to M, this means we need an accumulated value of M until we can claim there was a wheel click. So after enabling the multiplier and setting it to the maximum (like Windows): - M units are 15deg rotation → 1 unit is 2620/M micron (see below). This is the `REL_WHEEL_HI_RES` value. - wheel diameter 20mm: 15 deg rotation is 2.62mm, 2620 micron (pi * 20mm / (360deg/15deg)) - For every M units accumulated, send one `REL_WHEEL` event The problem here is that we've now hardcoded 20mm/15 deg into the kernel and we have no way of getting the size of the wheel or the click angle into the kernel. In userspace we now have to undo the kernel's calculation. If our click angle is e.g. 20 degree we have to undo the (lossy) calculation from the kernel and calculate the correct angle instead. This also means the 15 is a hardcoded option forever and cannot be changed. In hid-logitech-hidpp.c, the microns per unit is hardcoded per device. Harry, did you measure those by hand? We'd need to update the kernel for every device and there are 10 years worth of devices from MS alone. The multiplier default is 8 which is in the right ballpark, so I'm pretty sure this is the same as the Resolution Multiplier, just in HID++ lingo. And given that the 120 magic factor is what Windows uses in the end, I can't imagine Logitech rolling their own thing here. Nestor? And we're already fairly inaccurate with the microns anyway. The MX Anywhere 2S has a click angle of 20 degrees (18 stops) and a 17mm wheel, so a wheel notch is approximately 2.67mm, one event at multiplier 8 (1/8 of a notch) would be 334 micron. That's only 80% of the fallback value of 406 in the kernel. Multiplier 6 gives us 445micron (10% off). I'm assuming multiplier 7 doesn't exist because it's not a factor of 120. Summary: Best option may be to simply do what Windows is doing, all the HW manufacturers have to use that approach after all. Switch `REL_WHEEL_HI_RES` to report in fractions of 120, with 120 being one notch and divide that by the multiplier for the actual events. So e.g. the Logitech multiplier 8 would send value 15 for each event in hi-res mode. This can be converted in userspace to whatever userspace needs (combined with a hwdb there that tells you wheel size/click angle/...). Conflicts: include/uapi/linux/input-event-codes.h -> I kept the new reserved event in the code, so I had to adapt the revert slightly Signed-off-by: Benjamin Tissoires Acked-by: Harry Cutts Acked-by: Dmitry Torokhov Acked-by: Jiri Kosina --- Documentation/input/event-codes.rst | 11 +---------- include/uapi/linux/input-event-codes.h | 10 ---------- 2 files changed, 1 insertion(+), 20 deletions(-) (limited to 'include') diff --git a/Documentation/input/event-codes.rst b/Documentation/input/event-codes.rst index cef220c176a4..a8c0873beb95 100644 --- a/Documentation/input/event-codes.rst +++ b/Documentation/input/event-codes.rst @@ -190,16 +190,7 @@ A few EV_REL codes have special meanings: * REL_WHEEL, REL_HWHEEL: - These codes are used for vertical and horizontal scroll wheels, - respectively. The value is the number of "notches" moved on the wheel, the - physical size of which varies by device. For high-resolution wheels (which - report multiple events for each notch of movement, or do not have notches) - this may be an approximation based on the high-resolution scroll events. - -* REL_WHEEL_HI_RES: - - - If a vertical scroll wheel supports high-resolution scrolling, this code - will be emitted in addition to REL_WHEEL. The value is the (approximate) - distance travelled by the user's finger, in microns. + respectively. EV_ABS ------ diff --git a/include/uapi/linux/input-event-codes.h b/include/uapi/linux/input-event-codes.h index 6d180cc60a5d..3eb5a4c3d60a 100644 --- a/include/uapi/linux/input-event-codes.h +++ b/include/uapi/linux/input-event-codes.h @@ -716,7 +716,6 @@ * the situation described above. */ #define REL_RESERVED 0x0a -#define REL_WHEEL_HI_RES 0x0b #define REL_MAX 0x0f #define REL_CNT (REL_MAX+1) @@ -753,15 +752,6 @@ #define ABS_MISC 0x28 -/* - * 0x2e is reserved and should not be used in input drivers. - * It was used by HID as ABS_MISC+6 and userspace needs to detect if - * the next ABS_* event is correct or is just ABS_MISC + n. - * We define here ABS_RESERVED so userspace can rely on it and detect - * the situation described above. - */ -#define ABS_RESERVED 0x2e - #define ABS_MT_SLOT 0x2f /* MT slot being modified */ #define ABS_MT_TOUCH_MAJOR 0x30 /* Major axis of touching ellipse */ #define ABS_MT_TOUCH_MINOR 0x31 /* Minor axis (omit if circular) */ -- cgit v1.2.3 From 0211dda68a4f6531923a2f72d8e8959207f59fba Mon Sep 17 00:00:00 2001 From: Tal Gilboa Date: Wed, 21 Nov 2018 16:28:23 +0200 Subject: net/dim: Update DIM start sample after each DIM iteration On every iteration of net_dim, the algorithm may choose to check for the system state by comparing current data sample with previous data sample. After each of these comparison, regardless of the action taken, the sample used as baseline is needed to be updated. This patch fixes a bug that causes DIM to take wrong decisions, due to never updating the baseline sample for comparison between iterations. This way, DIM always compares current sample with zeros. Although this is a functional fix, it also improves and stabilizes performance as the algorithm works properly now. Performance: Tested single UDP TX stream with pktgen: samples/pktgen/pktgen_sample03_burst_single_flow.sh -i p4p2 -d 1.1.1.1 -m 24:8a:07:88:26:8b -f 3 -b 128 ConnectX-5 100GbE packet rate improved from 15-19Mpps to 19-20Mpps. Also, toggling between profiles is less frequent with the fix. Fixes: 8115b750dbcb ("net/dim: use struct net_dim_sample as arg to net_dim") Signed-off-by: Tal Gilboa Reviewed-by: Tariq Toukan Signed-off-by: David S. Miller --- include/linux/net_dim.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/net_dim.h b/include/linux/net_dim.h index c79e859408e6..fd458389f7d1 100644 --- a/include/linux/net_dim.h +++ b/include/linux/net_dim.h @@ -406,6 +406,8 @@ static inline void net_dim(struct net_dim *dim, } /* fall through */ case NET_DIM_START_MEASURE: + net_dim_sample(end_sample.event_ctr, end_sample.pkt_ctr, end_sample.byte_ctr, + &dim->start_sample); dim->state = NET_DIM_MEASURE_IN_PROGRESS; break; case NET_DIM_APPLY_NEW_PROFILE: -- cgit v1.2.3 From 5cd8d46ea1562be80063f53c7c6a5f40224de623 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Tue, 20 Nov 2018 13:00:18 -0500 Subject: packet: copy user buffers before orphan or clone tpacket_snd sends packets with user pages linked into skb frags. It notifies that pages can be reused when the skb is released by setting skb->destructor to tpacket_destruct_skb. This can cause data corruption if the skb is orphaned (e.g., on transmit through veth) or cloned (e.g., on mirror to another psock). Create a kernel-private copy of data in these cases, same as tun/tap zerocopy transmission. Reuse that infrastructure: mark the skb as SKBTX_ZEROCOPY_FRAG, which will trigger copy in skb_orphan_frags(_rx). Unlike other zerocopy packets, do not set shinfo destructor_arg to struct ubuf_info. tpacket_destruct_skb already uses that ptr to notify when the original skb is released and a timestamp is recorded. Do not change this timestamp behavior. The ubuf_info->callback is not needed anyway, as no zerocopy notification is expected. Mark destructor_arg as not-a-uarg by setting the lower bit to 1. The resulting value is not a valid ubuf_info pointer, nor a valid tpacket_snd frame address. Add skb_zcopy_.._nouarg helpers for this. The fix relies on features introduced in commit 52267790ef52 ("sock: add MSG_ZEROCOPY"), so can be backported as is only to 4.14. Tested with from `./in_netns.sh ./txring_overwrite` from http://github.com/wdebruij/kerneltools/tests Fixes: 69e3c75f4d54 ("net: TX_RING and packet mmap") Reported-by: Anand H. Krishnan Signed-off-by: Willem de Bruijn Signed-off-by: David S. Miller --- include/linux/skbuff.h | 18 +++++++++++++++++- net/packet/af_packet.c | 4 ++-- 2 files changed, 19 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 0ba687454267..0d1b2c3f127b 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1326,6 +1326,22 @@ static inline void skb_zcopy_set(struct sk_buff *skb, struct ubuf_info *uarg) } } +static inline void skb_zcopy_set_nouarg(struct sk_buff *skb, void *val) +{ + skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t) val | 0x1UL); + skb_shinfo(skb)->tx_flags |= SKBTX_ZEROCOPY_FRAG; +} + +static inline bool skb_zcopy_is_nouarg(struct sk_buff *skb) +{ + return (uintptr_t) skb_shinfo(skb)->destructor_arg & 0x1UL; +} + +static inline void *skb_zcopy_get_nouarg(struct sk_buff *skb) +{ + return (void *)((uintptr_t) skb_shinfo(skb)->destructor_arg & ~0x1UL); +} + /* Release a reference on a zerocopy structure */ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy) { @@ -1335,7 +1351,7 @@ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy) if (uarg->callback == sock_zerocopy_callback) { uarg->zerocopy = uarg->zerocopy && zerocopy; sock_zerocopy_put(uarg); - } else { + } else if (!skb_zcopy_is_nouarg(skb)) { uarg->callback(uarg, zerocopy); } diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index ec3095f13aae..a74650e98f42 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2394,7 +2394,7 @@ static void tpacket_destruct_skb(struct sk_buff *skb) void *ph; __u32 ts; - ph = skb_shinfo(skb)->destructor_arg; + ph = skb_zcopy_get_nouarg(skb); packet_dec_pending(&po->tx_ring); ts = __packet_set_timestamp(po, ph, skb); @@ -2461,7 +2461,7 @@ static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, skb->mark = po->sk.sk_mark; skb->tstamp = sockc->transmit_time; sock_tx_timestamp(&po->sk, sockc->tsflags, &skb_shinfo(skb)->tx_flags); - skb_shinfo(skb)->destructor_arg = ph.raw; + skb_zcopy_set_nouarg(skb, ph.raw); skb_reserve(skb, hlen); skb_reset_network_header(skb); -- cgit v1.2.3