From a1bf646f719f14464be93e163cfa7d003c472e6e Mon Sep 17 00:00:00 2001 From: Gabriel Fernandez Date: Fri, 6 Apr 2018 08:39:31 +0200 Subject: clk: stm32mp1: add missing tzc2 clock This patch adds tzc2 clock and rename tzc clock into tzc1 Signed-off-by: Gabriel Fernandez Signed-off-by: Stephen Boyd --- include/dt-bindings/clock/stm32mp1-clks.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/dt-bindings/clock/stm32mp1-clks.h b/include/dt-bindings/clock/stm32mp1-clks.h index 86e3ec662ef4..6c807fd2c8d5 100644 --- a/include/dt-bindings/clock/stm32mp1-clks.h +++ b/include/dt-bindings/clock/stm32mp1-clks.h @@ -76,7 +76,7 @@ #define I2C6 63 #define USART1 64 #define RTCAPB 65 -#define TZC 66 +#define TZC1 66 #define TZPC 67 #define IWDG1 68 #define BSEC 69 @@ -123,6 +123,7 @@ #define CRC1 110 #define USBH 111 #define ETHSTP 112 +#define TZC2 113 /* Kernel clocks */ #define SDMMC1_K 118 -- cgit v1.2.3 From ccf719b8845d6ca9f24a8708bb39e0df008f4e3f Mon Sep 17 00:00:00 2001 From: Gabriel Fernandez Date: Fri, 6 Apr 2018 08:39:33 +0200 Subject: clk: stm32mp1: remove ck_apb_dbg clock It's recommended to use only clk_sys_dbg clock instead to activate debug IP. Signed-off-by: Gabriel Fernandez Signed-off-by: Stephen Boyd --- drivers/clk/clk-stm32mp1.c | 4 ---- include/dt-bindings/clock/stm32mp1-clks.h | 1 - 2 files changed, 5 deletions(-) (limited to 'include') diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c index b7b53615b67c..edd3cf451401 100644 --- a/drivers/clk/clk-stm32mp1.c +++ b/drivers/clk/clk-stm32mp1.c @@ -1988,10 +1988,6 @@ static const struct clock_config stm32mp1_clock_cfg[] = { _DIV(RCC_MCO2CFGR, 4, 4, 0, NULL)), /* Debug clocks */ - FIXED_FACTOR(NO_ID, "ck_axi_div2", "ck_axi", 0, 1, 2), - - GATE(DBG, "ck_apb_dbg", "ck_axi_div2", 0, RCC_DBGCFGR, 8, 0), - GATE(CK_DBG, "ck_sys_dbg", "ck_axi", 0, RCC_DBGCFGR, 8, 0), COMPOSITE(CK_TRACE, "ck_trace", ck_trace_src, CLK_OPS_PARENT_ENABLE, diff --git a/include/dt-bindings/clock/stm32mp1-clks.h b/include/dt-bindings/clock/stm32mp1-clks.h index 6c807fd2c8d5..90ec780bfc68 100644 --- a/include/dt-bindings/clock/stm32mp1-clks.h +++ b/include/dt-bindings/clock/stm32mp1-clks.h @@ -229,7 +229,6 @@ #define CK_MCO2 212 /* TRACE & DEBUG clocks */ -#define DBG 213 #define CK_DBG 214 #define CK_TRACE 215 -- cgit v1.2.3 From a70cdb9eddcfd4ba20d69b84149b4a38648455ac Mon Sep 17 00:00:00 2001 From: Nayan Deshmukh Date: Thu, 29 Mar 2018 22:36:33 +0530 Subject: drm/scheduler: move the tracepoints file from the include directory MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Move it with the scheduler code. This is mostly a straight forward rename with no code change except for updating the TRACE_INCLUDE_PATH Signed-off-by: Nayan Deshmukh Suggested-by: Christian König Reviewed-by: Christian König Acked-by: Lucas Stach Signed-off-by: Alex Deucher --- drivers/gpu/drm/scheduler/gpu_scheduler.c | 2 +- drivers/gpu/drm/scheduler/gpu_scheduler_trace.h | 82 +++++++++++++++++++++++++ include/drm/gpu_scheduler_trace.h | 82 ------------------------- 3 files changed, 83 insertions(+), 83 deletions(-) create mode 100644 drivers/gpu/drm/scheduler/gpu_scheduler_trace.h delete mode 100644 include/drm/gpu_scheduler_trace.h (limited to 'include') diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c index 1d368bc66ac2..310275eaf128 100644 --- a/drivers/gpu/drm/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c @@ -30,7 +30,7 @@ #include #define CREATE_TRACE_POINTS -#include +#include "gpu_scheduler_trace.h" #define to_drm_sched_job(sched_job) \ container_of((sched_job), struct drm_sched_job, queue_node) diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h new file mode 100644 index 000000000000..4998ad950a48 --- /dev/null +++ b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h @@ -0,0 +1,82 @@ +/* + * Copyright 2017 Advanced Micro Devices, Inc. + * + * 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, sublicense, + * 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 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 NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) 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. + * + */ + +#if !defined(_GPU_SCHED_TRACE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _GPU_SCHED_TRACE_H_ + +#include +#include +#include + +#include + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM gpu_scheduler +#define TRACE_INCLUDE_FILE gpu_scheduler_trace + +TRACE_EVENT(drm_sched_job, + TP_PROTO(struct drm_sched_job *sched_job, struct drm_sched_entity *entity), + TP_ARGS(sched_job, entity), + TP_STRUCT__entry( + __field(struct drm_sched_entity *, entity) + __field(struct dma_fence *, fence) + __field(const char *, name) + __field(uint64_t, id) + __field(u32, job_count) + __field(int, hw_job_count) + ), + + TP_fast_assign( + __entry->entity = entity; + __entry->id = sched_job->id; + __entry->fence = &sched_job->s_fence->finished; + __entry->name = sched_job->sched->name; + __entry->job_count = spsc_queue_count(&entity->job_queue); + __entry->hw_job_count = atomic_read( + &sched_job->sched->hw_rq_count); + ), + TP_printk("entity=%p, id=%llu, fence=%p, ring=%s, job count:%u, hw job count:%d", + __entry->entity, __entry->id, + __entry->fence, __entry->name, + __entry->job_count, __entry->hw_job_count) +); + +TRACE_EVENT(drm_sched_process_job, + TP_PROTO(struct drm_sched_fence *fence), + TP_ARGS(fence), + TP_STRUCT__entry( + __field(struct dma_fence *, fence) + ), + + TP_fast_assign( + __entry->fence = &fence->finished; + ), + TP_printk("fence=%p signaled", __entry->fence) +); + +#endif + +/* This part must be outside protection */ +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH ../../drivers/gpu/drm/scheduler +#include diff --git a/include/drm/gpu_scheduler_trace.h b/include/drm/gpu_scheduler_trace.h deleted file mode 100644 index 0789e8d0a0e1..000000000000 --- a/include/drm/gpu_scheduler_trace.h +++ /dev/null @@ -1,82 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * 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, sublicense, - * 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 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 NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) 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. - * - */ - -#if !defined(_GPU_SCHED_TRACE_H) || defined(TRACE_HEADER_MULTI_READ) -#define _GPU_SCHED_TRACE_H_ - -#include -#include -#include - -#include - -#undef TRACE_SYSTEM -#define TRACE_SYSTEM gpu_scheduler -#define TRACE_INCLUDE_FILE gpu_scheduler_trace - -TRACE_EVENT(drm_sched_job, - TP_PROTO(struct drm_sched_job *sched_job, struct drm_sched_entity *entity), - TP_ARGS(sched_job, entity), - TP_STRUCT__entry( - __field(struct drm_sched_entity *, entity) - __field(struct dma_fence *, fence) - __field(const char *, name) - __field(uint64_t, id) - __field(u32, job_count) - __field(int, hw_job_count) - ), - - TP_fast_assign( - __entry->entity = entity; - __entry->id = sched_job->id; - __entry->fence = &sched_job->s_fence->finished; - __entry->name = sched_job->sched->name; - __entry->job_count = spsc_queue_count(&entity->job_queue); - __entry->hw_job_count = atomic_read( - &sched_job->sched->hw_rq_count); - ), - TP_printk("entity=%p, id=%llu, fence=%p, ring=%s, job count:%u, hw job count:%d", - __entry->entity, __entry->id, - __entry->fence, __entry->name, - __entry->job_count, __entry->hw_job_count) -); - -TRACE_EVENT(drm_sched_process_job, - TP_PROTO(struct drm_sched_fence *fence), - TP_ARGS(fence), - TP_STRUCT__entry( - __field(struct dma_fence *, fence) - ), - - TP_fast_assign( - __entry->fence = &fence->finished; - ), - TP_printk("fence=%p signaled", __entry->fence) -); - -#endif - -/* This part must be outside protection */ -#undef TRACE_INCLUDE_PATH -#define TRACE_INCLUDE_PATH . -#include -- cgit v1.2.3 From 552825b28ddac200b6080d9e79f4121b68e1517d Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Mon, 2 Apr 2018 11:20:44 +0800 Subject: drm/amdgpu: add new bo flag that indicates BOs don't need fallback (v2) user cases: 1. KFD wraps amdgpu_bo_create, they have no fallback case which is different with amdgpu_gem_object_create. since upstream branch has no amdgpu_amdkfd_gpuvm.c, which need KFD guys add this flag to __alloc_memory_of_gpu: + flags |= AMDGPU_GEM_CREATE_NO_FALLBACK; 2. UMD can specify this flag for their allocation as well if they like. v2: squash in merge conflict fix (Chunming) Signed-off-by: Chunming Zhou Acked-by: Felix Kuehling Cc: felix.kuehling@amd.com Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 3 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 ++++- include/uapi/drm/amdgpu_drm.h | 2 ++ 3 files changed, 8 insertions(+), 2 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 dc34b50e6b29..d7d7ce1507ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -386,7 +386,8 @@ retry: bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT) p->bytes_moved_vis += ctx.bytes_moved; - if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { + if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains && + !(bo->flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) { domain = bo->allowed_domains; goto retry; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 04d6830347ec..9e23d6f6f3f3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -388,6 +388,8 @@ retry: drm_gem_private_object_init(adev->ddev, &bo->gem_base, size); INIT_LIST_HEAD(&bo->shadow_list); INIT_LIST_HEAD(&bo->va); + bo->preferred_domains = preferred_domains; + bo->allowed_domains = allowed_domains; bo->flags = flags; @@ -424,7 +426,8 @@ retry: r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type, &bo->placement, page_align, &ctx, acc_size, NULL, resv, &amdgpu_ttm_bo_destroy); - if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device) { + if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device && + !(flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) { if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) { flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; goto retry; diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index c363b67f2d0a..4f5a27d64c54 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -95,6 +95,8 @@ extern "C" { #define AMDGPU_GEM_CREATE_VM_ALWAYS_VALID (1 << 6) /* Flag that BO sharing will be explicitly synchronized */ #define AMDGPU_GEM_CREATE_EXPLICIT_SYNC (1 << 7) +/* Flag that BO doesn't need fallback */ +#define AMDGPU_GEM_CREATE_NO_FALLBACK (1 << 8) struct drm_amdgpu_gem_create_in { /** the requested memory size */ -- cgit v1.2.3 From 1a61ee07211c543bf43e635fa703c162a78af0e1 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Wed, 4 Apr 2018 15:32:51 -0700 Subject: drm/sched: Extend the documentation. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit These comments answer all the questions I had for myself when implementing a driver using the GPU scheduler. Signed-off-by: Eric Anholt Reviewed-by: Christian König Signed-off-by: Alex Deucher --- include/drm/gpu_scheduler.h | 46 +++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 42 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index dfd54fb94e10..c053a32341bf 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -43,10 +43,12 @@ enum drm_sched_priority { }; /** - * A scheduler entity is a wrapper around a job queue or a group - * of other entities. Entities take turns emitting jobs from their - * job queues to corresponding hardware ring based on scheduling - * policy. + * drm_sched_entity - A wrapper around a job queue (typically attached + * to the DRM file_priv). + * + * Entities will emit jobs in order to their corresponding hardware + * ring, and the scheduler will alternate between entities based on + * scheduling policy. */ struct drm_sched_entity { struct list_head list; @@ -78,7 +80,18 @@ struct drm_sched_rq { struct drm_sched_fence { struct dma_fence scheduled; + + /* This fence is what will be signaled by the scheduler when + * the job is completed. + * + * When setting up an out fence for the job, you should use + * this, since it's available immediately upon + * drm_sched_job_init(), and the fence returned by the driver + * from run_job() won't be created until the dependencies have + * resolved. + */ struct dma_fence finished; + struct dma_fence_cb cb; struct dma_fence *parent; struct drm_gpu_scheduler *sched; @@ -88,6 +101,13 @@ struct drm_sched_fence { struct drm_sched_fence *to_drm_sched_fence(struct dma_fence *f); +/** + * drm_sched_job - A job to be run by an entity. + * + * A job is created by the driver using drm_sched_job_init(), and + * should call drm_sched_entity_push_job() once it wants the scheduler + * to schedule the job. + */ struct drm_sched_job { struct spsc_node queue_node; struct drm_gpu_scheduler *sched; @@ -112,10 +132,28 @@ static inline bool drm_sched_invalidate_job(struct drm_sched_job *s_job, * these functions should be implemented in driver side */ struct drm_sched_backend_ops { + /* Called when the scheduler is considering scheduling this + * job next, to get another struct dma_fence for this job to + * block on. Once it returns NULL, run_job() may be called. + */ struct dma_fence *(*dependency)(struct drm_sched_job *sched_job, struct drm_sched_entity *s_entity); + + /* Called to execute the job once all of the dependencies have + * been resolved. This may be called multiple times, if + * timedout_job() has happened and drm_sched_job_recovery() + * decides to try it again. + */ struct dma_fence *(*run_job)(struct drm_sched_job *sched_job); + + /* Called when a job has taken too long to execute, to trigger + * GPU recovery. + */ void (*timedout_job)(struct drm_sched_job *sched_job); + + /* Called once the job's finished fence has been signaled and + * it's time to clean it up. + */ void (*free_job)(struct drm_sched_job *sched_job); }; -- cgit v1.2.3 From b48c05ab5d32af2af4bc63851c153782d1c6ba42 Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Mon, 16 Apr 2018 07:50:09 +0200 Subject: xfrm: Fix warning in xfrm6_tunnel_net_exit. We need to make sure that all states are really deleted before we check that the state lists are empty. Otherwise we trigger a warning. Fixes: baeb0dbbb5659 ("xfrm6_tunnel: exit_net cleanup check added") Reported-and-tested-by:syzbot+777bf170a89e7b326405@syzkaller.appspotmail.com Signed-off-by: Steffen Klassert --- include/net/xfrm.h | 1 + net/ipv6/xfrm6_tunnel.c | 3 +++ net/xfrm/xfrm_state.c | 6 ++++++ 3 files changed, 10 insertions(+) (limited to 'include') diff --git a/include/net/xfrm.h b/include/net/xfrm.h index a872379b69da..45e75c36b738 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -375,6 +375,7 @@ struct xfrm_input_afinfo { int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo); int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo); +void xfrm_flush_gc(void); void xfrm_state_delete_tunnel(struct xfrm_state *x); struct xfrm_type { diff --git a/net/ipv6/xfrm6_tunnel.c b/net/ipv6/xfrm6_tunnel.c index f85f0d7480ac..4a46df8441c9 100644 --- a/net/ipv6/xfrm6_tunnel.c +++ b/net/ipv6/xfrm6_tunnel.c @@ -341,6 +341,9 @@ static void __net_exit xfrm6_tunnel_net_exit(struct net *net) struct xfrm6_tunnel_net *xfrm6_tn = xfrm6_tunnel_pernet(net); unsigned int i; + xfrm_state_flush(net, IPSEC_PROTO_ANY, false); + xfrm_flush_gc(); + for (i = 0; i < XFRM6_TUNNEL_SPI_BYADDR_HSIZE; i++) WARN_ON_ONCE(!hlist_empty(&xfrm6_tn->spi_byaddr[i])); diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index f9d2f2233f09..6c177ae7a6d9 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -2175,6 +2175,12 @@ struct xfrm_state_afinfo *xfrm_state_get_afinfo(unsigned int family) return afinfo; } +void xfrm_flush_gc(void) +{ + flush_work(&xfrm_state_gc_work); +} +EXPORT_SYMBOL(xfrm_flush_gc); + /* Temporarily located here until net/xfrm/xfrm_tunnel.c is created */ void xfrm_state_delete_tunnel(struct xfrm_state *x) { -- cgit v1.2.3 From 4ad69b80e886a845f56ce0a3d10211208693d92b Mon Sep 17 00:00:00 2001 From: Jerome Brunet Date: Mon, 9 Apr 2018 15:59:20 +0200 Subject: clk: honor CLK_MUX_ROUND_CLOSEST in generic clk mux CLK_MUX_ROUND_CLOSEST is part of the clk_mux documentation but clk_mux directly calls __clk_mux_determine_rate(), which overrides the flag. As result, if clk_mux is instantiated with CLK_MUX_ROUND_CLOSEST, the flag will be ignored and the clock rounded down. To solve this, this patch expose clk_mux_determine_rate_flags() in the clk-provider API and uses it in the determine_rate() callback of clk_mux. Fixes: 15a02c1f6dd7 ("clk: Add __clk_mux_determine_rate_closest") Signed-off-by: Jerome Brunet Signed-off-by: Stephen Boyd --- drivers/clk/clk-mux.c | 10 +++++++++- drivers/clk/clk.c | 7 ++++--- include/linux/clk-provider.h | 3 +++ 3 files changed, 16 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/clk/clk-mux.c b/drivers/clk/clk-mux.c index ac4a042f8658..1628b93655ed 100644 --- a/drivers/clk/clk-mux.c +++ b/drivers/clk/clk-mux.c @@ -112,10 +112,18 @@ static int clk_mux_set_parent(struct clk_hw *hw, u8 index) return 0; } +static int clk_mux_determine_rate(struct clk_hw *hw, + struct clk_rate_request *req) +{ + struct clk_mux *mux = to_clk_mux(hw); + + return clk_mux_determine_rate_flags(hw, req, mux->flags); +} + const struct clk_ops clk_mux_ops = { .get_parent = clk_mux_get_parent, .set_parent = clk_mux_set_parent, - .determine_rate = __clk_mux_determine_rate, + .determine_rate = clk_mux_determine_rate, }; EXPORT_SYMBOL_GPL(clk_mux_ops); diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c index ea67ac81c6f9..7af555f0e60c 100644 --- a/drivers/clk/clk.c +++ b/drivers/clk/clk.c @@ -426,9 +426,9 @@ static bool mux_is_better_rate(unsigned long rate, unsigned long now, return now <= rate && now > best; } -static int -clk_mux_determine_rate_flags(struct clk_hw *hw, struct clk_rate_request *req, - unsigned long flags) +int clk_mux_determine_rate_flags(struct clk_hw *hw, + struct clk_rate_request *req, + unsigned long flags) { struct clk_core *core = hw->core, *parent, *best_parent = NULL; int i, num_parents, ret; @@ -488,6 +488,7 @@ out: return 0; } +EXPORT_SYMBOL_GPL(clk_mux_determine_rate_flags); struct clk *__clk_lookup(const char *name) { diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h index 210a890008f9..1d25e149c1c5 100644 --- a/include/linux/clk-provider.h +++ b/include/linux/clk-provider.h @@ -765,6 +765,9 @@ int __clk_mux_determine_rate(struct clk_hw *hw, int __clk_determine_rate(struct clk_hw *core, struct clk_rate_request *req); int __clk_mux_determine_rate_closest(struct clk_hw *hw, struct clk_rate_request *req); +int clk_mux_determine_rate_flags(struct clk_hw *hw, + struct clk_rate_request *req, + unsigned long flags); void clk_hw_reparent(struct clk_hw *hw, struct clk_hw *new_parent); void clk_hw_set_rate_range(struct clk_hw *hw, unsigned long min_rate, unsigned long max_rate); -- cgit v1.2.3 From a7cfebcb7594a24609268f91299ab85ba064bf82 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 3 Apr 2018 14:33:49 +0200 Subject: cfg80211: limit wiphy names to 128 bytes There's currently no limit on wiphy names, other than netlink message size and memory limitations, but that causes issues when, for example, the wiphy name is used in a uevent, e.g. in rfkill where we use the same name for the rfkill instance, and then the buffer there is "only" 2k for the environment variables. This was reported by syzkaller, which used a 4k name. Limit the name to something reasonable, I randomly picked 128. Reported-by: syzbot+230d9e642a85d3fec29c@syzkaller.appspotmail.com Signed-off-by: Johannes Berg --- include/uapi/linux/nl80211.h | 2 ++ net/wireless/core.c | 3 +++ 2 files changed, 5 insertions(+) (limited to 'include') diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index 15daf5e2638d..9c3630146cec 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -2698,6 +2698,8 @@ enum nl80211_attrs { #define NL80211_ATTR_KEYS NL80211_ATTR_KEYS #define NL80211_ATTR_FEATURE_FLAGS NL80211_ATTR_FEATURE_FLAGS +#define NL80211_WIPHY_NAME_MAXLEN 128 + #define NL80211_MAX_SUPP_RATES 32 #define NL80211_MAX_SUPP_HT_RATES 77 #define NL80211_MAX_SUPP_REG_RULES 64 diff --git a/net/wireless/core.c b/net/wireless/core.c index a6f3cac8c640..c0fd8a85e7f7 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -95,6 +95,9 @@ static int cfg80211_dev_check_name(struct cfg80211_registered_device *rdev, ASSERT_RTNL(); + if (strlen(newname) > NL80211_WIPHY_NAME_MAXLEN) + return -EINVAL; + /* prohibit calling the thing phy%d when %d is not its number */ sscanf(newname, PHY_NAME "%d%n", &wiphy_idx, &taken); if (taken == strlen(newname) && wiphy_idx != rdev->wiphy_idx) { -- cgit v1.2.3 From d50e14abe2d0024aa527b89c7990147df5d531a5 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Fri, 20 Apr 2018 09:49:10 -0600 Subject: uapi: Fix SPDX tags for files referring to the 'OpenIB.org' license Based on discussion with Kate Stewart this license is not a BSD-2-Clause, but is now formally identified as Linux-OpenIB by SPDX. The key difference between the licenses is in the 'warranty' paragraph. if_infiniband.h refers to the 'OpenIB.org' license, but does not include the text, instead it links to an obsolete web site that contains a license that matches the BSD-2-Clause SPX. There is no 'three clause' version of the OpenIB.org license. Signed-off-by: Jason Gunthorpe Acked-by: David S. Miller Signed-off-by: Doug Ledford --- include/uapi/linux/if_infiniband.h | 2 +- include/uapi/linux/rds.h | 2 +- include/uapi/linux/tls.h | 2 +- include/uapi/rdma/cxgb3-abi.h | 2 +- include/uapi/rdma/cxgb4-abi.h | 2 +- include/uapi/rdma/hns-abi.h | 2 +- include/uapi/rdma/ib_user_cm.h | 2 +- include/uapi/rdma/ib_user_ioctl_verbs.h | 2 +- include/uapi/rdma/ib_user_mad.h | 2 +- include/uapi/rdma/ib_user_sa.h | 2 +- include/uapi/rdma/ib_user_verbs.h | 2 +- include/uapi/rdma/mlx4-abi.h | 2 +- include/uapi/rdma/mlx5-abi.h | 2 +- include/uapi/rdma/mthca-abi.h | 2 +- include/uapi/rdma/nes-abi.h | 2 +- include/uapi/rdma/qedr-abi.h | 2 +- include/uapi/rdma/rdma_user_cm.h | 2 +- include/uapi/rdma/rdma_user_ioctl.h | 2 +- include/uapi/rdma/rdma_user_rxe.h | 2 +- 19 files changed, 19 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/if_infiniband.h b/include/uapi/linux/if_infiniband.h index 050b92dcf8cf..0fc33bf30e45 100644 --- a/include/uapi/linux/if_infiniband.h +++ b/include/uapi/linux/if_infiniband.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ /* * This software is available to you under a choice of one of two * licenses. You may choose to be licensed under the terms of the GNU diff --git a/include/uapi/linux/rds.h b/include/uapi/linux/rds.h index a66b213de3d7..20c6bd0b0007 100644 --- a/include/uapi/linux/rds.h +++ b/include/uapi/linux/rds.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2008 Oracle. All rights reserved. * diff --git a/include/uapi/linux/tls.h b/include/uapi/linux/tls.h index c6633e97eca4..ff02287495ac 100644 --- a/include/uapi/linux/tls.h +++ b/include/uapi/linux/tls.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2016-2017, Mellanox Technologies. All rights reserved. * diff --git a/include/uapi/rdma/cxgb3-abi.h b/include/uapi/rdma/cxgb3-abi.h index 9acb4b7a6246..85aed672f43e 100644 --- a/include/uapi/rdma/cxgb3-abi.h +++ b/include/uapi/rdma/cxgb3-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2006 Chelsio, Inc. All rights reserved. * diff --git a/include/uapi/rdma/cxgb4-abi.h b/include/uapi/rdma/cxgb4-abi.h index 1fefd0140c26..a159ba8dcf8f 100644 --- a/include/uapi/rdma/cxgb4-abi.h +++ b/include/uapi/rdma/cxgb4-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved. * diff --git a/include/uapi/rdma/hns-abi.h b/include/uapi/rdma/hns-abi.h index 7092c8de4bd8..78613b609fa8 100644 --- a/include/uapi/rdma/hns-abi.h +++ b/include/uapi/rdma/hns-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2016 Hisilicon Limited. * diff --git a/include/uapi/rdma/ib_user_cm.h b/include/uapi/rdma/ib_user_cm.h index 4a8f9562f7cd..e2709bb8cb18 100644 --- a/include/uapi/rdma/ib_user_cm.h +++ b/include/uapi/rdma/ib_user_cm.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2005 Topspin Communications. All rights reserved. * Copyright (c) 2005 Intel Corporation. All rights reserved. diff --git a/include/uapi/rdma/ib_user_ioctl_verbs.h b/include/uapi/rdma/ib_user_ioctl_verbs.h index 04e46ea517d3..625545d862d7 100644 --- a/include/uapi/rdma/ib_user_ioctl_verbs.h +++ b/include/uapi/rdma/ib_user_ioctl_verbs.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2017-2018, Mellanox Technologies inc. All rights reserved. * diff --git a/include/uapi/rdma/ib_user_mad.h b/include/uapi/rdma/ib_user_mad.h index ef92118dad97..90c0cf228020 100644 --- a/include/uapi/rdma/ib_user_mad.h +++ b/include/uapi/rdma/ib_user_mad.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2004 Topspin Communications. All rights reserved. * Copyright (c) 2005 Voltaire, Inc. All rights reserved. diff --git a/include/uapi/rdma/ib_user_sa.h b/include/uapi/rdma/ib_user_sa.h index 0d2607f0cd20..435155d6e1c6 100644 --- a/include/uapi/rdma/ib_user_sa.h +++ b/include/uapi/rdma/ib_user_sa.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2005 Intel Corporation. All rights reserved. * diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h index 9be07394fdbe..6aeb03315b0b 100644 --- a/include/uapi/rdma/ib_user_verbs.h +++ b/include/uapi/rdma/ib_user_verbs.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2005 Topspin Communications. All rights reserved. * Copyright (c) 2005, 2006 Cisco Systems. All rights reserved. diff --git a/include/uapi/rdma/mlx4-abi.h b/include/uapi/rdma/mlx4-abi.h index 04f64bc4045f..f74557528175 100644 --- a/include/uapi/rdma/mlx4-abi.h +++ b/include/uapi/rdma/mlx4-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2007 Cisco Systems, Inc. All rights reserved. * Copyright (c) 2007, 2008 Mellanox Technologies. All rights reserved. diff --git a/include/uapi/rdma/mlx5-abi.h b/include/uapi/rdma/mlx5-abi.h index cb4a02c4a1ce..fdaf00e20649 100644 --- a/include/uapi/rdma/mlx5-abi.h +++ b/include/uapi/rdma/mlx5-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2013-2015, Mellanox Technologies. All rights reserved. * diff --git a/include/uapi/rdma/mthca-abi.h b/include/uapi/rdma/mthca-abi.h index ac756cd9e807..91b12e1a6f43 100644 --- a/include/uapi/rdma/mthca-abi.h +++ b/include/uapi/rdma/mthca-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2005 Topspin Communications. All rights reserved. * Copyright (c) 2005, 2006 Cisco Systems. All rights reserved. diff --git a/include/uapi/rdma/nes-abi.h b/include/uapi/rdma/nes-abi.h index 35bfd4015d07..f80495baa969 100644 --- a/include/uapi/rdma/nes-abi.h +++ b/include/uapi/rdma/nes-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2006 - 2011 Intel Corporation. All rights reserved. * Copyright (c) 2005 Topspin Communications. All rights reserved. diff --git a/include/uapi/rdma/qedr-abi.h b/include/uapi/rdma/qedr-abi.h index 8ba098900e9a..24c658b3c790 100644 --- a/include/uapi/rdma/qedr-abi.h +++ b/include/uapi/rdma/qedr-abi.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* QLogic qedr NIC Driver * Copyright (c) 2015-2016 QLogic Corporation * diff --git a/include/uapi/rdma/rdma_user_cm.h b/include/uapi/rdma/rdma_user_cm.h index e1269024af47..0d1e78ebad05 100644 --- a/include/uapi/rdma/rdma_user_cm.h +++ b/include/uapi/rdma/rdma_user_cm.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2005-2006 Intel Corporation. All rights reserved. * diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h index d223f4164a0f..d92d2721b28c 100644 --- a/include/uapi/rdma/rdma_user_ioctl.h +++ b/include/uapi/rdma/rdma_user_ioctl.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2016 Mellanox Technologies, LTD. All rights reserved. * diff --git a/include/uapi/rdma/rdma_user_rxe.h b/include/uapi/rdma/rdma_user_rxe.h index 1f8a9e7daea4..44ef6a3b7afc 100644 --- a/include/uapi/rdma/rdma_user_rxe.h +++ b/include/uapi/rdma/rdma_user_rxe.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ +/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */ /* * Copyright (c) 2016 Mellanox Technologies Ltd. All rights reserved. * -- cgit v1.2.3 From 6899b32b5b2dee358936b82b8363b716607a138f Mon Sep 17 00:00:00 2001 From: Roman Gushchin Date: Mon, 23 Apr 2018 18:09:21 +0100 Subject: bpf: disable and restore preemption in __BPF_PROG_RUN_ARRAY Running bpf programs requires disabled preemption, however at least some* of the BPF_PROG_RUN_ARRAY users do not follow this rule. To fix this bug, and also to make it not happen in the future, let's add explicit preemption disabling/re-enabling to the __BPF_PROG_RUN_ARRAY code. * for example: [ 17.624472] RIP: 0010:__cgroup_bpf_run_filter_sk+0x1c4/0x1d0 ... [ 17.640890] inet6_create+0x3eb/0x520 [ 17.641405] __sock_create+0x242/0x340 [ 17.641939] __sys_socket+0x57/0xe0 [ 17.642370] ? trace_hardirqs_off_thunk+0x1a/0x1c [ 17.642944] SyS_socket+0xa/0x10 [ 17.643357] do_syscall_64+0x79/0x220 [ 17.643879] entry_SYSCALL_64_after_hwframe+0x42/0xb7 Signed-off-by: Roman Gushchin Cc: Alexei Starovoitov Cc: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- include/linux/bpf.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 486e65e3db26..dc586cc64bc2 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -351,6 +351,7 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array, struct bpf_prog **_prog, *__prog; \ struct bpf_prog_array *_array; \ u32 _ret = 1; \ + preempt_disable(); \ rcu_read_lock(); \ _array = rcu_dereference(array); \ if (unlikely(check_non_null && !_array))\ @@ -362,6 +363,7 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array, } \ _out: \ rcu_read_unlock(); \ + preempt_enable_no_resched(); \ _ret; \ }) -- cgit v1.2.3 From ba6b8de423f8d0dee48d6030288ed81c03ddf9f0 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Mon, 23 Apr 2018 15:39:23 -0700 Subject: bpf: sockmap, map_release does not hold refcnt for pinned maps Relying on map_release hook to decrement the reference counts when a map is removed only works if the map is not being pinned. In the pinned case the ref is decremented immediately and the BPF programs released. After this BPF programs may not be in-use which is not what the user would expect. This patch moves the release logic into bpf_map_put_uref() and brings sockmap in-line with how a similar case is handled in prog array maps. Fixes: 3d9e952697de ("bpf: sockmap, fix leaking maps with attached but not detached progs") Signed-off-by: John Fastabend Signed-off-by: Daniel Borkmann --- include/linux/bpf.h | 2 +- kernel/bpf/arraymap.c | 3 ++- kernel/bpf/sockmap.c | 4 ++-- kernel/bpf/syscall.c | 4 ++-- 4 files changed, 7 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/bpf.h b/include/linux/bpf.h index dc586cc64bc2..469b20e1dd7e 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -31,6 +31,7 @@ struct bpf_map_ops { void (*map_release)(struct bpf_map *map, struct file *map_file); void (*map_free)(struct bpf_map *map); int (*map_get_next_key)(struct bpf_map *map, void *key, void *next_key); + void (*map_release_uref)(struct bpf_map *map); /* funcs callable from userspace and from eBPF programs */ void *(*map_lookup_elem)(struct bpf_map *map, void *key); @@ -436,7 +437,6 @@ int bpf_stackmap_copy(struct bpf_map *map, void *key, void *value); int bpf_fd_array_map_update_elem(struct bpf_map *map, struct file *map_file, void *key, void *value, u64 map_flags); int bpf_fd_array_map_lookup_elem(struct bpf_map *map, void *key, u32 *value); -void bpf_fd_array_map_clear(struct bpf_map *map); int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file, void *key, void *value, u64 map_flags); int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value); diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index 14750e7c5ee4..027107f4be53 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -476,7 +476,7 @@ static u32 prog_fd_array_sys_lookup_elem(void *ptr) } /* decrement refcnt of all bpf_progs that are stored in this map */ -void bpf_fd_array_map_clear(struct bpf_map *map) +static void bpf_fd_array_map_clear(struct bpf_map *map) { struct bpf_array *array = container_of(map, struct bpf_array, map); int i; @@ -495,6 +495,7 @@ const struct bpf_map_ops prog_array_map_ops = { .map_fd_get_ptr = prog_fd_array_get_ptr, .map_fd_put_ptr = prog_fd_array_put_ptr, .map_fd_sys_lookup_elem = prog_fd_array_sys_lookup_elem, + .map_release_uref = bpf_fd_array_map_clear, }; static struct bpf_event_entry *bpf_event_entry_gen(struct file *perf_file, diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index a3b21385e947..a73d484b6e4c 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -1831,7 +1831,7 @@ static int sock_map_update_elem(struct bpf_map *map, return err; } -static void sock_map_release(struct bpf_map *map, struct file *map_file) +static void sock_map_release(struct bpf_map *map) { struct bpf_stab *stab = container_of(map, struct bpf_stab, map); struct bpf_prog *orig; @@ -1855,7 +1855,7 @@ const struct bpf_map_ops sock_map_ops = { .map_get_next_key = sock_map_get_next_key, .map_update_elem = sock_map_update_elem, .map_delete_elem = sock_map_delete_elem, - .map_release = sock_map_release, + .map_release_uref = sock_map_release, }; BPF_CALL_4(bpf_sock_map_update, struct bpf_sock_ops_kern *, bpf_sock, diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 4ca46df19c9a..ebfe9f29dae8 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -257,8 +257,8 @@ static void bpf_map_free_deferred(struct work_struct *work) static void bpf_map_put_uref(struct bpf_map *map) { if (atomic_dec_and_test(&map->usercnt)) { - if (map->map_type == BPF_MAP_TYPE_PROG_ARRAY) - bpf_fd_array_map_clear(map); + if (map->ops->map_release_uref) + map->ops->map_release_uref(map); } } -- cgit v1.2.3 From 672e314b21dc614894e69bb56a2b55cc7d256810 Mon Sep 17 00:00:00 2001 From: Matt Atwood Date: Mon, 23 Apr 2018 15:28:03 -0700 Subject: drm/i915/kbl: Add KBL GT2 sku Adding a missing GT2 sku discovered off hardware. Signed-off-by: Matt Atwood Reviewed-by: Clint Taylor Signed-off-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/1524522483-19987-1-git-send-email-matthew.s.atwood@intel.com --- include/drm/i915_pciids.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/drm/i915_pciids.h b/include/drm/i915_pciids.h index 70f0c2535b87..bab70ff6e78b 100644 --- a/include/drm/i915_pciids.h +++ b/include/drm/i915_pciids.h @@ -349,6 +349,7 @@ #define INTEL_KBL_GT2_IDS(info) \ INTEL_VGA_DEVICE(0x5916, info), /* ULT GT2 */ \ INTEL_VGA_DEVICE(0x5917, info), /* Mobile GT2 */ \ + INTEL_VGA_DEVICE(0x591C, info), /* ULX GT2 */ \ INTEL_VGA_DEVICE(0x5921, info), /* ULT GT2F */ \ INTEL_VGA_DEVICE(0x591E, info), /* ULX GT2 */ \ INTEL_VGA_DEVICE(0x5912, info), /* DT GT2 */ \ -- cgit v1.2.3 From f0316f93897c4c4e67278b175bfbfd3a95ba650a Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 5 Dec 2015 18:41:28 +0000 Subject: drm/i2c: tda9950: add CEC driver Add a CEC driver for the TDA9950, which is a stand-alone I2C CEC device, but is also integrated into HDMI transceivers such as the TDA9989 and TDA19989. The TDA9950 contains a command processor which handles retransmissions and the low level bus protocol. The driver just has to read and write the messages, and handle error conditions. Reviewed-by: Hans Verkuil Signed-off-by: Russell King --- drivers/gpu/drm/i2c/Kconfig | 5 + drivers/gpu/drm/i2c/Makefile | 1 + drivers/gpu/drm/i2c/tda9950.c | 509 ++++++++++++++++++++++++++++++++++ include/linux/platform_data/tda9950.h | 16 ++ 4 files changed, 531 insertions(+) create mode 100644 drivers/gpu/drm/i2c/tda9950.c create mode 100644 include/linux/platform_data/tda9950.h (limited to 'include') diff --git a/drivers/gpu/drm/i2c/Kconfig b/drivers/gpu/drm/i2c/Kconfig index a6c92beb410a..3a232f5ff0a1 100644 --- a/drivers/gpu/drm/i2c/Kconfig +++ b/drivers/gpu/drm/i2c/Kconfig @@ -26,4 +26,9 @@ config DRM_I2C_NXP_TDA998X help Support for NXP Semiconductors TDA998X HDMI encoders. +config DRM_I2C_NXP_TDA9950 + tristate "NXP Semiconductors TDA9950/TDA998X HDMI CEC" + select CEC_NOTIFIER + select CEC_CORE + endmenu diff --git a/drivers/gpu/drm/i2c/Makefile b/drivers/gpu/drm/i2c/Makefile index b20100c18ffb..a962f6f08568 100644 --- a/drivers/gpu/drm/i2c/Makefile +++ b/drivers/gpu/drm/i2c/Makefile @@ -7,3 +7,4 @@ obj-$(CONFIG_DRM_I2C_SIL164) += sil164.o tda998x-y := tda998x_drv.o obj-$(CONFIG_DRM_I2C_NXP_TDA998X) += tda998x.o +obj-$(CONFIG_DRM_I2C_NXP_TDA9950) += tda9950.o diff --git a/drivers/gpu/drm/i2c/tda9950.c b/drivers/gpu/drm/i2c/tda9950.c new file mode 100644 index 000000000000..3f7396caad48 --- /dev/null +++ b/drivers/gpu/drm/i2c/tda9950.c @@ -0,0 +1,509 @@ +/* + * TDA9950 Consumer Electronics Control driver + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * The NXP TDA9950 implements the HDMI Consumer Electronics Control + * interface. The host interface is similar to a mailbox: the data + * registers starting at REG_CDR0 are written to send a command to the + * internal CPU, and replies are read from these registers. + * + * As the data registers represent a mailbox, they must be accessed + * as a single I2C transaction. See the TDA9950 data sheet for details. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include + +enum { + REG_CSR = 0x00, + CSR_BUSY = BIT(7), + CSR_INT = BIT(6), + CSR_ERR = BIT(5), + + REG_CER = 0x01, + + REG_CVR = 0x02, + + REG_CCR = 0x03, + CCR_RESET = BIT(7), + CCR_ON = BIT(6), + + REG_ACKH = 0x04, + REG_ACKL = 0x05, + + REG_CCONR = 0x06, + CCONR_ENABLE_ERROR = BIT(4), + CCONR_RETRY_MASK = 7, + + REG_CDR0 = 0x07, + + CDR1_REQ = 0x00, + CDR1_CNF = 0x01, + CDR1_IND = 0x81, + CDR1_ERR = 0x82, + CDR1_IER = 0x83, + + CDR2_CNF_SUCCESS = 0x00, + CDR2_CNF_OFF_STATE = 0x80, + CDR2_CNF_BAD_REQ = 0x81, + CDR2_CNF_CEC_ACCESS = 0x82, + CDR2_CNF_ARB_ERROR = 0x83, + CDR2_CNF_BAD_TIMING = 0x84, + CDR2_CNF_NACK_ADDR = 0x85, + CDR2_CNF_NACK_DATA = 0x86, +}; + +struct tda9950_priv { + struct i2c_client *client; + struct device *hdmi; + struct cec_adapter *adap; + struct tda9950_glue *glue; + u16 addresses; + struct cec_msg rx_msg; + struct cec_notifier *notify; + bool open; +}; + +static int tda9950_write_range(struct i2c_client *client, u8 addr, u8 *p, int cnt) +{ + struct i2c_msg msg; + u8 buf[cnt + 1]; + int ret; + + buf[0] = addr; + memcpy(buf + 1, p, cnt); + + msg.addr = client->addr; + msg.flags = 0; + msg.len = cnt + 1; + msg.buf = buf; + + dev_dbg(&client->dev, "wr 0x%02x: %*ph\n", addr, cnt, p); + + ret = i2c_transfer(client->adapter, &msg, 1); + if (ret < 0) + dev_err(&client->dev, "Error %d writing to cec:0x%x\n", ret, addr); + return ret < 0 ? ret : 0; +} + +static void tda9950_write(struct i2c_client *client, u8 addr, u8 val) +{ + tda9950_write_range(client, addr, &val, 1); +} + +static int tda9950_read_range(struct i2c_client *client, u8 addr, u8 *p, int cnt) +{ + struct i2c_msg msg[2]; + int ret; + + msg[0].addr = client->addr; + msg[0].flags = 0; + msg[0].len = 1; + msg[0].buf = &addr; + msg[1].addr = client->addr; + msg[1].flags = I2C_M_RD; + msg[1].len = cnt; + msg[1].buf = p; + + ret = i2c_transfer(client->adapter, msg, 2); + if (ret < 0) + dev_err(&client->dev, "Error %d reading from cec:0x%x\n", ret, addr); + + dev_dbg(&client->dev, "rd 0x%02x: %*ph\n", addr, cnt, p); + + return ret; +} + +static u8 tda9950_read(struct i2c_client *client, u8 addr) +{ + int ret; + u8 val; + + ret = tda9950_read_range(client, addr, &val, 1); + if (ret < 0) + val = 0; + + return val; +} + +static irqreturn_t tda9950_irq(int irq, void *data) +{ + struct tda9950_priv *priv = data; + unsigned int tx_status; + u8 csr, cconr, buf[19]; + u8 arb_lost_cnt, nack_cnt, err_cnt; + + if (!priv->open) + return IRQ_NONE; + + csr = tda9950_read(priv->client, REG_CSR); + if (!(csr & CSR_INT)) + return IRQ_NONE; + + cconr = tda9950_read(priv->client, REG_CCONR) & CCONR_RETRY_MASK; + + tda9950_read_range(priv->client, REG_CDR0, buf, sizeof(buf)); + + /* + * This should never happen: the data sheet says that there will + * always be a valid message if the interrupt line is asserted. + */ + if (buf[0] == 0) { + dev_warn(&priv->client->dev, "interrupt pending, but no message?\n"); + return IRQ_NONE; + } + + switch (buf[1]) { + case CDR1_CNF: /* transmit result */ + arb_lost_cnt = nack_cnt = err_cnt = 0; + switch (buf[2]) { + case CDR2_CNF_SUCCESS: + tx_status = CEC_TX_STATUS_OK; + break; + + case CDR2_CNF_ARB_ERROR: + tx_status = CEC_TX_STATUS_ARB_LOST; + arb_lost_cnt = cconr; + break; + + case CDR2_CNF_NACK_ADDR: + tx_status = CEC_TX_STATUS_NACK; + nack_cnt = cconr; + break; + + default: /* some other error, refer to TDA9950 docs */ + dev_err(&priv->client->dev, "CNF reply error 0x%02x\n", + buf[2]); + tx_status = CEC_TX_STATUS_ERROR; + err_cnt = cconr; + break; + } + /* TDA9950 executes all retries for us */ + tx_status |= CEC_TX_STATUS_MAX_RETRIES; + cec_transmit_done(priv->adap, tx_status, arb_lost_cnt, + nack_cnt, 0, err_cnt); + break; + + case CDR1_IND: + priv->rx_msg.len = buf[0] - 2; + if (priv->rx_msg.len > CEC_MAX_MSG_SIZE) + priv->rx_msg.len = CEC_MAX_MSG_SIZE; + + memcpy(priv->rx_msg.msg, buf + 2, priv->rx_msg.len); + cec_received_msg(priv->adap, &priv->rx_msg); + break; + + default: /* unknown */ + dev_err(&priv->client->dev, "unknown service id 0x%02x\n", + buf[1]); + break; + } + + return IRQ_HANDLED; +} + +static int tda9950_cec_transmit(struct cec_adapter *adap, u8 attempts, + u32 signal_free_time, struct cec_msg *msg) +{ + struct tda9950_priv *priv = adap->priv; + u8 buf[CEC_MAX_MSG_SIZE + 2]; + + buf[0] = 2 + msg->len; + buf[1] = CDR1_REQ; + memcpy(buf + 2, msg->msg, msg->len); + + if (attempts > 5) + attempts = 5; + + tda9950_write(priv->client, REG_CCONR, attempts); + + return tda9950_write_range(priv->client, REG_CDR0, buf, 2 + msg->len); +} + +static int tda9950_cec_adap_log_addr(struct cec_adapter *adap, u8 addr) +{ + struct tda9950_priv *priv = adap->priv; + u16 addresses; + u8 buf[2]; + + if (addr == CEC_LOG_ADDR_INVALID) + addresses = priv->addresses = 0; + else + addresses = priv->addresses |= BIT(addr); + + /* TDA9950 doesn't want address 15 set */ + addresses &= 0x7fff; + buf[0] = addresses >> 8; + buf[1] = addresses; + + return tda9950_write_range(priv->client, REG_ACKH, buf, 2); +} + +/* + * When operating as part of the TDA998x, we need additional handling + * to initialise and shut down the TDA9950 part of the device. These + * two hooks are provided to allow the TDA998x code to perform those + * activities. + */ +static int tda9950_glue_open(struct tda9950_priv *priv) +{ + int ret = 0; + + if (priv->glue && priv->glue->open) + ret = priv->glue->open(priv->glue->data); + + priv->open = true; + + return ret; +} + +static void tda9950_glue_release(struct tda9950_priv *priv) +{ + priv->open = false; + + if (priv->glue && priv->glue->release) + priv->glue->release(priv->glue->data); +} + +static int tda9950_open(struct tda9950_priv *priv) +{ + struct i2c_client *client = priv->client; + int ret; + + ret = tda9950_glue_open(priv); + if (ret) + return ret; + + /* Reset the TDA9950, and wait 250ms for it to recover */ + tda9950_write(client, REG_CCR, CCR_RESET); + msleep(250); + + tda9950_cec_adap_log_addr(priv->adap, CEC_LOG_ADDR_INVALID); + + /* Start the command processor */ + tda9950_write(client, REG_CCR, CCR_ON); + + return 0; +} + +static void tda9950_release(struct tda9950_priv *priv) +{ + struct i2c_client *client = priv->client; + int timeout = 50; + u8 csr; + + /* Stop the command processor */ + tda9950_write(client, REG_CCR, 0); + + /* Wait up to .5s for it to signal non-busy */ + do { + csr = tda9950_read(client, REG_CSR); + if (!(csr & CSR_BUSY) || --timeout) + break; + msleep(10); + } while (1); + + /* Warn the user that their IRQ may die if it's shared. */ + if (csr & CSR_BUSY) + dev_warn(&client->dev, "command processor failed to stop, irq%d may die (csr=0x%02x)\n", + client->irq, csr); + + tda9950_glue_release(priv); +} + +static int tda9950_cec_adap_enable(struct cec_adapter *adap, bool enable) +{ + struct tda9950_priv *priv = adap->priv; + + if (!enable) { + tda9950_release(priv); + return 0; + } else { + return tda9950_open(priv); + } +} + +static const struct cec_adap_ops tda9950_cec_ops = { + .adap_enable = tda9950_cec_adap_enable, + .adap_log_addr = tda9950_cec_adap_log_addr, + .adap_transmit = tda9950_cec_transmit, +}; + +/* + * When operating as part of the TDA998x, we need to claim additional + * resources. These two hooks permit the management of those resources. + */ +static void tda9950_devm_glue_exit(void *data) +{ + struct tda9950_glue *glue = data; + + if (glue && glue->exit) + glue->exit(glue->data); +} + +static int tda9950_devm_glue_init(struct device *dev, struct tda9950_glue *glue) +{ + int ret; + + if (glue && glue->init) { + ret = glue->init(glue->data); + if (ret) + return ret; + } + + ret = devm_add_action(dev, tda9950_devm_glue_exit, glue); + if (ret) + tda9950_devm_glue_exit(glue); + + return ret; +} + +static void tda9950_cec_del(void *data) +{ + struct tda9950_priv *priv = data; + + cec_delete_adapter(priv->adap); +} + +static int tda9950_probe(struct i2c_client *client, + const struct i2c_device_id *id) +{ + struct tda9950_glue *glue = client->dev.platform_data; + struct device *dev = &client->dev; + struct tda9950_priv *priv; + unsigned long irqflags; + int ret; + u8 cvr; + + /* + * We must have I2C functionality: our multi-byte accesses + * must be performed as a single contiguous transaction. + */ + if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) { + dev_err(&client->dev, + "adapter does not support I2C functionality\n"); + return -ENXIO; + } + + /* We must have an interrupt to be functional. */ + if (client->irq <= 0) { + dev_err(&client->dev, "driver requires an interrupt\n"); + return -ENXIO; + } + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + priv->client = client; + priv->glue = glue; + + i2c_set_clientdata(client, priv); + + /* + * If we're part of a TDA998x, we want the class devices to be + * associated with the HDMI Tx so we have a tight relationship + * between the HDMI interface and the CEC interface. + */ + priv->hdmi = dev; + if (glue && glue->parent) + priv->hdmi = glue->parent; + + priv->adap = cec_allocate_adapter(&tda9950_cec_ops, priv, "tda9950", + CEC_CAP_DEFAULTS, + CEC_MAX_LOG_ADDRS); + if (IS_ERR(priv->adap)) + return PTR_ERR(priv->adap); + + ret = devm_add_action(dev, tda9950_cec_del, priv); + if (ret) { + cec_delete_adapter(priv->adap); + return ret; + } + + ret = tda9950_devm_glue_init(dev, glue); + if (ret) + return ret; + + ret = tda9950_glue_open(priv); + if (ret) + return ret; + + cvr = tda9950_read(client, REG_CVR); + + dev_info(&client->dev, + "TDA9950 CEC interface, hardware version %u.%u\n", + cvr >> 4, cvr & 15); + + tda9950_glue_release(priv); + + irqflags = IRQF_TRIGGER_FALLING; + if (glue) + irqflags = glue->irq_flags; + + ret = devm_request_threaded_irq(dev, client->irq, NULL, tda9950_irq, + irqflags | IRQF_SHARED | IRQF_ONESHOT, + dev_name(&client->dev), priv); + if (ret < 0) + return ret; + + priv->notify = cec_notifier_get(priv->hdmi); + if (!priv->notify) + return -ENOMEM; + + ret = cec_register_adapter(priv->adap, priv->hdmi); + if (ret < 0) { + cec_notifier_put(priv->notify); + return ret; + } + + /* + * CEC documentation says we must not call cec_delete_adapter + * after a successful call to cec_register_adapter(). + */ + devm_remove_action(dev, tda9950_cec_del, priv); + + cec_register_cec_notifier(priv->adap, priv->notify); + + return 0; +} + +static int tda9950_remove(struct i2c_client *client) +{ + struct tda9950_priv *priv = i2c_get_clientdata(client); + + cec_unregister_adapter(priv->adap); + cec_notifier_put(priv->notify); + + return 0; +} + +static struct i2c_device_id tda9950_ids[] = { + { "tda9950", 0 }, + { }, +}; +MODULE_DEVICE_TABLE(i2c, tda9950_ids); + +static struct i2c_driver tda9950_driver = { + .probe = tda9950_probe, + .remove = tda9950_remove, + .driver = { + .name = "tda9950", + }, + .id_table = tda9950_ids, +}; + +module_i2c_driver(tda9950_driver); + +MODULE_AUTHOR("Russell King "); +MODULE_DESCRIPTION("TDA9950/TDA998x Consumer Electronics Control Driver"); +MODULE_LICENSE("GPL v2"); diff --git a/include/linux/platform_data/tda9950.h b/include/linux/platform_data/tda9950.h new file mode 100644 index 000000000000..c65efd461102 --- /dev/null +++ b/include/linux/platform_data/tda9950.h @@ -0,0 +1,16 @@ +#ifndef LINUX_PLATFORM_DATA_TDA9950_H +#define LINUX_PLATFORM_DATA_TDA9950_H + +struct device; + +struct tda9950_glue { + struct device *parent; + unsigned long irq_flags; + void *data; + int (*init)(void *); + void (*exit)(void *); + int (*open)(void *); + void (*release)(void *); +}; + +#endif -- cgit v1.2.3 From fcd58037f28bf70eb17157a51fbf94d466634a7d Mon Sep 17 00:00:00 2001 From: Arnaud Pouliquen Date: Tue, 10 Apr 2018 10:57:25 +0200 Subject: remoteproc: fix crashed parameter logic on stop call Fix rproc_add_subdev parameter name and inverse the crashed logic. Fixes: 880f5b388252 ("remoteproc: Pass type of shutdown to subdev remove") Reviewed-by: Alex Elder Signed-off-by: Arnaud Pouliquen Signed-off-by: Bjorn Andersson --- drivers/remoteproc/remoteproc_core.c | 4 ++-- include/linux/remoteproc.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c index 6d9c5832ce47..a9609d971f7f 100644 --- a/drivers/remoteproc/remoteproc_core.c +++ b/drivers/remoteproc/remoteproc_core.c @@ -1163,7 +1163,7 @@ int rproc_trigger_recovery(struct rproc *rproc) if (ret) return ret; - ret = rproc_stop(rproc, false); + ret = rproc_stop(rproc, true); if (ret) goto unlock_mutex; @@ -1316,7 +1316,7 @@ void rproc_shutdown(struct rproc *rproc) if (!atomic_dec_and_test(&rproc->power)) goto out; - ret = rproc_stop(rproc, true); + ret = rproc_stop(rproc, false); if (ret) { atomic_inc(&rproc->power); goto out; diff --git a/include/linux/remoteproc.h b/include/linux/remoteproc.h index d09a9c7af109..dfdaede9139e 100644 --- a/include/linux/remoteproc.h +++ b/include/linux/remoteproc.h @@ -569,7 +569,7 @@ static inline struct rproc *vdev_to_rproc(struct virtio_device *vdev) void rproc_add_subdev(struct rproc *rproc, struct rproc_subdev *subdev, int (*probe)(struct rproc_subdev *subdev), - void (*remove)(struct rproc_subdev *subdev, bool graceful)); + void (*remove)(struct rproc_subdev *subdev, bool crashed)); void rproc_remove_subdev(struct rproc *rproc, struct rproc_subdev *subdev); -- cgit v1.2.3 From bf0ddaba65ddbb2715af97041da8e7a45b2d8628 Mon Sep 17 00:00:00 2001 From: Omar Sandoval Date: Thu, 26 Apr 2018 00:21:59 -0700 Subject: blk-mq: fix sysfs inflight counter When the blk-mq inflight implementation was added, /proc/diskstats was converted to use it, but /sys/block/$dev/inflight was not. Fix it by adding another helper to count in-flight requests by data direction. Fixes: f299b7c7a9de ("blk-mq: provide internal in-flight variant") Signed-off-by: Omar Sandoval Signed-off-by: Jens Axboe --- block/blk-mq.c | 19 +++++++++++++++++++ block/blk-mq.h | 4 +++- block/genhd.c | 12 ++++++++++++ block/partition-generic.c | 10 ++++++---- include/linux/genhd.h | 4 +++- 5 files changed, 43 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/block/blk-mq.c b/block/blk-mq.c index 5450cbc61f8d..9ce9cac16c3f 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -115,6 +115,25 @@ void blk_mq_in_flight(struct request_queue *q, struct hd_struct *part, blk_mq_queue_tag_busy_iter(q, blk_mq_check_inflight, &mi); } +static void blk_mq_check_inflight_rw(struct blk_mq_hw_ctx *hctx, + struct request *rq, void *priv, + bool reserved) +{ + struct mq_inflight *mi = priv; + + if (rq->part == mi->part) + mi->inflight[rq_data_dir(rq)]++; +} + +void blk_mq_in_flight_rw(struct request_queue *q, struct hd_struct *part, + unsigned int inflight[2]) +{ + struct mq_inflight mi = { .part = part, .inflight = inflight, }; + + inflight[0] = inflight[1] = 0; + blk_mq_queue_tag_busy_iter(q, blk_mq_check_inflight_rw, &mi); +} + void blk_freeze_queue_start(struct request_queue *q) { int freeze_depth; diff --git a/block/blk-mq.h b/block/blk-mq.h index 89b5cd3a6c70..e1bb420dc5d6 100644 --- a/block/blk-mq.h +++ b/block/blk-mq.h @@ -188,7 +188,9 @@ static inline bool blk_mq_hw_queue_mapped(struct blk_mq_hw_ctx *hctx) } void blk_mq_in_flight(struct request_queue *q, struct hd_struct *part, - unsigned int inflight[2]); + unsigned int inflight[2]); +void blk_mq_in_flight_rw(struct request_queue *q, struct hd_struct *part, + unsigned int inflight[2]); static inline void blk_mq_put_dispatch_budget(struct blk_mq_hw_ctx *hctx) { diff --git a/block/genhd.c b/block/genhd.c index dc7e089373b9..c4513fe1adda 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -82,6 +82,18 @@ void part_in_flight(struct request_queue *q, struct hd_struct *part, } } +void part_in_flight_rw(struct request_queue *q, struct hd_struct *part, + unsigned int inflight[2]) +{ + if (q->mq_ops) { + blk_mq_in_flight_rw(q, part, inflight); + return; + } + + inflight[0] = atomic_read(&part->in_flight[0]); + inflight[1] = atomic_read(&part->in_flight[1]); +} + struct hd_struct *__disk_get_part(struct gendisk *disk, int partno) { struct disk_part_tbl *ptbl = rcu_dereference(disk->part_tbl); diff --git a/block/partition-generic.c b/block/partition-generic.c index 08dabcd8b6ae..db57cced9b98 100644 --- a/block/partition-generic.c +++ b/block/partition-generic.c @@ -145,13 +145,15 @@ ssize_t part_stat_show(struct device *dev, jiffies_to_msecs(part_stat_read(p, time_in_queue))); } -ssize_t part_inflight_show(struct device *dev, - struct device_attribute *attr, char *buf) +ssize_t part_inflight_show(struct device *dev, struct device_attribute *attr, + char *buf) { struct hd_struct *p = dev_to_part(dev); + struct request_queue *q = part_to_disk(p)->queue; + unsigned int inflight[2]; - return sprintf(buf, "%8u %8u\n", atomic_read(&p->in_flight[0]), - atomic_read(&p->in_flight[1])); + part_in_flight_rw(q, p, inflight); + return sprintf(buf, "%8u %8u\n", inflight[0], inflight[1]); } #ifdef CONFIG_FAIL_MAKE_REQUEST diff --git a/include/linux/genhd.h b/include/linux/genhd.h index c826b0b5232a..6cb8a5789668 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -368,7 +368,9 @@ static inline void free_part_stats(struct hd_struct *part) part_stat_add(cpu, gendiskp, field, -subnd) void part_in_flight(struct request_queue *q, struct hd_struct *part, - unsigned int inflight[2]); + unsigned int inflight[2]); +void part_in_flight_rw(struct request_queue *q, struct hd_struct *part, + unsigned int inflight[2]); void part_dec_in_flight(struct request_queue *q, struct hd_struct *part, int rw); void part_inc_in_flight(struct request_queue *q, struct hd_struct *part, -- cgit v1.2.3 From 0566e40ce7c493d39006cdd7edf17bfdc52eb2ac Mon Sep 17 00:00:00 2001 From: Rishabh Bhatnagar Date: Wed, 25 Apr 2018 16:42:35 -0700 Subject: tracing: initcall: Ordered comparison of function pointers Using initcall_t in the __field macro generates the following warning with clang version 6.0: include/trace/events/initcall.h:34:3: warning: ordered comparison of function pointers ('initcall_t' (aka 'int (*)(void)') and 'initcall_t') __field macro expands to __field_ext macro which does is_signed_type check on the type argument. Since initcall_t is defined as a function pointer, using it as the type in the __field macro, leads to an ordered comparison of function pointer warning, inside the check. Using __field_struct macro avoids the issue. Link: http://lkml.kernel.org/r/1524699755-29388-1-git-send-email-rishabhb@codeaurora.org Signed-off-by: Rishabh Bhatnagar [ Added comment to why we are using field_struct() ] Signed-off-by: Steven Rostedt (VMware) --- include/trace/events/initcall.h | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/trace/events/initcall.h b/include/trace/events/initcall.h index 8d6cf10d27c9..eb903c3f195f 100644 --- a/include/trace/events/initcall.h +++ b/include/trace/events/initcall.h @@ -31,7 +31,11 @@ TRACE_EVENT(initcall_start, TP_ARGS(func), TP_STRUCT__entry( - __field(initcall_t, func) + /* + * Use field_struct to avoid is_signed_type() + * comparison of a function pointer + */ + __field_struct(initcall_t, func) ), TP_fast_assign( @@ -48,8 +52,12 @@ TRACE_EVENT(initcall_finish, TP_ARGS(func, ret), TP_STRUCT__entry( - __field(initcall_t, func) - __field(int, ret) + /* + * Use field_struct to avoid is_signed_type() + * comparison of a function pointer + */ + __field_struct(initcall_t, func) + __field(int, ret) ), TP_fast_assign( -- cgit v1.2.3 From 6082d9c9c94a408d7409b5f2e4e42ac9e8b16d0d Mon Sep 17 00:00:00 2001 From: Israel Rukshin Date: Thu, 12 Apr 2018 09:49:11 +0000 Subject: net/mlx5: Fix mlx5_get_vector_affinity function Adding the vector offset when calling to mlx5_vector2eqn() is wrong. This is because mlx5_vector2eqn() checks if EQ index is equal to vector number and the fact that the internal completion vectors that mlx5 allocates don't get an EQ index. The second problem here is that using effective_affinity_mask gives the same CPU for different vectors. This leads to unmapped queues when calling it from blk_mq_rdma_map_queues(). This doesn't happen when using affinity_hint mask. Fixes: 2572cf57d75a ("mlx5: fix mlx5_get_vector_affinity to start from completion vector 0") Fixes: 05e0cc84e00c ("net/mlx5: Fix get vector affinity helper function") Signed-off-by: Israel Rukshin Reviewed-by: Max Gurtovoy Reviewed-by: Sagi Grimberg --- drivers/infiniband/hw/mlx5/main.c | 2 +- include/linux/mlx5/driver.h | 12 +++--------- 2 files changed, 4 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index daa919e5a442..241cf4ff9901 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -4757,7 +4757,7 @@ mlx5_ib_get_vector_affinity(struct ib_device *ibdev, int comp_vector) { struct mlx5_ib_dev *dev = to_mdev(ibdev); - return mlx5_get_vector_affinity(dev->mdev, comp_vector); + return mlx5_get_vector_affinity_hint(dev->mdev, comp_vector); } /* The mlx5_ib_multiport_mutex should be held when calling this function */ diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index 767d193c269a..2a156c5dfadd 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -1284,25 +1284,19 @@ enum { }; static inline const struct cpumask * -mlx5_get_vector_affinity(struct mlx5_core_dev *dev, int vector) +mlx5_get_vector_affinity_hint(struct mlx5_core_dev *dev, int vector) { - const struct cpumask *mask; struct irq_desc *desc; unsigned int irq; int eqn; int err; - err = mlx5_vector2eqn(dev, MLX5_EQ_VEC_COMP_BASE + vector, &eqn, &irq); + err = mlx5_vector2eqn(dev, vector, &eqn, &irq); if (err) return NULL; desc = irq_to_desc(irq); -#ifdef CONFIG_GENERIC_IRQ_EFFECTIVE_AFF_MASK - mask = irq_data_get_effective_affinity_mask(&desc->irq_data); -#else - mask = desc->irq_common_data.affinity; -#endif - return mask; + return desc->affinity_hint; } #endif /* MLX5_DRIVER_H */ -- cgit v1.2.3 From ed769520727edbf526e0f510e6c501fc6ba3824a Mon Sep 17 00:00:00 2001 From: Joel Pepper Date: Thu, 26 Apr 2018 20:26:08 +0200 Subject: usb: gadget: composite Allow for larger configuration descriptors The composite framework allows us to create gadgets composed from many different functions, which need to fit into a single configuration descriptor. Some functions (like uvc) can produce configuration descriptors upwards of 2500 bytes on their own. This patch increases the limit from 1024 bytes to 4096. Signed-off-by: Joel Pepper Signed-off-by: Felipe Balbi --- include/linux/usb/composite.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/usb/composite.h b/include/linux/usb/composite.h index 4b6b9283fa7b..8675e145ea8b 100644 --- a/include/linux/usb/composite.h +++ b/include/linux/usb/composite.h @@ -52,7 +52,7 @@ #define USB_GADGET_DELAYED_STATUS 0x7fff /* Impossibly large value */ /* big enough to hold our biggest descriptor */ -#define USB_COMP_EP0_BUFSIZ 1024 +#define USB_COMP_EP0_BUFSIZ 4096 /* OS feature descriptor length <= 4kB */ #define USB_COMP_EP0_OS_DESC_BUFSIZ 4096 -- cgit v1.2.3 From 53692908b0f594285aba18ab848318262332ed25 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Wed, 18 Apr 2018 10:39:04 +0100 Subject: KVM: arm/arm64: vgic: Fix source vcpu issues for GICv2 SGI Now that we make sure we don't inject multiple instances of the same GICv2 SGI at the same time, we've made another bug more obvious: If we exit with an active SGI, we completely lose track of which vcpu it came from. On the next entry, we restore it with 0 as a source, and if that wasn't the right one, too bad. While this doesn't seem to trouble GIC-400, the architectural model gets offended and doesn't deactivate the interrupt on EOI. Another connected issue is that we will happilly make pending an interrupt from another vcpu, overriding the above zero with something that is just as inconsistent. Don't do that. The final issue is that we signal a maintenance interrupt when no pending interrupts are present in the LR. Assuming we've fixed the two issues above, we end-up in a situation where we keep exiting as soon as we've reached the active state, and not be able to inject the following pending. The fix comes in 3 parts: - GICv2 SGIs have their source vcpu saved if they are active on exit, and restored on entry - Multi-SGIs cannot go via the Pending+Active state, as this would corrupt the source field - Multi-SGIs are converted to using MI on EOI instead of NPIE Fixes: 16ca6a607d84bef0 ("KVM: arm/arm64: vgic: Don't populate multiple LRs with the same vintid") Reported-by: Mark Rutland Tested-by: Mark Rutland Reviewed-by: Christoffer Dall Signed-off-by: Marc Zyngier --- include/kvm/arm_vgic.h | 1 + virt/kvm/arm/vgic/vgic-mmio.c | 10 +++++++-- virt/kvm/arm/vgic/vgic-v2.c | 38 +++++++++++++++++++-------------- virt/kvm/arm/vgic/vgic-v3.c | 49 +++++++++++++++++++++++++------------------ virt/kvm/arm/vgic/vgic.c | 30 +++++++------------------- virt/kvm/arm/vgic/vgic.h | 14 +++++++++++++ 6 files changed, 81 insertions(+), 61 deletions(-) (limited to 'include') diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h index 24f03941ada8..e7efe12a81bd 100644 --- a/include/kvm/arm_vgic.h +++ b/include/kvm/arm_vgic.h @@ -131,6 +131,7 @@ struct vgic_irq { u32 mpidr; /* GICv3 target VCPU */ }; u8 source; /* GICv2 SGIs only */ + u8 active_source; /* GICv2 SGIs only */ u8 priority; enum vgic_irq_config config; /* Level or edge */ diff --git a/virt/kvm/arm/vgic/vgic-mmio.c b/virt/kvm/arm/vgic/vgic-mmio.c index dbe99d635c80..ff9655cfeb2f 100644 --- a/virt/kvm/arm/vgic/vgic-mmio.c +++ b/virt/kvm/arm/vgic/vgic-mmio.c @@ -289,10 +289,16 @@ static void vgic_mmio_change_active(struct kvm_vcpu *vcpu, struct vgic_irq *irq, irq->vcpu->cpu != -1) /* VCPU thread is running */ cond_resched_lock(&irq->irq_lock); - if (irq->hw) + if (irq->hw) { vgic_hw_irq_change_active(vcpu, irq, active, !requester_vcpu); - else + } else { + u32 model = vcpu->kvm->arch.vgic.vgic_model; + irq->active = active; + if (model == KVM_DEV_TYPE_ARM_VGIC_V2 && + active && vgic_irq_is_sgi(irq->intid)) + irq->active_source = requester_vcpu->vcpu_id; + } if (irq->active) vgic_queue_irq_unlock(vcpu->kvm, irq, flags); diff --git a/virt/kvm/arm/vgic/vgic-v2.c b/virt/kvm/arm/vgic/vgic-v2.c index 45aa433f018f..a5f2e44f1c33 100644 --- a/virt/kvm/arm/vgic/vgic-v2.c +++ b/virt/kvm/arm/vgic/vgic-v2.c @@ -37,13 +37,6 @@ void vgic_v2_init_lrs(void) vgic_v2_write_lr(i, 0); } -void vgic_v2_set_npie(struct kvm_vcpu *vcpu) -{ - struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2; - - cpuif->vgic_hcr |= GICH_HCR_NPIE; -} - void vgic_v2_set_underflow(struct kvm_vcpu *vcpu) { struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2; @@ -71,13 +64,18 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu) int lr; unsigned long flags; - cpuif->vgic_hcr &= ~(GICH_HCR_UIE | GICH_HCR_NPIE); + cpuif->vgic_hcr &= ~GICH_HCR_UIE; for (lr = 0; lr < vgic_cpu->used_lrs; lr++) { u32 val = cpuif->vgic_lr[lr]; - u32 intid = val & GICH_LR_VIRTUALID; + u32 cpuid, intid = val & GICH_LR_VIRTUALID; struct vgic_irq *irq; + /* Extract the source vCPU id from the LR */ + cpuid = val & GICH_LR_PHYSID_CPUID; + cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT; + cpuid &= 7; + /* Notify fds when the guest EOI'ed a level-triggered SPI */ if (lr_signals_eoi_mi(val) && vgic_valid_spi(vcpu->kvm, intid)) kvm_notify_acked_irq(vcpu->kvm, 0, @@ -90,17 +88,16 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu) /* Always preserve the active bit */ irq->active = !!(val & GICH_LR_ACTIVE_BIT); + if (irq->active && vgic_irq_is_sgi(intid)) + irq->active_source = cpuid; + /* Edge is the only case where we preserve the pending bit */ if (irq->config == VGIC_CONFIG_EDGE && (val & GICH_LR_PENDING_BIT)) { irq->pending_latch = true; - if (vgic_irq_is_sgi(intid)) { - u32 cpuid = val & GICH_LR_PHYSID_CPUID; - - cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT; + if (vgic_irq_is_sgi(intid)) irq->source |= (1 << cpuid); - } } /* @@ -152,8 +149,15 @@ void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr) u32 val = irq->intid; bool allow_pending = true; - if (irq->active) + if (irq->active) { val |= GICH_LR_ACTIVE_BIT; + if (vgic_irq_is_sgi(irq->intid)) + val |= irq->active_source << GICH_LR_PHYSID_CPUID_SHIFT; + if (vgic_irq_is_multi_sgi(irq)) { + allow_pending = false; + val |= GICH_LR_EOI; + } + } if (irq->hw) { val |= GICH_LR_HW; @@ -190,8 +194,10 @@ void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr) BUG_ON(!src); val |= (src - 1) << GICH_LR_PHYSID_CPUID_SHIFT; irq->source &= ~(1 << (src - 1)); - if (irq->source) + if (irq->source) { irq->pending_latch = true; + val |= GICH_LR_EOI; + } } } diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c index 8195f52ae6f0..c7423f3768e5 100644 --- a/virt/kvm/arm/vgic/vgic-v3.c +++ b/virt/kvm/arm/vgic/vgic-v3.c @@ -27,13 +27,6 @@ static bool group1_trap; static bool common_trap; static bool gicv4_enable; -void vgic_v3_set_npie(struct kvm_vcpu *vcpu) -{ - struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3; - - cpuif->vgic_hcr |= ICH_HCR_NPIE; -} - void vgic_v3_set_underflow(struct kvm_vcpu *vcpu) { struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3; @@ -55,17 +48,23 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu) int lr; unsigned long flags; - cpuif->vgic_hcr &= ~(ICH_HCR_UIE | ICH_HCR_NPIE); + cpuif->vgic_hcr &= ~ICH_HCR_UIE; for (lr = 0; lr < vgic_cpu->used_lrs; lr++) { u64 val = cpuif->vgic_lr[lr]; - u32 intid; + u32 intid, cpuid; struct vgic_irq *irq; + bool is_v2_sgi = false; - if (model == KVM_DEV_TYPE_ARM_VGIC_V3) + cpuid = val & GICH_LR_PHYSID_CPUID; + cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT; + + if (model == KVM_DEV_TYPE_ARM_VGIC_V3) { intid = val & ICH_LR_VIRTUAL_ID_MASK; - else + } else { intid = val & GICH_LR_VIRTUALID; + is_v2_sgi = vgic_irq_is_sgi(intid); + } /* Notify fds when the guest EOI'ed a level-triggered IRQ */ if (lr_signals_eoi_mi(val) && vgic_valid_spi(vcpu->kvm, intid)) @@ -81,18 +80,16 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu) /* Always preserve the active bit */ irq->active = !!(val & ICH_LR_ACTIVE_BIT); + if (irq->active && is_v2_sgi) + irq->active_source = cpuid; + /* Edge is the only case where we preserve the pending bit */ if (irq->config == VGIC_CONFIG_EDGE && (val & ICH_LR_PENDING_BIT)) { irq->pending_latch = true; - if (vgic_irq_is_sgi(intid) && - model == KVM_DEV_TYPE_ARM_VGIC_V2) { - u32 cpuid = val & GICH_LR_PHYSID_CPUID; - - cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT; + if (is_v2_sgi) irq->source |= (1 << cpuid); - } } /* @@ -133,10 +130,20 @@ void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr) { u32 model = vcpu->kvm->arch.vgic.vgic_model; u64 val = irq->intid; - bool allow_pending = true; + bool allow_pending = true, is_v2_sgi; - if (irq->active) + is_v2_sgi = (vgic_irq_is_sgi(irq->intid) && + model == KVM_DEV_TYPE_ARM_VGIC_V2); + + if (irq->active) { val |= ICH_LR_ACTIVE_BIT; + if (is_v2_sgi) + val |= irq->active_source << GICH_LR_PHYSID_CPUID_SHIFT; + if (vgic_irq_is_multi_sgi(irq)) { + allow_pending = false; + val |= ICH_LR_EOI; + } + } if (irq->hw) { val |= ICH_LR_HW; @@ -174,8 +181,10 @@ void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr) BUG_ON(!src); val |= (src - 1) << GICH_LR_PHYSID_CPUID_SHIFT; irq->source &= ~(1 << (src - 1)); - if (irq->source) + if (irq->source) { irq->pending_latch = true; + val |= ICH_LR_EOI; + } } } diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c index 4b6d72939c42..568c65f852e1 100644 --- a/virt/kvm/arm/vgic/vgic.c +++ b/virt/kvm/arm/vgic/vgic.c @@ -719,14 +719,6 @@ static inline void vgic_set_underflow(struct kvm_vcpu *vcpu) vgic_v3_set_underflow(vcpu); } -static inline void vgic_set_npie(struct kvm_vcpu *vcpu) -{ - if (kvm_vgic_global_state.type == VGIC_V2) - vgic_v2_set_npie(vcpu); - else - vgic_v3_set_npie(vcpu); -} - /* Requires the ap_list_lock to be held. */ static int compute_ap_list_depth(struct kvm_vcpu *vcpu, bool *multi_sgi) @@ -740,17 +732,15 @@ static int compute_ap_list_depth(struct kvm_vcpu *vcpu, DEBUG_SPINLOCK_BUG_ON(!spin_is_locked(&vgic_cpu->ap_list_lock)); list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list) { + int w; + spin_lock(&irq->irq_lock); /* GICv2 SGIs can count for more than one... */ - if (vgic_irq_is_sgi(irq->intid) && irq->source) { - int w = hweight8(irq->source); - - count += w; - *multi_sgi |= (w > 1); - } else { - count++; - } + w = vgic_irq_get_lr_count(irq); spin_unlock(&irq->irq_lock); + + count += w; + *multi_sgi |= (w > 1); } return count; } @@ -761,7 +751,6 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu) struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu; struct vgic_irq *irq; int count; - bool npie = false; bool multi_sgi; u8 prio = 0xff; @@ -791,10 +780,8 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu) if (likely(vgic_target_oracle(irq) == vcpu)) { vgic_populate_lr(vcpu, irq, count++); - if (irq->source) { - npie = true; + if (irq->source) prio = irq->priority; - } } spin_unlock(&irq->irq_lock); @@ -807,9 +794,6 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu) } } - if (npie) - vgic_set_npie(vcpu); - vcpu->arch.vgic_cpu.used_lrs = count; /* Nuke remaining LRs */ diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h index 830e815748a0..32c25d42c93f 100644 --- a/virt/kvm/arm/vgic/vgic.h +++ b/virt/kvm/arm/vgic/vgic.h @@ -110,6 +110,20 @@ static inline bool vgic_irq_is_mapped_level(struct vgic_irq *irq) return irq->config == VGIC_CONFIG_LEVEL && irq->hw; } +static inline int vgic_irq_get_lr_count(struct vgic_irq *irq) +{ + /* Account for the active state as an interrupt */ + if (vgic_irq_is_sgi(irq->intid) && irq->source) + return hweight8(irq->source) + irq->active; + + return irq_is_pending(irq) || irq->active; +} + +static inline bool vgic_irq_is_multi_sgi(struct vgic_irq *irq) +{ + return vgic_irq_get_lr_count(irq) > 1; +} + /* * This struct provides an intermediate representation of the fields contained * in the GICH_VMCR and ICH_VMCR registers, such that code exporting the GIC -- cgit v1.2.3 From 98eb6cf25f0317395d9a799d18f3d46ba26a00d3 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 1 May 2018 11:37:19 -0400 Subject: sunrpc: Fix latency trace point crashes If the rpc_task survived longer than the transport, task->tk_xprt points to freed memory by the time rpc_count_iostats_metrics runs. Replace the references to task->tk_xprt with references to the task's tk_client. Reported-by: syzbot+27db1f90e2b972a5f2d3@syzkaller.appspotmail.com Fixes: 40bf7eb304b5 ('sunrpc: Add static trace point to report ...') Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 76887d60f0c0..7f1204a179b9 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -224,6 +224,8 @@ TRACE_EVENT(rpc_stats_latency, TP_ARGS(task, backlog, rtt, execute), TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) __field(u32, xid) __field(int, version) __string(progname, task->tk_client->cl_program->name) @@ -231,13 +233,11 @@ TRACE_EVENT(rpc_stats_latency, __field(unsigned long, backlog) __field(unsigned long, rtt) __field(unsigned long, execute) - __string(addr, - task->tk_xprt->address_strings[RPC_DISPLAY_ADDR]) - __string(port, - task->tk_xprt->address_strings[RPC_DISPLAY_PORT]) ), TP_fast_assign( + __entry->client_id = task->tk_client->cl_clid; + __entry->task_id = task->tk_pid; __entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid); __entry->version = task->tk_client->cl_vers; __assign_str(progname, task->tk_client->cl_program->name) @@ -245,14 +245,10 @@ TRACE_EVENT(rpc_stats_latency, __entry->backlog = ktime_to_us(backlog); __entry->rtt = ktime_to_us(rtt); __entry->execute = ktime_to_us(execute); - __assign_str(addr, - task->tk_xprt->address_strings[RPC_DISPLAY_ADDR]); - __assign_str(port, - task->tk_xprt->address_strings[RPC_DISPLAY_PORT]); ), - TP_printk("peer=[%s]:%s xid=0x%08x %sv%d %s backlog=%lu rtt=%lu execute=%lu", - __get_str(addr), __get_str(port), __entry->xid, + TP_printk("task:%u@%d xid=0x%08x %sv%d %s backlog=%lu rtt=%lu execute=%lu", + __entry->task_id, __entry->client_id, __entry->xid, __get_str(progname), __entry->version, __get_str(procname), __entry->backlog, __entry->rtt, __entry->execute) ); -- cgit v1.2.3 From de08481a253ac658433a8304a303ce9f018d71e5 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Fri, 27 Apr 2018 19:02:05 +0300 Subject: vhost: make msg padding explicit There's a 32 bit hole just after type. It's best to give it a name, this way compiler is forced to initialize it with rest of the structure. Reported-by: Kevin Easton Signed-off-by: Michael S. Tsirkin Signed-off-by: David S. Miller --- include/uapi/linux/vhost.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h index c51f8e5cc608..5a8ad064445b 100644 --- a/include/uapi/linux/vhost.h +++ b/include/uapi/linux/vhost.h @@ -68,6 +68,7 @@ struct vhost_iotlb_msg { struct vhost_msg { int type; + int padding0; union { struct vhost_iotlb_msg iotlb; __u8 padding[64]; -- cgit v1.2.3 From edd7ceb78296fb1574958991b6655c3c2cedf124 Mon Sep 17 00:00:00 2001 From: Thomas Winter Date: Tue, 1 May 2018 09:15:29 +1200 Subject: ipv6: Allow non-gateway ECMP for IPv6 It is valid to have static routes where the nexthop is an interface not an address such as tunnels. For IPv4 it was possible to use ECMP on these routes but not for IPv6. Signed-off-by: Thomas Winter Cc: David Ahern Cc: "David S. Miller" Cc: Alexey Kuznetsov Cc: Hideaki YOSHIFUJI Acked-by: David Ahern Signed-off-by: David S. Miller --- include/net/ip6_route.h | 3 +-- net/ipv6/ip6_fib.c | 3 --- 2 files changed, 1 insertion(+), 5 deletions(-) (limited to 'include') diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h index 08b132381984..abceb5864d99 100644 --- a/include/net/ip6_route.h +++ b/include/net/ip6_route.h @@ -68,8 +68,7 @@ static inline bool rt6_need_strict(const struct in6_addr *daddr) static inline bool rt6_qualify_for_ecmp(const struct rt6_info *rt) { - return (rt->rt6i_flags & (RTF_GATEWAY|RTF_ADDRCONF|RTF_DYNAMIC)) == - RTF_GATEWAY; + return (rt->rt6i_flags & (RTF_ADDRCONF | RTF_DYNAMIC)) == 0; } void ip6_route_input(struct sk_buff *skb); diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index deab2db6692e..3c97c29d4401 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -934,9 +934,6 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt, * list. * Only static routes (which don't have flag * RTF_EXPIRES) are used for ECMPv6. - * - * To avoid long list, we only had siblings if the - * route have a gateway. */ if (rt_can_ecmp && rt6_qualify_for_ecmp(iter)) -- cgit v1.2.3 From c212d2c7fc4736d49be102fb7a1a545cdc2f1fea Mon Sep 17 00:00:00 2001 From: Dave Watson Date: Tue, 1 May 2018 13:05:39 -0700 Subject: net/tls: Don't recursively call push_record during tls_write_space callbacks It is reported that in some cases, write_space may be called in do_tcp_sendpages, such that we recursively invoke do_tcp_sendpages again: [ 660.468802] ? do_tcp_sendpages+0x8d/0x580 [ 660.468826] ? tls_push_sg+0x74/0x130 [tls] [ 660.468852] ? tls_push_record+0x24a/0x390 [tls] [ 660.468880] ? tls_write_space+0x6a/0x80 [tls] ... tls_push_sg already does a loop over all sending sg's, so ignore any tls_write_space notifications until we are done sending. We then have to call the previous write_space to wake up poll() waiters after we are done with the send loop. Reported-by: Andre Tomt Signed-off-by: Dave Watson Signed-off-by: David S. Miller --- include/net/tls.h | 1 + net/tls/tls_main.c | 7 +++++++ 2 files changed, 8 insertions(+) (limited to 'include') diff --git a/include/net/tls.h b/include/net/tls.h index 3da8e13a6d96..b400d0bb7448 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -148,6 +148,7 @@ struct tls_context { struct scatterlist *partially_sent_record; u16 partially_sent_offset; unsigned long flags; + bool in_tcp_sendpages; u16 pending_open_record_frags; int (*push_pending_record)(struct sock *sk, int flags); diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c index 0d379970960e..cc03e00785c7 100644 --- a/net/tls/tls_main.c +++ b/net/tls/tls_main.c @@ -114,6 +114,7 @@ int tls_push_sg(struct sock *sk, size = sg->length - offset; offset += sg->offset; + ctx->in_tcp_sendpages = true; while (1) { if (sg_is_last(sg)) sendpage_flags = flags; @@ -148,6 +149,8 @@ retry: } clear_bit(TLS_PENDING_CLOSED_RECORD, &ctx->flags); + ctx->in_tcp_sendpages = false; + ctx->sk_write_space(sk); return 0; } @@ -217,6 +220,10 @@ static void tls_write_space(struct sock *sk) { struct tls_context *ctx = tls_get_ctx(sk); + /* We are already sending pages, ignore notification */ + if (ctx->in_tcp_sendpages) + return; + if (!sk->sk_write_pending && tls_is_pending_closed_record(ctx)) { gfp_t sk_allocation = sk->sk_allocation; int rc; -- cgit v1.2.3 From c818aa88d2d0cfc4938bfa9e226c0792af2dc45f Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Wed, 2 May 2018 17:19:05 +0300 Subject: Revert "vhost: make msg padding explicit" This reverts commit 93c0d549c4c5a7382ad70de6b86610b7aae57406. Unfortunately the padding will break 32 bit userspace. Ouch. Need to add some compat code, revert for now. Signed-off-by: Michael S. Tsirkin Signed-off-by: David S. Miller --- include/uapi/linux/vhost.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h index 5a8ad064445b..c51f8e5cc608 100644 --- a/include/uapi/linux/vhost.h +++ b/include/uapi/linux/vhost.h @@ -68,7 +68,6 @@ struct vhost_iotlb_msg { struct vhost_msg { int type; - int padding0; union { struct vhost_iotlb_msg iotlb; __u8 padding[64]; -- cgit v1.2.3 From 30ca22e4a5d0063dd9a9cdf35cd139c5807cbeb3 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Wed, 2 May 2018 22:41:56 +0300 Subject: ipv6: Revert "ipv6: Allow non-gateway ECMP for IPv6" This reverts commit edd7ceb78296 ("ipv6: Allow non-gateway ECMP for IPv6"). Eric reported a division by zero in rt6_multipath_rebalance() which is caused by above commit that considers identical local routes to be siblings. The division by zero happens because a nexthop weight is not set for local routes. Revert the commit as it does not fix a bug and has side effects. To reproduce: # ip -6 address add 2001:db8::1/64 dev dummy0 # ip -6 address add 2001:db8::1/64 dev dummy1 Fixes: edd7ceb78296 ("ipv6: Allow non-gateway ECMP for IPv6") Signed-off-by: Ido Schimmel Reported-by: Eric Dumazet Tested-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/ip6_route.h | 3 ++- net/ipv6/ip6_fib.c | 3 +++ 2 files changed, 5 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h index abceb5864d99..08b132381984 100644 --- a/include/net/ip6_route.h +++ b/include/net/ip6_route.h @@ -68,7 +68,8 @@ static inline bool rt6_need_strict(const struct in6_addr *daddr) static inline bool rt6_qualify_for_ecmp(const struct rt6_info *rt) { - return (rt->rt6i_flags & (RTF_ADDRCONF | RTF_DYNAMIC)) == 0; + return (rt->rt6i_flags & (RTF_GATEWAY|RTF_ADDRCONF|RTF_DYNAMIC)) == + RTF_GATEWAY; } void ip6_route_input(struct sk_buff *skb); diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index 3c97c29d4401..deab2db6692e 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -934,6 +934,9 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt, * list. * Only static routes (which don't have flag * RTF_EXPIRES) are used for ECMPv6. + * + * To avoid long list, we only had siblings if the + * route have a gateway. */ if (rt_can_ecmp && rt6_qualify_for_ecmp(iter)) -- cgit v1.2.3 From 85f1abe0019fcb3ea10df7029056cf42702283a8 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 1 May 2018 18:14:45 +0200 Subject: kthread, sched/wait: Fix kthread_parkme() completion issue Even with the wait-loop fixed, there is a further issue with kthread_parkme(). Upon hotplug, when we do takedown_cpu(), smpboot_park_threads() can return before all those threads are in fact blocked, due to the placement of the complete() in __kthread_parkme(). When that happens, sched_cpu_dying() -> migrate_tasks() can end up migrating such a still runnable task onto another CPU. Normally the task will have hit schedule() and gone to sleep by the time we do kthread_unpark(), which will then do __kthread_bind() to re-bind the task to the correct CPU. However, when we loose the initial TASK_PARKED store to the concurrent wakeup issue described previously, do the complete(), get migrated, it is possible to either: - observe kthread_unpark()'s clearing of SHOULD_PARK and terminate the park and set TASK_RUNNING, or - __kthread_bind()'s wait_task_inactive() to observe the competing TASK_RUNNING store. Either way the WARN() in __kthread_bind() will trigger and fail to correctly set the CPU affinity. Fix this by only issuing the complete() when the kthread has scheduled out. This does away with all the icky 'still running' nonsense. The alternative is to promote TASK_PARKED to a special state, this guarantees wait_task_inactive() cannot observe a 'stale' TASK_RUNNING and we'll end up doing the right thing, but this preserves the whole icky business of potentially migating the still runnable thing. Reported-by: Gaurav Kohli Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Thomas Gleixner Signed-off-by: Ingo Molnar --- include/linux/kthread.h | 1 + kernel/kthread.c | 43 +++++++++++++++++++------------------------ kernel/sched/core.c | 32 +++++++++++++++++++++----------- 3 files changed, 41 insertions(+), 35 deletions(-) (limited to 'include') diff --git a/include/linux/kthread.h b/include/linux/kthread.h index c1961761311d..2803264c512f 100644 --- a/include/linux/kthread.h +++ b/include/linux/kthread.h @@ -62,6 +62,7 @@ void *kthread_probe_data(struct task_struct *k); int kthread_park(struct task_struct *k); void kthread_unpark(struct task_struct *k); void kthread_parkme(void); +void kthread_park_complete(struct task_struct *k); int kthreadd(void *unused); extern struct task_struct *kthreadd_task; diff --git a/kernel/kthread.c b/kernel/kthread.c index cbee858e5815..2017a39ab490 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -55,7 +55,6 @@ enum KTHREAD_BITS { KTHREAD_IS_PER_CPU = 0, KTHREAD_SHOULD_STOP, KTHREAD_SHOULD_PARK, - KTHREAD_IS_PARKED, }; static inline void set_kthread_struct(void *kthread) @@ -181,11 +180,8 @@ static void __kthread_parkme(struct kthread *self) set_current_state(TASK_PARKED); if (!test_bit(KTHREAD_SHOULD_PARK, &self->flags)) break; - if (!test_and_set_bit(KTHREAD_IS_PARKED, &self->flags)) - complete(&self->parked); schedule(); } - clear_bit(KTHREAD_IS_PARKED, &self->flags); __set_current_state(TASK_RUNNING); } @@ -195,6 +191,11 @@ void kthread_parkme(void) } EXPORT_SYMBOL_GPL(kthread_parkme); +void kthread_park_complete(struct task_struct *k) +{ + complete(&to_kthread(k)->parked); +} + static int kthread(void *_create) { /* Copy data: it's on kthread's stack */ @@ -451,22 +452,15 @@ void kthread_unpark(struct task_struct *k) { struct kthread *kthread = to_kthread(k); - clear_bit(KTHREAD_SHOULD_PARK, &kthread->flags); /* - * We clear the IS_PARKED bit here as we don't wait - * until the task has left the park code. So if we'd - * park before that happens we'd see the IS_PARKED bit - * which might be about to be cleared. + * Newly created kthread was parked when the CPU was offline. + * The binding was lost and we need to set it again. */ - if (test_and_clear_bit(KTHREAD_IS_PARKED, &kthread->flags)) { - /* - * Newly created kthread was parked when the CPU was offline. - * The binding was lost and we need to set it again. - */ - if (test_bit(KTHREAD_IS_PER_CPU, &kthread->flags)) - __kthread_bind(k, kthread->cpu, TASK_PARKED); - wake_up_state(k, TASK_PARKED); - } + if (test_bit(KTHREAD_IS_PER_CPU, &kthread->flags)) + __kthread_bind(k, kthread->cpu, TASK_PARKED); + + clear_bit(KTHREAD_SHOULD_PARK, &kthread->flags); + wake_up_state(k, TASK_PARKED); } EXPORT_SYMBOL_GPL(kthread_unpark); @@ -489,12 +483,13 @@ int kthread_park(struct task_struct *k) if (WARN_ON(k->flags & PF_EXITING)) return -ENOSYS; - if (!test_bit(KTHREAD_IS_PARKED, &kthread->flags)) { - set_bit(KTHREAD_SHOULD_PARK, &kthread->flags); - if (k != current) { - wake_up_process(k); - wait_for_completion(&kthread->parked); - } + if (WARN_ON_ONCE(test_bit(KTHREAD_SHOULD_PARK, &kthread->flags))) + return -EBUSY; + + set_bit(KTHREAD_SHOULD_PARK, &kthread->flags); + if (k != current) { + wake_up_process(k); + wait_for_completion(&kthread->parked); } return 0; diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 5e10aaeebfcc..7ad60e00a6a8 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -7,6 +7,8 @@ */ #include "sched.h" +#include + #include #include @@ -2718,20 +2720,28 @@ static struct rq *finish_task_switch(struct task_struct *prev) membarrier_mm_sync_core_before_usermode(mm); mmdrop(mm); } - if (unlikely(prev_state == TASK_DEAD)) { - if (prev->sched_class->task_dead) - prev->sched_class->task_dead(prev); + if (unlikely(prev_state & (TASK_DEAD|TASK_PARKED))) { + switch (prev_state) { + case TASK_DEAD: + if (prev->sched_class->task_dead) + prev->sched_class->task_dead(prev); - /* - * Remove function-return probe instances associated with this - * task and put them back on the free list. - */ - kprobe_flush_task(prev); + /* + * Remove function-return probe instances associated with this + * task and put them back on the free list. + */ + kprobe_flush_task(prev); - /* Task is done with its stack. */ - put_task_stack(prev); + /* Task is done with its stack. */ + put_task_stack(prev); - put_task_struct(prev); + put_task_struct(prev); + break; + + case TASK_PARKED: + kthread_park_complete(prev); + break; + } } tick_nohz_task_switch(); -- cgit v1.2.3 From 8236b0ae31c837d2b3a2565c5f8d77f637e824cc Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Wed, 2 May 2018 07:07:55 +0900 Subject: bdi: wake up concurrent wb_shutdown() callers. syzbot is reporting hung tasks at wait_on_bit(WB_shutting_down) in wb_shutdown() [1]. This seems to be because commit 5318ce7d46866e1d ("bdi: Shutdown writeback on all cgwbs in cgwb_bdi_destroy()") forgot to call wake_up_bit(WB_shutting_down) after clear_bit(WB_shutting_down). Introduce a helper function clear_and_wake_up_bit() and use it, in order to avoid similar errors in future. [1] https://syzkaller.appspot.com/bug?id=b297474817af98d5796bc544e1bb806fc3da0e5e Signed-off-by: Tetsuo Handa Reported-by: syzbot Fixes: 5318ce7d46866e1d ("bdi: Shutdown writeback on all cgwbs in cgwb_bdi_destroy()") Cc: Tejun Heo Reviewed-by: Jan Kara Suggested-by: Linus Torvalds Signed-off-by: Jens Axboe --- include/linux/wait_bit.h | 17 +++++++++++++++++ mm/backing-dev.c | 2 +- 2 files changed, 18 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/wait_bit.h b/include/linux/wait_bit.h index 9318b2166439..2b0072fa5e92 100644 --- a/include/linux/wait_bit.h +++ b/include/linux/wait_bit.h @@ -305,4 +305,21 @@ do { \ __ret; \ }) +/** + * clear_and_wake_up_bit - clear a bit and wake up anyone waiting on that bit + * + * @bit: the bit of the word being waited on + * @word: the word being waited on, a kernel virtual address + * + * You can use this helper if bitflags are manipulated atomically rather than + * non-atomically under a lock. + */ +static inline void clear_and_wake_up_bit(int bit, void *word) +{ + clear_bit_unlock(bit, word); + /* See wake_up_bit() for which memory barrier you need to use. */ + smp_mb__after_atomic(); + wake_up_bit(word, bit); +} + #endif /* _LINUX_WAIT_BIT_H */ diff --git a/mm/backing-dev.c b/mm/backing-dev.c index 023190c69dce..fa5e6d7406d1 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -383,7 +383,7 @@ static void wb_shutdown(struct bdi_writeback *wb) * the barrier provided by test_and_clear_bit() above. */ smp_wmb(); - clear_bit(WB_shutting_down, &wb->state); + clear_and_wake_up_bit(WB_shutting_down, &wb->state); } static void wb_exit(struct bdi_writeback *wb) -- cgit v1.2.3 From b5bf9a90bbebffba888c9144c5a8a10317b04064 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 30 Apr 2018 14:51:01 +0200 Subject: sched/core: Introduce set_special_state() Gaurav reported a perceived problem with TASK_PARKED, which turned out to be a broken wait-loop pattern in __kthread_parkme(), but the reported issue can (and does) in fact happen for states that do not do condition based sleeps. When the 'current->state = TASK_RUNNING' store of a previous (concurrent) try_to_wake_up() collides with the setting of a 'special' sleep state, we can loose the sleep state. Normal condition based wait-loops are immune to this problem, but for sleep states that are not condition based are subject to this problem. There already is a fix for TASK_DEAD. Abstract that and also apply it to TASK_STOPPED and TASK_TRACED, both of which are also without condition based wait-loop. Reported-by: Gaurav Kohli Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Oleg Nesterov Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Signed-off-by: Ingo Molnar --- include/linux/sched.h | 50 +++++++++++++++++++++++++++++++++++++++----- include/linux/sched/signal.h | 2 +- kernel/sched/core.c | 17 +-------------- kernel/signal.c | 17 +++++++++++++-- 4 files changed, 62 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/include/linux/sched.h b/include/linux/sched.h index b3d697f3b573..c2413703f45d 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -112,17 +112,36 @@ struct task_group; #ifdef CONFIG_DEBUG_ATOMIC_SLEEP +/* + * Special states are those that do not use the normal wait-loop pattern. See + * the comment with set_special_state(). + */ +#define is_special_task_state(state) \ + ((state) & (__TASK_STOPPED | __TASK_TRACED | TASK_DEAD)) + #define __set_current_state(state_value) \ do { \ + WARN_ON_ONCE(is_special_task_state(state_value));\ current->task_state_change = _THIS_IP_; \ current->state = (state_value); \ } while (0) + #define set_current_state(state_value) \ do { \ + WARN_ON_ONCE(is_special_task_state(state_value));\ current->task_state_change = _THIS_IP_; \ smp_store_mb(current->state, (state_value)); \ } while (0) +#define set_special_state(state_value) \ + do { \ + unsigned long flags; /* may shadow */ \ + WARN_ON_ONCE(!is_special_task_state(state_value)); \ + raw_spin_lock_irqsave(¤t->pi_lock, flags); \ + current->task_state_change = _THIS_IP_; \ + current->state = (state_value); \ + raw_spin_unlock_irqrestore(¤t->pi_lock, flags); \ + } while (0) #else /* * set_current_state() includes a barrier so that the write of current->state @@ -144,8 +163,8 @@ struct task_group; * * The above is typically ordered against the wakeup, which does: * - * need_sleep = false; - * wake_up_state(p, TASK_UNINTERRUPTIBLE); + * need_sleep = false; + * wake_up_state(p, TASK_UNINTERRUPTIBLE); * * Where wake_up_state() (and all other wakeup primitives) imply enough * barriers to order the store of the variable against wakeup. @@ -154,12 +173,33 @@ struct task_group; * once it observes the TASK_UNINTERRUPTIBLE store the waking CPU can issue a * TASK_RUNNING store which can collide with __set_current_state(TASK_RUNNING). * - * This is obviously fine, since they both store the exact same value. + * However, with slightly different timing the wakeup TASK_RUNNING store can + * also collide with the TASK_UNINTERRUPTIBLE store. Loosing that store is not + * a problem either because that will result in one extra go around the loop + * and our @cond test will save the day. * * Also see the comments of try_to_wake_up(). */ -#define __set_current_state(state_value) do { current->state = (state_value); } while (0) -#define set_current_state(state_value) smp_store_mb(current->state, (state_value)) +#define __set_current_state(state_value) \ + current->state = (state_value) + +#define set_current_state(state_value) \ + smp_store_mb(current->state, (state_value)) + +/* + * set_special_state() should be used for those states when the blocking task + * can not use the regular condition based wait-loop. In that case we must + * serialize against wakeups such that any possible in-flight TASK_RUNNING stores + * will not collide with our state change. + */ +#define set_special_state(state_value) \ + do { \ + unsigned long flags; /* may shadow */ \ + raw_spin_lock_irqsave(¤t->pi_lock, flags); \ + current->state = (state_value); \ + raw_spin_unlock_irqrestore(¤t->pi_lock, flags); \ + } while (0) + #endif /* Task command name length: */ diff --git a/include/linux/sched/signal.h b/include/linux/sched/signal.h index a7ce74c74e49..113d1ad1ced7 100644 --- a/include/linux/sched/signal.h +++ b/include/linux/sched/signal.h @@ -280,7 +280,7 @@ static inline void kernel_signal_stop(void) { spin_lock_irq(¤t->sighand->siglock); if (current->jobctl & JOBCTL_STOP_DEQUEUED) - __set_current_state(TASK_STOPPED); + set_special_state(TASK_STOPPED); spin_unlock_irq(¤t->sighand->siglock); schedule(); diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 7ad60e00a6a8..ffde9eebc846 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -3508,23 +3508,8 @@ static void __sched notrace __schedule(bool preempt) void __noreturn do_task_dead(void) { - /* - * The setting of TASK_RUNNING by try_to_wake_up() may be delayed - * when the following two conditions become true. - * - There is race condition of mmap_sem (It is acquired by - * exit_mm()), and - * - SMI occurs before setting TASK_RUNINNG. - * (or hypervisor of virtual machine switches to other guest) - * As a result, we may become TASK_RUNNING after becoming TASK_DEAD - * - * To avoid it, we have to wait for releasing tsk->pi_lock which - * is held by try_to_wake_up() - */ - raw_spin_lock_irq(¤t->pi_lock); - raw_spin_unlock_irq(¤t->pi_lock); - /* Causes final put_task_struct in finish_task_switch(): */ - __set_current_state(TASK_DEAD); + set_special_state(TASK_DEAD); /* Tell freezer to ignore us: */ current->flags |= PF_NOFREEZE; diff --git a/kernel/signal.c b/kernel/signal.c index d4ccea599692..9c33163a6165 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1961,14 +1961,27 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) return; } + set_special_state(TASK_TRACED); + /* * We're committing to trapping. TRACED should be visible before * TRAPPING is cleared; otherwise, the tracer might fail do_wait(). * Also, transition to TRACED and updates to ->jobctl should be * atomic with respect to siglock and should be done after the arch * hook as siglock is released and regrabbed across it. + * + * TRACER TRACEE + * + * ptrace_attach() + * [L] wait_on_bit(JOBCTL_TRAPPING) [S] set_special_state(TRACED) + * do_wait() + * set_current_state() smp_wmb(); + * ptrace_do_wait() + * wait_task_stopped() + * task_stopped_code() + * [L] task_is_traced() [S] task_clear_jobctl_trapping(); */ - set_current_state(TASK_TRACED); + smp_wmb(); current->last_siginfo = info; current->exit_code = exit_code; @@ -2176,7 +2189,7 @@ static bool do_signal_stop(int signr) if (task_participate_group_stop(current)) notify = CLD_STOPPED; - __set_current_state(TASK_STOPPED); + set_special_state(TASK_STOPPED); spin_unlock_irq(¤t->sighand->siglock); /* -- cgit v1.2.3 From 3259081991a9398434f6f49468b960f136ac0158 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Wed, 25 Apr 2018 05:34:48 -0400 Subject: MAINTAINERS & files: Canonize the e-mails I use at files From now on, I'll start using my @kernel.org as my development e-mail. As such, let's remove the entries that point to the old mchehab@s-opensource.com at MAINTAINERS file. For the files written with a copyright with mchehab@s-opensource, let's keep Samsung on their names, using mchehab+samsung@kernel.org, in order to keep pointing to my employer, with sponsors the work. For the files written before I join Samsung (on July, 4 2013), let's just use mchehab@kernel.org. For bug reports, we can simply point to just kernel.org, as this will reach my mchehab+samsung inbox anyway. Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Brian Warner Signed-off-by: Mauro Carvalho Chehab --- Documentation/doc-guide/parse-headers.rst | 4 ++-- Documentation/media/uapi/rc/keytable.c.rst | 2 +- Documentation/media/uapi/v4l/v4l2grab.c.rst | 2 +- Documentation/sphinx/parse-headers.pl | 4 ++-- .../translations/zh_CN/video4linux/v4l2-framework.txt | 4 ++-- MAINTAINERS | 17 ----------------- drivers/media/i2c/saa7115.c | 2 +- drivers/media/i2c/saa711x_regs.h | 2 +- drivers/media/i2c/tda7432.c | 2 +- drivers/media/i2c/tvp5150.c | 2 +- drivers/media/i2c/tvp5150_reg.h | 2 +- drivers/media/i2c/tvp7002.c | 2 +- drivers/media/i2c/tvp7002_reg.h | 2 +- drivers/media/media-devnode.c | 2 +- drivers/media/pci/bt8xx/bttv-audio-hook.c | 2 +- drivers/media/pci/bt8xx/bttv-audio-hook.h | 2 +- drivers/media/pci/bt8xx/bttv-cards.c | 4 ++-- drivers/media/pci/bt8xx/bttv-driver.c | 2 +- drivers/media/pci/bt8xx/bttv-i2c.c | 2 +- drivers/media/pci/cx23885/cx23885-input.c | 2 +- drivers/media/pci/cx88/cx88-alsa.c | 4 ++-- drivers/media/pci/cx88/cx88-blackbird.c | 2 +- drivers/media/pci/cx88/cx88-core.c | 2 +- drivers/media/pci/cx88/cx88-i2c.c | 2 +- drivers/media/pci/cx88/cx88-video.c | 2 +- drivers/media/radio/radio-aimslab.c | 2 +- drivers/media/radio/radio-aztech.c | 2 +- drivers/media/radio/radio-gemtek.c | 2 +- drivers/media/radio/radio-maxiradio.c | 2 +- drivers/media/radio/radio-rtrack2.c | 2 +- drivers/media/radio/radio-sf16fmi.c | 2 +- drivers/media/radio/radio-terratec.c | 2 +- drivers/media/radio/radio-trust.c | 2 +- drivers/media/radio/radio-typhoon.c | 2 +- drivers/media/radio/radio-zoltrix.c | 2 +- drivers/media/rc/keymaps/rc-avermedia-m135a.c | 2 +- drivers/media/rc/keymaps/rc-encore-enltv-fm53.c | 2 +- drivers/media/rc/keymaps/rc-encore-enltv2.c | 2 +- drivers/media/rc/keymaps/rc-kaiomy.c | 2 +- drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c | 2 +- drivers/media/rc/keymaps/rc-pixelview-new.c | 2 +- drivers/media/tuners/tea5761.c | 4 ++-- drivers/media/tuners/tea5767.c | 4 ++-- drivers/media/tuners/tuner-xc2028-types.h | 2 +- drivers/media/tuners/tuner-xc2028.c | 4 ++-- drivers/media/tuners/tuner-xc2028.h | 2 +- drivers/media/usb/em28xx/em28xx-camera.c | 2 +- drivers/media/usb/em28xx/em28xx-cards.c | 2 +- drivers/media/usb/em28xx/em28xx-core.c | 4 ++-- drivers/media/usb/em28xx/em28xx-dvb.c | 4 ++-- drivers/media/usb/em28xx/em28xx-i2c.c | 2 +- drivers/media/usb/em28xx/em28xx-input.c | 2 +- drivers/media/usb/em28xx/em28xx-video.c | 4 ++-- drivers/media/usb/em28xx/em28xx.h | 2 +- drivers/media/usb/gspca/zc3xx-reg.h | 2 +- drivers/media/usb/tm6000/tm6000-cards.c | 2 +- drivers/media/usb/tm6000/tm6000-core.c | 2 +- drivers/media/usb/tm6000/tm6000-i2c.c | 2 +- drivers/media/usb/tm6000/tm6000-regs.h | 2 +- drivers/media/usb/tm6000/tm6000-usb-isoc.h | 2 +- drivers/media/usb/tm6000/tm6000-video.c | 2 +- drivers/media/usb/tm6000/tm6000.h | 2 +- drivers/media/v4l2-core/v4l2-dev.c | 4 ++-- drivers/media/v4l2-core/v4l2-ioctl.c | 2 +- drivers/media/v4l2-core/videobuf-core.c | 6 +++--- drivers/media/v4l2-core/videobuf-dma-contig.c | 2 +- drivers/media/v4l2-core/videobuf-dma-sg.c | 6 +++--- drivers/media/v4l2-core/videobuf-vmalloc.c | 4 ++-- include/media/i2c/tvp7002.h | 2 +- include/media/videobuf-core.h | 4 ++-- include/media/videobuf-dma-sg.h | 4 ++-- include/media/videobuf-vmalloc.h | 2 +- scripts/extract_xc3028.pl | 2 +- scripts/split-man.pl | 2 +- 74 files changed, 92 insertions(+), 109 deletions(-) (limited to 'include') diff --git a/Documentation/doc-guide/parse-headers.rst b/Documentation/doc-guide/parse-headers.rst index 96a0423d5dba..82a3e43b6864 100644 --- a/Documentation/doc-guide/parse-headers.rst +++ b/Documentation/doc-guide/parse-headers.rst @@ -177,14 +177,14 @@ BUGS **** -Report bugs to Mauro Carvalho Chehab +Report bugs to Mauro Carvalho Chehab COPYRIGHT ********* -Copyright (c) 2016 by Mauro Carvalho Chehab . +Copyright (c) 2016 by Mauro Carvalho Chehab . License GPLv2: GNU GPL version 2 . diff --git a/Documentation/media/uapi/rc/keytable.c.rst b/Documentation/media/uapi/rc/keytable.c.rst index e6ce1e3f5a78..217237f93b37 100644 --- a/Documentation/media/uapi/rc/keytable.c.rst +++ b/Documentation/media/uapi/rc/keytable.c.rst @@ -7,7 +7,7 @@ file: uapi/v4l/keytable.c /* keytable.c - This program allows checking/replacing keys at IR - Copyright (C) 2006-2009 Mauro Carvalho Chehab + Copyright (C) 2006-2009 Mauro Carvalho Chehab This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by diff --git a/Documentation/media/uapi/v4l/v4l2grab.c.rst b/Documentation/media/uapi/v4l/v4l2grab.c.rst index 5aabd0b7b089..f0d0ab6abd41 100644 --- a/Documentation/media/uapi/v4l/v4l2grab.c.rst +++ b/Documentation/media/uapi/v4l/v4l2grab.c.rst @@ -6,7 +6,7 @@ file: media/v4l/v4l2grab.c .. code-block:: c /* V4L2 video picture grabber - Copyright (C) 2009 Mauro Carvalho Chehab + Copyright (C) 2009 Mauro Carvalho Chehab This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by diff --git a/Documentation/sphinx/parse-headers.pl b/Documentation/sphinx/parse-headers.pl index a958d8b5e99d..d410f47567e9 100755 --- a/Documentation/sphinx/parse-headers.pl +++ b/Documentation/sphinx/parse-headers.pl @@ -387,11 +387,11 @@ tree for more details. =head1 BUGS -Report bugs to Mauro Carvalho Chehab +Report bugs to Mauro Carvalho Chehab =head1 COPYRIGHT -Copyright (c) 2016 by Mauro Carvalho Chehab . +Copyright (c) 2016 by Mauro Carvalho Chehab . License GPLv2: GNU GPL version 2 . diff --git a/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt b/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt index 698660b7f21f..c77c0f060864 100644 --- a/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt +++ b/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt @@ -6,7 +6,7 @@ communicating in English you can also ask the Chinese maintainer for help. Contact the Chinese maintainer if this translation is outdated or if there is a problem with the translation. -Maintainer: Mauro Carvalho Chehab +Maintainer: Mauro Carvalho Chehab Chinese maintainer: Fu Wei --------------------------------------------------------------------- Documentation/video4linux/v4l2-framework.txt 的中文翻译 @@ -14,7 +14,7 @@ Documentation/video4linux/v4l2-framework.txt 的中文翻译 如果想评论或更新本文的内容，请直接联系原文档的维护者。如果你使用英文 交流有困难的话，也可以向中文版维护者求助。如果本翻译更新不及时或者翻 译存在问题，请联系中文版维护者。 -英文版维护者： Mauro Carvalho Chehab +英文版维护者： Mauro Carvalho Chehab 中文版维护者： 傅炜 Fu Wei 中文版翻译者： 傅炜 Fu Wei 中文版校译者： 傅炜 Fu Wei diff --git a/MAINTAINERS b/MAINTAINERS index 79bb02ff812f..64a315998d5d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2554,7 +2554,6 @@ F: Documentation/devicetree/bindings/sound/axentia,* F: sound/soc/atmel/tse850-pcm5142.c AZ6007 DVB DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -3083,7 +3082,6 @@ F: include/linux/btrfs* F: include/uapi/linux/btrfs* BTTV VIDEO4LINUX DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -3812,7 +3810,6 @@ S: Maintained F: drivers/media/dvb-frontends/cx24120* CX88 VIDEO4LINUX DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -5053,7 +5050,6 @@ F: drivers/edac/thunderx_edac* EDAC-CORE M: Borislav Petkov -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-edac@vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp.git for-next @@ -5082,7 +5078,6 @@ S: Maintained F: drivers/edac/fsl_ddr_edac.* EDAC-GHES -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-edac@vger.kernel.org S: Maintained @@ -5099,21 +5094,18 @@ S: Maintained F: drivers/edac/i5000_edac.c EDAC-I5400 -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-edac@vger.kernel.org S: Maintained F: drivers/edac/i5400_edac.c EDAC-I7300 -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-edac@vger.kernel.org S: Maintained F: drivers/edac/i7300_edac.c EDAC-I7CORE -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-edac@vger.kernel.org S: Maintained @@ -5163,7 +5155,6 @@ S: Maintained F: drivers/edac/r82600_edac.c EDAC-SBRIDGE -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-edac@vger.kernel.org S: Maintained @@ -5222,7 +5213,6 @@ S: Maintained F: drivers/net/ethernet/ibm/ehea/ EM28XX VIDEO4LINUX DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -8871,7 +8861,6 @@ F: Documentation/devicetree/bindings/media/nvidia,tegra-vde.txt F: drivers/staging/media/tegra-vde/ MEDIA INPUT INFRASTRUCTURE (V4L/DVB) -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab P: LinuxTV.org Project L: linux-media@vger.kernel.org @@ -12259,7 +12248,6 @@ S: Odd Fixes F: drivers/media/i2c/saa6588* SAA7134 VIDEO4LINUX DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -12763,7 +12751,6 @@ S: Maintained F: drivers/media/radio/si4713/radio-usb-si4713.c SIANO DVB DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -13754,7 +13741,6 @@ S: Maintained F: drivers/media/i2c/tda9840* TEA5761 TUNER DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -13763,7 +13749,6 @@ S: Odd fixes F: drivers/media/tuners/tea5761.* TEA5767 TUNER DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -14180,7 +14165,6 @@ F: Documentation/networking/tlan.txt F: drivers/net/ethernet/ti/tlan.* TM6000 VIDEO4LINUX DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org @@ -15407,7 +15391,6 @@ S: Maintained F: arch/x86/entry/vdso/ XC2028/3028 TUNER DRIVER -M: Mauro Carvalho Chehab M: Mauro Carvalho Chehab L: linux-media@vger.kernel.org W: https://linuxtv.org diff --git a/drivers/media/i2c/saa7115.c b/drivers/media/i2c/saa7115.c index e216cd768409..b07114b5efb2 100644 --- a/drivers/media/i2c/saa7115.c +++ b/drivers/media/i2c/saa7115.c @@ -20,7 +20,7 @@ // // VBI support (2004) and cleanups (2005) by Hans Verkuil // -// Copyright (c) 2005-2006 Mauro Carvalho Chehab +// Copyright (c) 2005-2006 Mauro Carvalho Chehab // SAA7111, SAA7113 and SAA7118 support #include "saa711x_regs.h" diff --git a/drivers/media/i2c/saa711x_regs.h b/drivers/media/i2c/saa711x_regs.h index a50d480e101a..44fabe08234d 100644 --- a/drivers/media/i2c/saa711x_regs.h +++ b/drivers/media/i2c/saa711x_regs.h @@ -2,7 +2,7 @@ * SPDX-License-Identifier: GPL-2.0+ * saa711x - Philips SAA711x video decoder register specifications * - * Copyright (c) 2006 Mauro Carvalho Chehab + * Copyright (c) 2006 Mauro Carvalho Chehab */ #define R_00_CHIP_VERSION 0x00 diff --git a/drivers/media/i2c/tda7432.c b/drivers/media/i2c/tda7432.c index 1c5c61d829d6..9b4f21237810 100644 --- a/drivers/media/i2c/tda7432.c +++ b/drivers/media/i2c/tda7432.c @@ -8,7 +8,7 @@ * Muting and tone control by Jonathan Isom * * Copyright (c) 2000 Eric Sandeen - * Copyright (c) 2006 Mauro Carvalho Chehab + * Copyright (c) 2006 Mauro Carvalho Chehab * This code is placed under the terms of the GNU General Public License * Based on tda9855.c by Steve VanDeBogart (vandebo@uclink.berkeley.edu) * Which was based on tda8425.c by Greg Alexander (c) 1998 diff --git a/drivers/media/i2c/tvp5150.c b/drivers/media/i2c/tvp5150.c index 2476d812f669..1734ed4ede33 100644 --- a/drivers/media/i2c/tvp5150.c +++ b/drivers/media/i2c/tvp5150.c @@ -2,7 +2,7 @@ // // tvp5150 - Texas Instruments TVP5150A/AM1 and TVP5151 video decoder driver // -// Copyright (c) 2005,2006 Mauro Carvalho Chehab +// Copyright (c) 2005,2006 Mauro Carvalho Chehab #include #include diff --git a/drivers/media/i2c/tvp5150_reg.h b/drivers/media/i2c/tvp5150_reg.h index c43b7b844021..d3a764cae1a0 100644 --- a/drivers/media/i2c/tvp5150_reg.h +++ b/drivers/media/i2c/tvp5150_reg.h @@ -3,7 +3,7 @@ * * tvp5150 - Texas Instruments TVP5150A/AM1 video decoder registers * - * Copyright (c) 2005,2006 Mauro Carvalho Chehab + * Copyright (c) 2005,2006 Mauro Carvalho Chehab */ #define TVP5150_VD_IN_SRC_SEL_1 0x00 /* Video input source selection #1 */ diff --git a/drivers/media/i2c/tvp7002.c b/drivers/media/i2c/tvp7002.c index a26c1a3f7183..4599b7e28a8d 100644 --- a/drivers/media/i2c/tvp7002.c +++ b/drivers/media/i2c/tvp7002.c @@ -5,7 +5,7 @@ * Author: Santiago Nunez-Corrales * * This code is partially based upon the TVP5150 driver - * written by Mauro Carvalho Chehab (mchehab@infradead.org), + * written by Mauro Carvalho Chehab , * the TVP514x driver written by Vaibhav Hiremath * and the TVP7002 driver in the TI LSP 2.10.00.14. Revisions by * Muralidharan Karicheri and Snehaprabha Narnakaje (TI). diff --git a/drivers/media/i2c/tvp7002_reg.h b/drivers/media/i2c/tvp7002_reg.h index 3c8c8b0a6a4c..7f56ba689dfe 100644 --- a/drivers/media/i2c/tvp7002_reg.h +++ b/drivers/media/i2c/tvp7002_reg.h @@ -5,7 +5,7 @@ * Author: Santiago Nunez-Corrales * * This code is partially based upon the TVP5150 driver - * written by Mauro Carvalho Chehab (mchehab@infradead.org), + * written by Mauro Carvalho Chehab , * the TVP514x driver written by Vaibhav Hiremath * and the TVP7002 driver in the TI LSP 2.10.00.14 * diff --git a/drivers/media/media-devnode.c b/drivers/media/media-devnode.c index 67ac51eff15c..6b87a721dc49 100644 --- a/drivers/media/media-devnode.c +++ b/drivers/media/media-devnode.c @@ -4,7 +4,7 @@ * Copyright (C) 2010 Nokia Corporation * * Based on drivers/media/video/v4l2_dev.c code authored by - * Mauro Carvalho Chehab (version 2) + * Mauro Carvalho Chehab (version 2) * Alan Cox, (version 1) * * Contacts: Laurent Pinchart diff --git a/drivers/media/pci/bt8xx/bttv-audio-hook.c b/drivers/media/pci/bt8xx/bttv-audio-hook.c index 9f1f9169fb5b..346fc7f58839 100644 --- a/drivers/media/pci/bt8xx/bttv-audio-hook.c +++ b/drivers/media/pci/bt8xx/bttv-audio-hook.c @@ -1,7 +1,7 @@ /* * Handlers for board audio hooks, splitted from bttv-cards * - * Copyright (c) 2006 Mauro Carvalho Chehab (mchehab@infradead.org) + * Copyright (c) 2006 Mauro Carvalho Chehab * This code is placed under the terms of the GNU General Public License */ diff --git a/drivers/media/pci/bt8xx/bttv-audio-hook.h b/drivers/media/pci/bt8xx/bttv-audio-hook.h index 159d07adeff8..be16a537a03a 100644 --- a/drivers/media/pci/bt8xx/bttv-audio-hook.h +++ b/drivers/media/pci/bt8xx/bttv-audio-hook.h @@ -1,7 +1,7 @@ /* * Handlers for board audio hooks, splitted from bttv-cards * - * Copyright (c) 2006 Mauro Carvalho Chehab (mchehab@infradead.org) + * Copyright (c) 2006 Mauro Carvalho Chehab * This code is placed under the terms of the GNU General Public License */ diff --git a/drivers/media/pci/bt8xx/bttv-cards.c b/drivers/media/pci/bt8xx/bttv-cards.c index 1902732f90e1..2616243b2c49 100644 --- a/drivers/media/pci/bt8xx/bttv-cards.c +++ b/drivers/media/pci/bt8xx/bttv-cards.c @@ -2447,7 +2447,7 @@ struct tvcard bttv_tvcards[] = { }, /* ---- card 0x88---------------------------------- */ [BTTV_BOARD_ACORP_Y878F] = { - /* Mauro Carvalho Chehab */ + /* Mauro Carvalho Chehab */ .name = "Acorp Y878F", .video_inputs = 3, /* .audio_inputs= 1, */ @@ -2688,7 +2688,7 @@ struct tvcard bttv_tvcards[] = { }, [BTTV_BOARD_ENLTV_FM_2] = { /* Encore TV Tuner Pro ENL TV-FM-2 - Mauro Carvalho Chehab */ .name = "Encore ENL TV-FM-2", .video_inputs = 3, /* .audio_inputs= 1, */ diff --git a/drivers/media/pci/bt8xx/bttv-driver.c b/drivers/media/pci/bt8xx/bttv-driver.c index 707f57a9f940..de3f44b8dec6 100644 --- a/drivers/media/pci/bt8xx/bttv-driver.c +++ b/drivers/media/pci/bt8xx/bttv-driver.c @@ -13,7 +13,7 @@ (c) 2005-2006 Nickolay V. Shmyrev Fixes to be fully V4L2 compliant by - (c) 2006 Mauro Carvalho Chehab + (c) 2006 Mauro Carvalho Chehab Cropping and overscan support Copyright (C) 2005, 2006 Michael H. Schimek diff --git a/drivers/media/pci/bt8xx/bttv-i2c.c b/drivers/media/pci/bt8xx/bttv-i2c.c index eccd1e3d717a..c76823eb399d 100644 --- a/drivers/media/pci/bt8xx/bttv-i2c.c +++ b/drivers/media/pci/bt8xx/bttv-i2c.c @@ -8,7 +8,7 @@ & Marcus Metzler (mocm@thp.uni-koeln.de) (c) 1999-2003 Gerd Knorr - (c) 2005 Mauro Carvalho Chehab + (c) 2005 Mauro Carvalho Chehab - Multituner support and i2c address binding This program is free software; you can redistribute it and/or modify diff --git a/drivers/media/pci/cx23885/cx23885-input.c b/drivers/media/pci/cx23885/cx23885-input.c index be49589a61d2..395ff9bba759 100644 --- a/drivers/media/pci/cx23885/cx23885-input.c +++ b/drivers/media/pci/cx23885/cx23885-input.c @@ -13,7 +13,7 @@ * Copyright (C) 2008 * Copyright (C) 2005 Ludovico Cavedon * Markus Rechberger - * Mauro Carvalho Chehab + * Mauro Carvalho Chehab * Sascha Sommer * Copyright (C) 2004, 2005 Chris Pascoe * Copyright (C) 2003, 2004 Gerd Knorr diff --git a/drivers/media/pci/cx88/cx88-alsa.c b/drivers/media/pci/cx88/cx88-alsa.c index ab09bb55cf45..8a28fda703a2 100644 --- a/drivers/media/pci/cx88/cx88-alsa.c +++ b/drivers/media/pci/cx88/cx88-alsa.c @@ -4,7 +4,7 @@ * * (c) 2007 Trent Piepho * (c) 2005,2006 Ricardo Cerqueira - * (c) 2005 Mauro Carvalho Chehab + * (c) 2005 Mauro Carvalho Chehab * Based on a dummy cx88 module by Gerd Knorr * Based on dummy.c by Jaroslav Kysela * @@ -103,7 +103,7 @@ MODULE_PARM_DESC(index, "Index value for cx88x capture interface(s)."); MODULE_DESCRIPTION("ALSA driver module for cx2388x based TV cards"); MODULE_AUTHOR("Ricardo Cerqueira"); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL"); MODULE_VERSION(CX88_VERSION); diff --git a/drivers/media/pci/cx88/cx88-blackbird.c b/drivers/media/pci/cx88/cx88-blackbird.c index 0e0952e60795..7a4876cf9f08 100644 --- a/drivers/media/pci/cx88/cx88-blackbird.c +++ b/drivers/media/pci/cx88/cx88-blackbird.c @@ -5,7 +5,7 @@ * (c) 2004 Jelle Foks * (c) 2004 Gerd Knorr * - * (c) 2005-2006 Mauro Carvalho Chehab + * (c) 2005-2006 Mauro Carvalho Chehab * - video_ioctl2 conversion * * Includes parts from the ivtv driver diff --git a/drivers/media/pci/cx88/cx88-core.c b/drivers/media/pci/cx88/cx88-core.c index 8bfa5b7ed91b..60988e95b637 100644 --- a/drivers/media/pci/cx88/cx88-core.c +++ b/drivers/media/pci/cx88/cx88-core.c @@ -4,7 +4,7 @@ * * (c) 2003 Gerd Knorr [SuSE Labs] * - * (c) 2005-2006 Mauro Carvalho Chehab + * (c) 2005-2006 Mauro Carvalho Chehab * - Multituner support * - video_ioctl2 conversion * - PAL/M fixes diff --git a/drivers/media/pci/cx88/cx88-i2c.c b/drivers/media/pci/cx88/cx88-i2c.c index f7692775fb5a..99f88a05a7c9 100644 --- a/drivers/media/pci/cx88/cx88-i2c.c +++ b/drivers/media/pci/cx88/cx88-i2c.c @@ -8,7 +8,7 @@ * (c) 2002 Yurij Sysoev * (c) 1999-2003 Gerd Knorr * - * (c) 2005 Mauro Carvalho Chehab + * (c) 2005 Mauro Carvalho Chehab * - Multituner support and i2c address binding * * This program is free software; you can redistribute it and/or modify diff --git a/drivers/media/pci/cx88/cx88-video.c b/drivers/media/pci/cx88/cx88-video.c index 9be682cdb644..7b113bad70d2 100644 --- a/drivers/media/pci/cx88/cx88-video.c +++ b/drivers/media/pci/cx88/cx88-video.c @@ -5,7 +5,7 @@ * * (c) 2003-04 Gerd Knorr [SuSE Labs] * - * (c) 2005-2006 Mauro Carvalho Chehab + * (c) 2005-2006 Mauro Carvalho Chehab * - Multituner support * - video_ioctl2 conversion * - PAL/M fixes diff --git a/drivers/media/radio/radio-aimslab.c b/drivers/media/radio/radio-aimslab.c index 5ef635e72e10..4c52ac6d8bc5 100644 --- a/drivers/media/radio/radio-aimslab.c +++ b/drivers/media/radio/radio-aimslab.c @@ -4,7 +4,7 @@ * Copyright 1997 M. Kirkwood * * Converted to the radio-isa framework by Hans Verkuil - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab * Converted to new API by Alan Cox * Various bugfixes and enhancements by Russell Kroll * diff --git a/drivers/media/radio/radio-aztech.c b/drivers/media/radio/radio-aztech.c index 9e12c6027359..840b7d60462b 100644 --- a/drivers/media/radio/radio-aztech.c +++ b/drivers/media/radio/radio-aztech.c @@ -2,7 +2,7 @@ * radio-aztech.c - Aztech radio card driver * * Converted to the radio-isa framework by Hans Verkuil - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab * Adapted to support the Video for Linux API by * Russell Kroll . Based on original tuner code by: * diff --git a/drivers/media/radio/radio-gemtek.c b/drivers/media/radio/radio-gemtek.c index 3ff4c4e1435f..f051f8694ab9 100644 --- a/drivers/media/radio/radio-gemtek.c +++ b/drivers/media/radio/radio-gemtek.c @@ -15,7 +15,7 @@ * Various bugfixes and enhancements by Russell Kroll * * Converted to the radio-isa framework by Hans Verkuil - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab * * Note: this card seems to swap the left and right audio channels! * diff --git a/drivers/media/radio/radio-maxiradio.c b/drivers/media/radio/radio-maxiradio.c index 95f06f3b35dc..e4e758739246 100644 --- a/drivers/media/radio/radio-maxiradio.c +++ b/drivers/media/radio/radio-maxiradio.c @@ -27,7 +27,7 @@ * BUGS: * - card unmutes if you change frequency * - * (c) 2006, 2007 by Mauro Carvalho Chehab : + * (c) 2006, 2007 by Mauro Carvalho Chehab : * - Conversion to V4L2 API * - Uses video_ioctl2 for parsing and to add debug support */ diff --git a/drivers/media/radio/radio-rtrack2.c b/drivers/media/radio/radio-rtrack2.c index abeaedd8d437..5a1470eb753e 100644 --- a/drivers/media/radio/radio-rtrack2.c +++ b/drivers/media/radio/radio-rtrack2.c @@ -7,7 +7,7 @@ * Various bugfixes and enhancements by Russell Kroll * * Converted to the radio-isa framework by Hans Verkuil - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab * * Fully tested with actual hardware and the v4l2-compliance tool. */ diff --git a/drivers/media/radio/radio-sf16fmi.c b/drivers/media/radio/radio-sf16fmi.c index fc4e63d36e4c..4f9b97edd9eb 100644 --- a/drivers/media/radio/radio-sf16fmi.c +++ b/drivers/media/radio/radio-sf16fmi.c @@ -13,7 +13,7 @@ * No volume control - only mute/unmute - you have to use line volume * control on SB-part of SF16-FMI/SF16-FMP/SF16-FMD * - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab */ #include /* __setup */ diff --git a/drivers/media/radio/radio-terratec.c b/drivers/media/radio/radio-terratec.c index 4f116ea294fb..1af8f29cc7d1 100644 --- a/drivers/media/radio/radio-terratec.c +++ b/drivers/media/radio/radio-terratec.c @@ -17,7 +17,7 @@ * Volume Control is done digitally * * Converted to the radio-isa framework by Hans Verkuil - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab */ #include /* Modules */ diff --git a/drivers/media/radio/radio-trust.c b/drivers/media/radio/radio-trust.c index 26a8c6002121..a4bad322ffff 100644 --- a/drivers/media/radio/radio-trust.c +++ b/drivers/media/radio/radio-trust.c @@ -12,7 +12,7 @@ * Scott McGrath (smcgrath@twilight.vtc.vsc.edu) * William McGrath (wmcgrath@twilight.vtc.vsc.edu) * - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab */ #include diff --git a/drivers/media/radio/radio-typhoon.c b/drivers/media/radio/radio-typhoon.c index eb72a4d13758..d0d67ad85b8f 100644 --- a/drivers/media/radio/radio-typhoon.c +++ b/drivers/media/radio/radio-typhoon.c @@ -25,7 +25,7 @@ * The frequency change is necessary since the card never seems to be * completely silent. * - * Converted to V4L2 API by Mauro Carvalho Chehab + * Converted to V4L2 API by Mauro Carvalho Chehab */ #include /* Modules */ diff --git a/drivers/media/radio/radio-zoltrix.c b/drivers/media/radio/radio-zoltrix.c index 026e88eef29c..6007cd09b328 100644 --- a/drivers/media/radio/radio-zoltrix.c +++ b/drivers/media/radio/radio-zoltrix.c @@ -27,7 +27,7 @@ * 2002-07-15 - Fix Stereo typo * * 2006-07-24 - Converted to V4L2 API - * by Mauro Carvalho Chehab + * by Mauro Carvalho Chehab * * Converted to the radio-isa framework by Hans Verkuil * diff --git a/drivers/media/rc/keymaps/rc-avermedia-m135a.c b/drivers/media/rc/keymaps/rc-avermedia-m135a.c index f6977df1a75b..d275d98d066a 100644 --- a/drivers/media/rc/keymaps/rc-avermedia-m135a.c +++ b/drivers/media/rc/keymaps/rc-avermedia-m135a.c @@ -12,7 +12,7 @@ * * On Avermedia M135A with IR model RM-JX, the same codes exist on both * Positivo (BR) and original IR, initial version and remote control codes - * added by Mauro Carvalho Chehab + * added by Mauro Carvalho Chehab * * Positivo also ships Avermedia M135A with model RM-K6, extra control * codes added by Herton Ronaldo Krzesinski diff --git a/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c b/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c index e4e78c1f4123..057c13b765ef 100644 --- a/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c +++ b/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c @@ -9,7 +9,7 @@ #include /* Encore ENLTV-FM v5.3 - Mauro Carvalho Chehab + Mauro Carvalho Chehab */ static struct rc_map_table encore_enltv_fm53[] = { diff --git a/drivers/media/rc/keymaps/rc-encore-enltv2.c b/drivers/media/rc/keymaps/rc-encore-enltv2.c index c3d4437a6fda..cd0555924456 100644 --- a/drivers/media/rc/keymaps/rc-encore-enltv2.c +++ b/drivers/media/rc/keymaps/rc-encore-enltv2.c @@ -9,7 +9,7 @@ #include /* Encore ENLTV2-FM - silver plastic - "Wand Media" written at the botton - Mauro Carvalho Chehab */ + Mauro Carvalho Chehab */ static struct rc_map_table encore_enltv2[] = { { 0x4c, KEY_POWER2 }, diff --git a/drivers/media/rc/keymaps/rc-kaiomy.c b/drivers/media/rc/keymaps/rc-kaiomy.c index f0f88df18606..a00051339842 100644 --- a/drivers/media/rc/keymaps/rc-kaiomy.c +++ b/drivers/media/rc/keymaps/rc-kaiomy.c @@ -9,7 +9,7 @@ #include /* Kaiomy TVnPC U2 - Mauro Carvalho Chehab + Mauro Carvalho Chehab */ static struct rc_map_table kaiomy[] = { diff --git a/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c b/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c index 453e04377de7..db5edde3eeb1 100644 --- a/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c +++ b/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c @@ -9,7 +9,7 @@ #include /* Kworld Plus TV Analog Lite PCI IR - Mauro Carvalho Chehab + Mauro Carvalho Chehab */ static struct rc_map_table kworld_plus_tv_analog[] = { diff --git a/drivers/media/rc/keymaps/rc-pixelview-new.c b/drivers/media/rc/keymaps/rc-pixelview-new.c index 791130f108ff..e4e34f2ccf74 100644 --- a/drivers/media/rc/keymaps/rc-pixelview-new.c +++ b/drivers/media/rc/keymaps/rc-pixelview-new.c @@ -9,7 +9,7 @@ #include /* - Mauro Carvalho Chehab + Mauro Carvalho Chehab present on PV MPEG 8000GT */ diff --git a/drivers/media/tuners/tea5761.c b/drivers/media/tuners/tea5761.c index 88b3e80c38ad..d78a2bdb3e36 100644 --- a/drivers/media/tuners/tea5761.c +++ b/drivers/media/tuners/tea5761.c @@ -2,7 +2,7 @@ // For Philips TEA5761 FM Chip // I2C address is always 0x20 (0x10 at 7-bit mode). // -// Copyright (c) 2005-2007 Mauro Carvalho Chehab (mchehab@infradead.org) +// Copyright (c) 2005-2007 Mauro Carvalho Chehab #include #include @@ -337,5 +337,5 @@ EXPORT_SYMBOL_GPL(tea5761_attach); EXPORT_SYMBOL_GPL(tea5761_autodetection); MODULE_DESCRIPTION("Philips TEA5761 FM tuner driver"); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL v2"); diff --git a/drivers/media/tuners/tea5767.c b/drivers/media/tuners/tea5767.c index 2b2c064d7dc3..016d0d5ec50b 100644 --- a/drivers/media/tuners/tea5767.c +++ b/drivers/media/tuners/tea5767.c @@ -2,7 +2,7 @@ // For Philips TEA5767 FM Chip used on some TV Cards like Prolink Pixelview // I2C address is always 0xC0. // -// Copyright (c) 2005 Mauro Carvalho Chehab (mchehab@infradead.org) +// Copyright (c) 2005 Mauro Carvalho Chehab // // tea5767 autodetection thanks to Torsten Seeboth and Atsushi Nakagawa // from their contributions on DScaler. @@ -469,5 +469,5 @@ EXPORT_SYMBOL_GPL(tea5767_attach); EXPORT_SYMBOL_GPL(tea5767_autodetection); MODULE_DESCRIPTION("Philips TEA5767 FM tuner driver"); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL v2"); diff --git a/drivers/media/tuners/tuner-xc2028-types.h b/drivers/media/tuners/tuner-xc2028-types.h index bb0437c36c03..50d017a4822a 100644 --- a/drivers/media/tuners/tuner-xc2028-types.h +++ b/drivers/media/tuners/tuner-xc2028-types.h @@ -5,7 +5,7 @@ * This file includes internal tipes to be used inside tuner-xc2028. * Shouldn't be included outside tuner-xc2028 * - * Copyright (c) 2007-2008 Mauro Carvalho Chehab (mchehab@infradead.org) + * Copyright (c) 2007-2008 Mauro Carvalho Chehab */ /* xc3028 firmware types */ diff --git a/drivers/media/tuners/tuner-xc2028.c b/drivers/media/tuners/tuner-xc2028.c index fca85e08ebd7..84744e138982 100644 --- a/drivers/media/tuners/tuner-xc2028.c +++ b/drivers/media/tuners/tuner-xc2028.c @@ -1,7 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 // tuner-xc2028 // -// Copyright (c) 2007-2008 Mauro Carvalho Chehab (mchehab@infradead.org) +// Copyright (c) 2007-2008 Mauro Carvalho Chehab // // Copyright (c) 2007 Michel Ludwig (michel.ludwig@gmail.com) // - frontend interface @@ -1518,7 +1518,7 @@ EXPORT_SYMBOL(xc2028_attach); MODULE_DESCRIPTION("Xceive xc2028/xc3028 tuner driver"); MODULE_AUTHOR("Michel Ludwig "); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL v2"); MODULE_FIRMWARE(XC2028_DEFAULT_FIRMWARE); MODULE_FIRMWARE(XC3028L_DEFAULT_FIRMWARE); diff --git a/drivers/media/tuners/tuner-xc2028.h b/drivers/media/tuners/tuner-xc2028.h index 03fd6d4233a4..7b58bc06e35c 100644 --- a/drivers/media/tuners/tuner-xc2028.h +++ b/drivers/media/tuners/tuner-xc2028.h @@ -2,7 +2,7 @@ * SPDX-License-Identifier: GPL-2.0 * tuner-xc2028 * - * Copyright (c) 2007-2008 Mauro Carvalho Chehab (mchehab@infradead.org) + * Copyright (c) 2007-2008 Mauro Carvalho Chehab */ #ifndef __TUNER_XC2028_H__ diff --git a/drivers/media/usb/em28xx/em28xx-camera.c b/drivers/media/usb/em28xx/em28xx-camera.c index 3c2694a16ed1..d1e66b503f4d 100644 --- a/drivers/media/usb/em28xx/em28xx-camera.c +++ b/drivers/media/usb/em28xx/em28xx-camera.c @@ -2,7 +2,7 @@ // // em28xx-camera.c - driver for Empia EM25xx/27xx/28xx USB video capture devices // -// Copyright (C) 2009 Mauro Carvalho Chehab +// Copyright (C) 2009 Mauro Carvalho Chehab // Copyright (C) 2013 Frank Schäfer // // This program is free software; you can redistribute it and/or modify diff --git a/drivers/media/usb/em28xx/em28xx-cards.c b/drivers/media/usb/em28xx/em28xx-cards.c index 6e0e67d23876..7c3203d7044b 100644 --- a/drivers/media/usb/em28xx/em28xx-cards.c +++ b/drivers/media/usb/em28xx/em28xx-cards.c @@ -5,7 +5,7 @@ // // Copyright (C) 2005 Ludovico Cavedon // Markus Rechberger -// Mauro Carvalho Chehab +// Mauro Carvalho Chehab // Sascha Sommer // Copyright (C) 2012 Frank Schäfer // diff --git a/drivers/media/usb/em28xx/em28xx-core.c b/drivers/media/usb/em28xx/em28xx-core.c index 36d341fb65dd..f28995383090 100644 --- a/drivers/media/usb/em28xx/em28xx-core.c +++ b/drivers/media/usb/em28xx/em28xx-core.c @@ -4,7 +4,7 @@ // // Copyright (C) 2005 Ludovico Cavedon // Markus Rechberger -// Mauro Carvalho Chehab +// Mauro Carvalho Chehab // Sascha Sommer // Copyright (C) 2012 Frank Schäfer // @@ -32,7 +32,7 @@ #define DRIVER_AUTHOR "Ludovico Cavedon , " \ "Markus Rechberger , " \ - "Mauro Carvalho Chehab , " \ + "Mauro Carvalho Chehab , " \ "Sascha Sommer " MODULE_AUTHOR(DRIVER_AUTHOR); diff --git a/drivers/media/usb/em28xx/em28xx-dvb.c b/drivers/media/usb/em28xx/em28xx-dvb.c index a54cb8dc52c9..3f493e0b0716 100644 --- a/drivers/media/usb/em28xx/em28xx-dvb.c +++ b/drivers/media/usb/em28xx/em28xx-dvb.c @@ -2,7 +2,7 @@ // // DVB device driver for em28xx // -// (c) 2008-2011 Mauro Carvalho Chehab +// (c) 2008-2011 Mauro Carvalho Chehab // // (c) 2008 Devin Heitmueller // - Fixes for the driver to properly work with HVR-950 @@ -63,7 +63,7 @@ #include "tc90522.h" #include "qm1d1c0042.h" -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL v2"); MODULE_DESCRIPTION(DRIVER_DESC " - digital TV interface"); MODULE_VERSION(EM28XX_VERSION); diff --git a/drivers/media/usb/em28xx/em28xx-i2c.c b/drivers/media/usb/em28xx/em28xx-i2c.c index 9151bccd859a..6458682bc6e2 100644 --- a/drivers/media/usb/em28xx/em28xx-i2c.c +++ b/drivers/media/usb/em28xx/em28xx-i2c.c @@ -4,7 +4,7 @@ // // Copyright (C) 2005 Ludovico Cavedon // Markus Rechberger -// Mauro Carvalho Chehab +// Mauro Carvalho Chehab // Sascha Sommer // Copyright (C) 2013 Frank Schäfer // diff --git a/drivers/media/usb/em28xx/em28xx-input.c b/drivers/media/usb/em28xx/em28xx-input.c index 2dc1be00b8b8..f84a1208d5d3 100644 --- a/drivers/media/usb/em28xx/em28xx-input.c +++ b/drivers/media/usb/em28xx/em28xx-input.c @@ -4,7 +4,7 @@ // // Copyright (C) 2005 Ludovico Cavedon // Markus Rechberger -// Mauro Carvalho Chehab +// Mauro Carvalho Chehab // Sascha Sommer // // This program is free software; you can redistribute it and/or modify diff --git a/drivers/media/usb/em28xx/em28xx-video.c b/drivers/media/usb/em28xx/em28xx-video.c index d70ee13cc52e..68571bf36d28 100644 --- a/drivers/media/usb/em28xx/em28xx-video.c +++ b/drivers/media/usb/em28xx/em28xx-video.c @@ -5,7 +5,7 @@ // // Copyright (C) 2005 Ludovico Cavedon // Markus Rechberger -// Mauro Carvalho Chehab +// Mauro Carvalho Chehab // Sascha Sommer // Copyright (C) 2012 Frank Schäfer // @@ -44,7 +44,7 @@ #define DRIVER_AUTHOR "Ludovico Cavedon , " \ "Markus Rechberger , " \ - "Mauro Carvalho Chehab , " \ + "Mauro Carvalho Chehab , " \ "Sascha Sommer " static unsigned int isoc_debug; diff --git a/drivers/media/usb/em28xx/em28xx.h b/drivers/media/usb/em28xx/em28xx.h index 63c7c6124707..b0378e77ddff 100644 --- a/drivers/media/usb/em28xx/em28xx.h +++ b/drivers/media/usb/em28xx/em28xx.h @@ -4,7 +4,7 @@ * * Copyright (C) 2005 Markus Rechberger * Ludovico Cavedon - * Mauro Carvalho Chehab + * Mauro Carvalho Chehab * Copyright (C) 2012 Frank Schäfer * * Based on the em2800 driver from Sascha Sommer diff --git a/drivers/media/usb/gspca/zc3xx-reg.h b/drivers/media/usb/gspca/zc3xx-reg.h index a1bd94e8ce52..71fda38e85e0 100644 --- a/drivers/media/usb/gspca/zc3xx-reg.h +++ b/drivers/media/usb/gspca/zc3xx-reg.h @@ -1,7 +1,7 @@ /* * zc030x registers * - * Copyright (c) 2008 Mauro Carvalho Chehab + * Copyright (c) 2008 Mauro Carvalho Chehab * * The register aliases used here came from this driver: * http://zc0302.sourceforge.net/zc0302.php diff --git a/drivers/media/usb/tm6000/tm6000-cards.c b/drivers/media/usb/tm6000/tm6000-cards.c index 70939e96b856..23df50aa0a4a 100644 --- a/drivers/media/usb/tm6000/tm6000-cards.c +++ b/drivers/media/usb/tm6000/tm6000-cards.c @@ -1,7 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 // tm6000-cards.c - driver for TM5600/TM6000/TM6010 USB video capture devices // -// Copyright (c) 2006-2007 Mauro Carvalho Chehab +// Copyright (c) 2006-2007 Mauro Carvalho Chehab #include #include diff --git a/drivers/media/usb/tm6000/tm6000-core.c b/drivers/media/usb/tm6000/tm6000-core.c index 23a1332d98e6..d3229aa45fcb 100644 --- a/drivers/media/usb/tm6000/tm6000-core.c +++ b/drivers/media/usb/tm6000/tm6000-core.c @@ -1,7 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 // tm6000-core.c - driver for TM5600/TM6000/TM6010 USB video capture devices // -// Copyright (c) 2006-2007 Mauro Carvalho Chehab +// Copyright (c) 2006-2007 Mauro Carvalho Chehab // // Copyright (c) 2007 Michel Ludwig // - DVB-T support diff --git a/drivers/media/usb/tm6000/tm6000-i2c.c b/drivers/media/usb/tm6000/tm6000-i2c.c index c9a62bbff27a..659b63febf85 100644 --- a/drivers/media/usb/tm6000/tm6000-i2c.c +++ b/drivers/media/usb/tm6000/tm6000-i2c.c @@ -1,7 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 // tm6000-i2c.c - driver for TM5600/TM6000/TM6010 USB video capture devices // -// Copyright (c) 2006-2007 Mauro Carvalho Chehab +// Copyright (c) 2006-2007 Mauro Carvalho Chehab // // Copyright (c) 2007 Michel Ludwig // - Fix SMBus Read Byte command diff --git a/drivers/media/usb/tm6000/tm6000-regs.h b/drivers/media/usb/tm6000/tm6000-regs.h index 21587fcf11e3..d10424673db9 100644 --- a/drivers/media/usb/tm6000/tm6000-regs.h +++ b/drivers/media/usb/tm6000/tm6000-regs.h @@ -2,7 +2,7 @@ * SPDX-License-Identifier: GPL-2.0 * tm6000-regs.h - driver for TM5600/TM6000/TM6010 USB video capture devices * - * Copyright (c) 2006-2007 Mauro Carvalho Chehab + * Copyright (c) 2006-2007 Mauro Carvalho Chehab */ /* diff --git a/drivers/media/usb/tm6000/tm6000-usb-isoc.h b/drivers/media/usb/tm6000/tm6000-usb-isoc.h index 5c615b0a7a46..b275dbce3a1b 100644 --- a/drivers/media/usb/tm6000/tm6000-usb-isoc.h +++ b/drivers/media/usb/tm6000/tm6000-usb-isoc.h @@ -2,7 +2,7 @@ * SPDX-License-Identifier: GPL-2.0 * tm6000-buf.c - driver for TM5600/TM6000/TM6010 USB video capture devices * - * Copyright (c) 2006-2007 Mauro Carvalho Chehab + * Copyright (c) 2006-2007 Mauro Carvalho Chehab */ #include diff --git a/drivers/media/usb/tm6000/tm6000-video.c b/drivers/media/usb/tm6000/tm6000-video.c index b2399d4266da..aa85fe31c835 100644 --- a/drivers/media/usb/tm6000/tm6000-video.c +++ b/drivers/media/usb/tm6000/tm6000-video.c @@ -1,7 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 // tm6000-video.c - driver for TM5600/TM6000/TM6010 USB video capture devices // -// Copyright (c) 2006-2007 Mauro Carvalho Chehab +// Copyright (c) 2006-2007 Mauro Carvalho Chehab // // Copyright (c) 2007 Michel Ludwig // - Fixed module load/unload diff --git a/drivers/media/usb/tm6000/tm6000.h b/drivers/media/usb/tm6000/tm6000.h index e1e45770e28d..0864ed7314eb 100644 --- a/drivers/media/usb/tm6000/tm6000.h +++ b/drivers/media/usb/tm6000/tm6000.h @@ -2,7 +2,7 @@ * SPDX-License-Identifier: GPL-2.0 * tm6000.h - driver for TM5600/TM6000/TM6010 USB video capture devices * - * Copyright (c) 2006-2007 Mauro Carvalho Chehab + * Copyright (c) 2006-2007 Mauro Carvalho Chehab * * Copyright (c) 2007 Michel Ludwig * - DVB-T support diff --git a/drivers/media/v4l2-core/v4l2-dev.c b/drivers/media/v4l2-core/v4l2-dev.c index 1d0b2208e8fb..c080dcc75393 100644 --- a/drivers/media/v4l2-core/v4l2-dev.c +++ b/drivers/media/v4l2-core/v4l2-dev.c @@ -10,7 +10,7 @@ * 2 of the License, or (at your option) any later version. * * Authors: Alan Cox, (version 1) - * Mauro Carvalho Chehab (version 2) + * Mauro Carvalho Chehab (version 2) * * Fixes: 20000516 Claudio Matsuoka * - Added procfs support @@ -1072,7 +1072,7 @@ static void __exit videodev_exit(void) subsys_initcall(videodev_init); module_exit(videodev_exit) -MODULE_AUTHOR("Alan Cox, Mauro Carvalho Chehab "); +MODULE_AUTHOR("Alan Cox, Mauro Carvalho Chehab "); MODULE_DESCRIPTION("Device registrar for Video4Linux drivers v2"); MODULE_LICENSE("GPL"); MODULE_ALIAS_CHARDEV_MAJOR(VIDEO_MAJOR); diff --git a/drivers/media/v4l2-core/v4l2-ioctl.c b/drivers/media/v4l2-core/v4l2-ioctl.c index f48c505550e0..de5d96dbe69e 100644 --- a/drivers/media/v4l2-core/v4l2-ioctl.c +++ b/drivers/media/v4l2-core/v4l2-ioctl.c @@ -9,7 +9,7 @@ * 2 of the License, or (at your option) any later version. * * Authors: Alan Cox, (version 1) - * Mauro Carvalho Chehab (version 2) + * Mauro Carvalho Chehab (version 2) */ #include diff --git a/drivers/media/v4l2-core/videobuf-core.c b/drivers/media/v4l2-core/videobuf-core.c index 2b3981842b4b..7491b337002c 100644 --- a/drivers/media/v4l2-core/videobuf-core.c +++ b/drivers/media/v4l2-core/videobuf-core.c @@ -1,11 +1,11 @@ /* * generic helper functions for handling video4linux capture buffers * - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * Highly based on video-buf written originally by: * (c) 2001,02 Gerd Knorr - * (c) 2006 Mauro Carvalho Chehab, + * (c) 2006 Mauro Carvalho Chehab, * (c) 2006 Ted Walther and John Sokol * * This program is free software; you can redistribute it and/or modify @@ -38,7 +38,7 @@ static int debug; module_param(debug, int, 0644); MODULE_DESCRIPTION("helper module to manage video4linux buffers"); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL"); #define dprintk(level, fmt, arg...) \ diff --git a/drivers/media/v4l2-core/videobuf-dma-contig.c b/drivers/media/v4l2-core/videobuf-dma-contig.c index e02353e340dd..f46132504d88 100644 --- a/drivers/media/v4l2-core/videobuf-dma-contig.c +++ b/drivers/media/v4l2-core/videobuf-dma-contig.c @@ -7,7 +7,7 @@ * Copyright (c) 2008 Magnus Damm * * Based on videobuf-vmalloc.c, - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by diff --git a/drivers/media/v4l2-core/videobuf-dma-sg.c b/drivers/media/v4l2-core/videobuf-dma-sg.c index add2edb23eac..7770034aae28 100644 --- a/drivers/media/v4l2-core/videobuf-dma-sg.c +++ b/drivers/media/v4l2-core/videobuf-dma-sg.c @@ -6,11 +6,11 @@ * into PAGE_SIZE chunks). They also assume the driver does not need * to touch the video data. * - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * Highly based on video-buf written originally by: * (c) 2001,02 Gerd Knorr - * (c) 2006 Mauro Carvalho Chehab, + * (c) 2006 Mauro Carvalho Chehab, * (c) 2006 Ted Walther and John Sokol * * This program is free software; you can redistribute it and/or modify @@ -48,7 +48,7 @@ static int debug; module_param(debug, int, 0644); MODULE_DESCRIPTION("helper module to manage video4linux dma sg buffers"); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL"); #define dprintk(level, fmt, arg...) \ diff --git a/drivers/media/v4l2-core/videobuf-vmalloc.c b/drivers/media/v4l2-core/videobuf-vmalloc.c index 2ff7fcc77b11..45fe781aeeec 100644 --- a/drivers/media/v4l2-core/videobuf-vmalloc.c +++ b/drivers/media/v4l2-core/videobuf-vmalloc.c @@ -6,7 +6,7 @@ * into PAGE_SIZE chunks). They also assume the driver does not need * to touch the video data. * - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -41,7 +41,7 @@ static int debug; module_param(debug, int, 0644); MODULE_DESCRIPTION("helper module to manage video4linux vmalloc buffers"); -MODULE_AUTHOR("Mauro Carvalho Chehab "); +MODULE_AUTHOR("Mauro Carvalho Chehab "); MODULE_LICENSE("GPL"); #define dprintk(level, fmt, arg...) \ diff --git a/include/media/i2c/tvp7002.h b/include/media/i2c/tvp7002.h index 5ee007c1cead..cb213c136089 100644 --- a/include/media/i2c/tvp7002.h +++ b/include/media/i2c/tvp7002.h @@ -5,7 +5,7 @@ * Author: Santiago Nunez-Corrales * * This code is partially based upon the TVP5150 driver - * written by Mauro Carvalho Chehab (mchehab@infradead.org), + * written by Mauro Carvalho Chehab , * the TVP514x driver written by Vaibhav Hiremath * and the TVP7002 driver in the TI LSP 2.10.00.14 * diff --git a/include/media/videobuf-core.h b/include/media/videobuf-core.h index 0bda0adc744f..60a664febba0 100644 --- a/include/media/videobuf-core.h +++ b/include/media/videobuf-core.h @@ -1,11 +1,11 @@ /* * generic helper functions for handling video4linux capture buffers * - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * Highly based on video-buf written originally by: * (c) 2001,02 Gerd Knorr - * (c) 2006 Mauro Carvalho Chehab, + * (c) 2006 Mauro Carvalho Chehab, * (c) 2006 Ted Walther and John Sokol * * This program is free software; you can redistribute it and/or modify diff --git a/include/media/videobuf-dma-sg.h b/include/media/videobuf-dma-sg.h index d8b27854e3bf..01bd142b979d 100644 --- a/include/media/videobuf-dma-sg.h +++ b/include/media/videobuf-dma-sg.h @@ -6,11 +6,11 @@ * into PAGE_SIZE chunks). They also assume the driver does not need * to touch the video data. * - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * Highly based on video-buf written originally by: * (c) 2001,02 Gerd Knorr - * (c) 2006 Mauro Carvalho Chehab, + * (c) 2006 Mauro Carvalho Chehab, * (c) 2006 Ted Walther and John Sokol * * This program is free software; you can redistribute it and/or modify diff --git a/include/media/videobuf-vmalloc.h b/include/media/videobuf-vmalloc.h index 486a97efdb56..36c6a4ad3504 100644 --- a/include/media/videobuf-vmalloc.h +++ b/include/media/videobuf-vmalloc.h @@ -6,7 +6,7 @@ * into PAGE_SIZE chunks). They also assume the driver does not need * to touch the video data. * - * (c) 2007 Mauro Carvalho Chehab, + * (c) 2007 Mauro Carvalho Chehab, * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by diff --git a/scripts/extract_xc3028.pl b/scripts/extract_xc3028.pl index 61d9b256c658..a1c51b7e4baf 100755 --- a/scripts/extract_xc3028.pl +++ b/scripts/extract_xc3028.pl @@ -1,6 +1,6 @@ #!/usr/bin/env perl -# Copyright (c) Mauro Carvalho Chehab +# Copyright (c) Mauro Carvalho Chehab # Released under GPLv2 # # In order to use, you need to: diff --git a/scripts/split-man.pl b/scripts/split-man.pl index bfe16cbe42df..c3db607ee9ec 100755 --- a/scripts/split-man.pl +++ b/scripts/split-man.pl @@ -1,7 +1,7 @@ #!/usr/bin/perl # SPDX-License-Identifier: GPL-2.0 # -# Author: Mauro Carvalho Chehab +# Author: Mauro Carvalho Chehab # # Produce manpages from kernel-doc. # See Documentation/doc-guide/kernel-doc.rst for instructions -- cgit v1.2.3 From 23b8392201e0681b76630c4cea68e1a2e1821ec6 Mon Sep 17 00:00:00 2001 From: Bhadram Varka Date: Wed, 2 May 2018 20:43:58 +0530 Subject: net: phy: broadcom: add support for BCM89610 PHY It adds support for BCM89610 (Single-Port 10/100/1000BASE-T) transceiver which is used in P3310 Tegra186 platform. Signed-off-by: Bhadram Varka Signed-off-by: David S. Miller --- drivers/net/phy/broadcom.c | 10 ++++++++++ include/linux/brcmphy.h | 1 + 2 files changed, 11 insertions(+) (limited to 'include') diff --git a/drivers/net/phy/broadcom.c b/drivers/net/phy/broadcom.c index 3bb6b66dc7bf..f9c25912eb98 100644 --- a/drivers/net/phy/broadcom.c +++ b/drivers/net/phy/broadcom.c @@ -720,6 +720,15 @@ static struct phy_driver broadcom_drivers[] = { .get_strings = bcm_phy_get_strings, .get_stats = bcm53xx_phy_get_stats, .probe = bcm53xx_phy_probe, +}, { + .phy_id = PHY_ID_BCM89610, + .phy_id_mask = 0xfffffff0, + .name = "Broadcom BCM89610", + .features = PHY_GBIT_FEATURES, + .flags = PHY_HAS_INTERRUPT, + .config_init = bcm54xx_config_init, + .ack_interrupt = bcm_phy_ack_intr, + .config_intr = bcm_phy_config_intr, } }; module_phy_driver(broadcom_drivers); @@ -741,6 +750,7 @@ static struct mdio_device_id __maybe_unused broadcom_tbl[] = { { PHY_ID_BCMAC131, 0xfffffff0 }, { PHY_ID_BCM5241, 0xfffffff0 }, { PHY_ID_BCM5395, 0xfffffff0 }, + { PHY_ID_BCM89610, 0xfffffff0 }, { } }; diff --git a/include/linux/brcmphy.h b/include/linux/brcmphy.h index d3339dd48b1a..b324e01ccf2d 100644 --- a/include/linux/brcmphy.h +++ b/include/linux/brcmphy.h @@ -25,6 +25,7 @@ #define PHY_ID_BCM54612E 0x03625e60 #define PHY_ID_BCM54616S 0x03625d10 #define PHY_ID_BCM57780 0x03625d90 +#define PHY_ID_BCM89610 0x03625cd0 #define PHY_ID_BCM7250 0xae025280 #define PHY_ID_BCM7260 0xae025190 -- cgit v1.2.3 From d1361b32e6aec7440c01d5c8fcc54189930a342d Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 26 Apr 2018 18:17:31 -0700 Subject: mac80211: fix kernel-doc "bad line" warning Fix 88 instances of a kernel-doc warning: ../include/net/mac80211.h:2083: warning: bad line: > Signed-off-by: Randy Dunlap Cc: linux-wireless@vger.kernel.org Cc: Johannes Berg Signed-off-by: Johannes Berg --- include/net/mac80211.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index d2279b2d61aa..b2f3a0c018e7 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -2080,7 +2080,7 @@ struct ieee80211_txq { * virtual interface might not be given air time for the transmission of * the frame, as it is not synced with the AP/P2P GO yet, and thus the * deauthentication frame might not be transmitted. - > + * * @IEEE80211_HW_DOESNT_SUPPORT_QOS_NDP: The driver (or firmware) doesn't * support QoS NDP for AP probing - that's most likely a driver bug. * -- cgit v1.2.3 From 53bc017f72b3b7f7f1bad3043c2d6dd87331ea44 Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Sun, 6 May 2018 13:23:52 +0200 Subject: net: flow_dissector: fix typo 'can by' to 'can be' Signed-off-by: Wolfram Sang Signed-off-by: David S. Miller --- include/net/flow_dissector.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/flow_dissector.h b/include/net/flow_dissector.h index 9a074776f70b..d1fcf2442a42 100644 --- a/include/net/flow_dissector.h +++ b/include/net/flow_dissector.h @@ -251,7 +251,7 @@ extern struct flow_dissector flow_keys_buf_dissector; * This structure is used to hold a digest of the full flow keys. This is a * larger "hash" of a flow to allow definitively matching specific flows where * the 32 bit skb->hash is not large enough. The size is limited to 16 bytes so - * that it can by used in CB of skb (see sch_choke for an example). + * that it can be used in CB of skb (see sch_choke for an example). */ #define FLOW_KEYS_DIGEST_LEN 16 struct flow_keys_digest { -- cgit v1.2.3 From 9913f74fe15705acd5163551ddf449568cf0048d Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Thu, 10 May 2018 08:46:36 +0900 Subject: drm/exynos: ipp: Add IPP v2 framework This patch adds Exynos IPP v2 subsystem and userspace API. New userspace API is focused ONLY on memory-to-memory image processing. The two remainging operation modes of obsolete IPP v1 API (framebuffer writeback and local-path output with image processing) can be implemented using standard DRM features: writeback connectors and additional DRM planes with scaling features. V2 IPP userspace API is based on stateless approach, which much better fits to memory-to-memory image processing model. It also provides support for all image formats, which are both already defined in DRM API and supported by the existing IPP hardware modules. The API consists of the following ioctls: - DRM_IOCTL_EXYNOS_IPP_GET_RESOURCES: to enumerate all available image processing modules, - DRM_IOCTL_EXYNOS_IPP_GET_CAPS: to query capabilities and supported image formats of given IPP module, - DRM_IOCTL_EXYNOS_IPP_GET_LIMITS: to query hardware limitiations for selected image format of given IPP module, - DRM_IOCTL_EXYNOS_IPP_COMMIT: to perform operation described by the provided structures (source and destination buffers, operation rectangle, transformation, etc). The proposed userspace API is extensible. In the future more advanced image processing operations can be defined to support for example blending. Userspace API is fully functional also on DRM render nodes, so it is not limited to the root/privileged client. Internal driver API also has been completely rewritten. New IPP core performs all possible input validation, checks and object life-time control. The drivers can focus only on writing configuration to hardware registers. Stateless nature of DRM_IOCTL_EXYNOS_IPP_COMMIT ioctl simplifies the driver API. Minimal driver needs to provide a single callback for starting processing and an array with supported image formats. Signed-off-by: Marek Szyprowski Tested-by: Hoegeun Kwon Merge conflict so merged manually. Signed-off-by: Inki Dae --- drivers/gpu/drm/exynos/Kconfig | 3 + drivers/gpu/drm/exynos/Makefile | 1 + drivers/gpu/drm/exynos/exynos_drm_drv.c | 22 +- drivers/gpu/drm/exynos/exynos_drm_ipp.c | 916 ++++++++++++++++++++++++++++++++ drivers/gpu/drm/exynos/exynos_drm_ipp.h | 175 ++++++ include/uapi/drm/exynos_drm.h | 240 +++++++++ 6 files changed, 1355 insertions(+), 2 deletions(-) create mode 100644 drivers/gpu/drm/exynos/exynos_drm_ipp.c create mode 100644 drivers/gpu/drm/exynos/exynos_drm_ipp.h (limited to 'include') diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig index 1548a784ef71..9e914655c430 100644 --- a/drivers/gpu/drm/exynos/Kconfig +++ b/drivers/gpu/drm/exynos/Kconfig @@ -95,6 +95,9 @@ config DRM_EXYNOS_G2D help Choose this option if you want to use Exynos G2D for DRM. +config DRM_EXYNOS_IPP + bool + config DRM_EXYNOS_FIMC bool "FIMC" depends on BROKEN && MFD_SYSCON diff --git a/drivers/gpu/drm/exynos/Makefile b/drivers/gpu/drm/exynos/Makefile index a51c5459bb13..bdf4212dde7b 100644 --- a/drivers/gpu/drm/exynos/Makefile +++ b/drivers/gpu/drm/exynos/Makefile @@ -18,6 +18,7 @@ exynosdrm-$(CONFIG_DRM_EXYNOS_MIXER) += exynos_mixer.o exynosdrm-$(CONFIG_DRM_EXYNOS_HDMI) += exynos_hdmi.o exynosdrm-$(CONFIG_DRM_EXYNOS_VIDI) += exynos_drm_vidi.o exynosdrm-$(CONFIG_DRM_EXYNOS_G2D) += exynos_drm_g2d.o +exynosdrm-$(CONFIG_DRM_EXYNOS_IPP) += exynos_drm_ipp.o exynosdrm-$(CONFIG_DRM_EXYNOS_FIMC) += exynos_drm_fimc.o exynosdrm-$(CONFIG_DRM_EXYNOS_ROTATOR) += exynos_drm_rotator.o exynosdrm-$(CONFIG_DRM_EXYNOS_GSC) += exynos_drm_gsc.o diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c index a518e9c6d6cc..37c0db759674 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_drv.c +++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c @@ -27,15 +27,23 @@ #include "exynos_drm_fb.h" #include "exynos_drm_gem.h" #include "exynos_drm_plane.h" +#include "exynos_drm_ipp.h" #include "exynos_drm_vidi.h" #include "exynos_drm_g2d.h" #include "exynos_drm_iommu.h" #define DRIVER_NAME "exynos" #define DRIVER_DESC "Samsung SoC DRM" -#define DRIVER_DATE "20110530" +#define DRIVER_DATE "20180330" + +/* + * Interface history: + * + * 1.0 - Original version + * 1.1 - Upgrade IPP driver to version 2.0 + */ #define DRIVER_MAJOR 1 -#define DRIVER_MINOR 0 +#define DRIVER_MINOR 1 int exynos_atomic_check(struct drm_device *dev, struct drm_atomic_state *state) @@ -108,6 +116,16 @@ static const struct drm_ioctl_desc exynos_ioctls[] = { DRM_AUTH | DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(EXYNOS_G2D_EXEC, exynos_g2d_exec_ioctl, DRM_AUTH | DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_RESOURCES, + exynos_drm_ipp_get_res_ioctl, + DRM_AUTH | DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_CAPS, exynos_drm_ipp_get_caps_ioctl, + DRM_AUTH | DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_LIMITS, + exynos_drm_ipp_get_limits_ioctl, + DRM_AUTH | DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(EXYNOS_IPP_COMMIT, exynos_drm_ipp_commit_ioctl, + DRM_AUTH | DRM_RENDER_ALLOW), }; static const struct file_operations exynos_drm_driver_fops = { diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c new file mode 100644 index 000000000000..26374e58c557 --- /dev/null +++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c @@ -0,0 +1,916 @@ +/* + * Copyright (C) 2017 Samsung Electronics Co.Ltd + * Authors: + * Marek Szyprowski + * + * Exynos DRM Image Post Processing (IPP) related functions + * + * 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, sublicense, + * 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 shall be included in + * all copies or substantial portions of the Software. + */ + + +#include +#include +#include + +#include "exynos_drm_drv.h" +#include "exynos_drm_gem.h" +#include "exynos_drm_ipp.h" + +static int num_ipp; +static LIST_HEAD(ipp_list); + +/** + * exynos_drm_ipp_register - Register a new picture processor hardware module + * @dev: DRM device + * @ipp: ipp module to init + * @funcs: callbacks for the new ipp object + * @caps: bitmask of ipp capabilities (%DRM_EXYNOS_IPP_CAP_*) + * @formats: array of supported formats + * @num_formats: size of the supported formats array + * @name: name (for debugging purposes) + * + * Initializes a ipp module. + * + * Returns: + * Zero on success, error code on failure. + */ +int exynos_drm_ipp_register(struct drm_device *dev, struct exynos_drm_ipp *ipp, + const struct exynos_drm_ipp_funcs *funcs, unsigned int caps, + const struct exynos_drm_ipp_formats *formats, + unsigned int num_formats, const char *name) +{ + WARN_ON(!ipp); + WARN_ON(!funcs); + WARN_ON(!formats); + WARN_ON(!num_formats); + + spin_lock_init(&ipp->lock); + INIT_LIST_HEAD(&ipp->todo_list); + init_waitqueue_head(&ipp->done_wq); + ipp->dev = dev; + ipp->funcs = funcs; + ipp->capabilities = caps; + ipp->name = name; + ipp->formats = formats; + ipp->num_formats = num_formats; + + /* ipp_list modification is serialized by component framework */ + list_add_tail(&ipp->head, &ipp_list); + ipp->id = num_ipp++; + + DRM_DEBUG_DRIVER("Registered ipp %d\n", ipp->id); + + return 0; +} + +/** + * exynos_drm_ipp_unregister - Unregister the picture processor module + * @dev: DRM device + * @ipp: ipp module + */ +void exynos_drm_ipp_unregister(struct drm_device *dev, + struct exynos_drm_ipp *ipp) +{ + WARN_ON(ipp->task); + WARN_ON(!list_empty(&ipp->todo_list)); + list_del(&ipp->head); +} + +/** + * exynos_drm_ipp_ioctl_get_res_ioctl - enumerate all ipp modules + * @dev: DRM device + * @data: ioctl data + * @file_priv: DRM file info + * + * Construct a list of ipp ids. + * + * Called by the user via ioctl. + * + * Returns: + * Zero on success, negative errno on failure. + */ +int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + struct drm_exynos_ioctl_ipp_get_res *resp = data; + struct exynos_drm_ipp *ipp; + uint32_t __user *ipp_ptr = (uint32_t __user *) + (unsigned long)resp->ipp_id_ptr; + unsigned int count = num_ipp, copied = 0; + + /* + * This ioctl is called twice, once to determine how much space is + * needed, and the 2nd time to fill it. + */ + if (count && resp->count_ipps >= count) { + list_for_each_entry(ipp, &ipp_list, head) { + if (put_user(ipp->id, ipp_ptr + copied)) + return -EFAULT; + copied++; + } + } + resp->count_ipps = count; + + return 0; +} + +static inline struct exynos_drm_ipp *__ipp_get(uint32_t id) +{ + struct exynos_drm_ipp *ipp; + + list_for_each_entry(ipp, &ipp_list, head) + if (ipp->id == id) + return ipp; + return NULL; +} + +/** + * exynos_drm_ipp_ioctl_get_caps - get ipp module capabilities and formats + * @dev: DRM device + * @data: ioctl data + * @file_priv: DRM file info + * + * Construct a structure describing ipp module capabilities. + * + * Called by the user via ioctl. + * + * Returns: + * Zero on success, negative errno on failure. + */ +int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + struct drm_exynos_ioctl_ipp_get_caps *resp = data; + void __user *ptr = (void __user *)(unsigned long)resp->formats_ptr; + struct exynos_drm_ipp *ipp; + int i; + + ipp = __ipp_get(resp->ipp_id); + if (!ipp) + return -ENOENT; + + resp->ipp_id = ipp->id; + resp->capabilities = ipp->capabilities; + + /* + * This ioctl is called twice, once to determine how much space is + * needed, and the 2nd time to fill it. + */ + if (resp->formats_count >= ipp->num_formats) { + for (i = 0; i < ipp->num_formats; i++) { + struct drm_exynos_ipp_format tmp = { + .fourcc = ipp->formats[i].fourcc, + .type = ipp->formats[i].type, + .modifier = ipp->formats[i].modifier, + }; + + if (copy_to_user(ptr, &tmp, sizeof(tmp))) + return -EFAULT; + ptr += sizeof(tmp); + } + } + resp->formats_count = ipp->num_formats; + + return 0; +} + +static inline const struct exynos_drm_ipp_formats *__ipp_format_get( + struct exynos_drm_ipp *ipp, uint32_t fourcc, + uint64_t mod, unsigned int type) +{ + int i; + + for (i = 0; i < ipp->num_formats; i++) { + if ((ipp->formats[i].type & type) && + ipp->formats[i].fourcc == fourcc && + ipp->formats[i].modifier == mod) + return &ipp->formats[i]; + } + return NULL; +} + +/** + * exynos_drm_ipp_get_limits_ioctl - get ipp module limits + * @dev: DRM device + * @data: ioctl data + * @file_priv: DRM file info + * + * Construct a structure describing ipp module limitations for provided + * picture format. + * + * Called by the user via ioctl. + * + * Returns: + * Zero on success, negative errno on failure. + */ +int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + struct drm_exynos_ioctl_ipp_get_limits *resp = data; + void __user *ptr = (void __user *)(unsigned long)resp->limits_ptr; + const struct exynos_drm_ipp_formats *format; + struct exynos_drm_ipp *ipp; + + if (resp->type != DRM_EXYNOS_IPP_FORMAT_SOURCE && + resp->type != DRM_EXYNOS_IPP_FORMAT_DESTINATION) + return -EINVAL; + + ipp = __ipp_get(resp->ipp_id); + if (!ipp) + return -ENOENT; + + format = __ipp_format_get(ipp, resp->fourcc, resp->modifier, + resp->type); + if (!format) + return -EINVAL; + + /* + * This ioctl is called twice, once to determine how much space is + * needed, and the 2nd time to fill it. + */ + if (format->num_limits && resp->limits_count >= format->num_limits) + if (copy_to_user((void __user *)ptr, format->limits, + sizeof(*format->limits) * format->num_limits)) + return -EFAULT; + resp->limits_count = format->num_limits; + + return 0; +} + +struct drm_pending_exynos_ipp_event { + struct drm_pending_event base; + struct drm_exynos_ipp_event event; +}; + +static inline struct exynos_drm_ipp_task * + exynos_drm_ipp_task_alloc(struct exynos_drm_ipp *ipp) +{ + struct exynos_drm_ipp_task *task; + + task = kzalloc(sizeof(*task), GFP_KERNEL); + if (!task) + return NULL; + + task->dev = ipp->dev; + task->ipp = ipp; + + /* some defaults */ + task->src.rect.w = task->dst.rect.w = UINT_MAX; + task->src.rect.h = task->dst.rect.h = UINT_MAX; + task->transform.rotation = DRM_MODE_ROTATE_0; + + DRM_DEBUG_DRIVER("Allocated task %pK\n", task); + + return task; +} + +static const struct exynos_drm_param_map { + unsigned int id; + unsigned int size; + unsigned int offset; +} exynos_drm_ipp_params_maps[] = { + { + DRM_EXYNOS_IPP_TASK_BUFFER | DRM_EXYNOS_IPP_TASK_TYPE_SOURCE, + sizeof(struct drm_exynos_ipp_task_buffer), + offsetof(struct exynos_drm_ipp_task, src.buf), + }, { + DRM_EXYNOS_IPP_TASK_BUFFER | + DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION, + sizeof(struct drm_exynos_ipp_task_buffer), + offsetof(struct exynos_drm_ipp_task, dst.buf), + }, { + DRM_EXYNOS_IPP_TASK_RECTANGLE | DRM_EXYNOS_IPP_TASK_TYPE_SOURCE, + sizeof(struct drm_exynos_ipp_task_rect), + offsetof(struct exynos_drm_ipp_task, src.rect), + }, { + DRM_EXYNOS_IPP_TASK_RECTANGLE | + DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION, + sizeof(struct drm_exynos_ipp_task_rect), + offsetof(struct exynos_drm_ipp_task, dst.rect), + }, { + DRM_EXYNOS_IPP_TASK_TRANSFORM, + sizeof(struct drm_exynos_ipp_task_transform), + offsetof(struct exynos_drm_ipp_task, transform), + }, { + DRM_EXYNOS_IPP_TASK_ALPHA, + sizeof(struct drm_exynos_ipp_task_alpha), + offsetof(struct exynos_drm_ipp_task, alpha), + }, +}; + +static int exynos_drm_ipp_task_set(struct exynos_drm_ipp_task *task, + struct drm_exynos_ioctl_ipp_commit *arg) +{ + const struct exynos_drm_param_map *map = exynos_drm_ipp_params_maps; + void __user *params = (void __user *)(unsigned long)arg->params_ptr; + unsigned int size = arg->params_size; + uint32_t id; + int i; + + while (size) { + if (get_user(id, (uint32_t __user *)params)) + return -EFAULT; + + for (i = 0; i < ARRAY_SIZE(exynos_drm_ipp_params_maps); i++) + if (map[i].id == id) + break; + if (i == ARRAY_SIZE(exynos_drm_ipp_params_maps) || + map[i].size > size) + return -EINVAL; + + if (copy_from_user((void *)task + map[i].offset, params, + map[i].size)) + return -EFAULT; + + params += map[i].size; + size -= map[i].size; + } + + DRM_DEBUG_DRIVER("Got task %pK configuration from userspace\n", task); + return 0; +} + +static int exynos_drm_ipp_task_setup_buffer(struct exynos_drm_ipp_buffer *buf, + struct drm_file *filp) +{ + int ret = 0; + int i; + + /* basic checks */ + if (buf->buf.width == 0 || buf->buf.height == 0) + return -EINVAL; + buf->format = drm_format_info(buf->buf.fourcc); + for (i = 0; i < buf->format->num_planes; i++) { + unsigned int width = (i == 0) ? buf->buf.width : + DIV_ROUND_UP(buf->buf.width, buf->format->hsub); + + if (buf->buf.pitch[i] == 0) + buf->buf.pitch[i] = width * buf->format->cpp[i]; + if (buf->buf.pitch[i] < width * buf->format->cpp[i]) + return -EINVAL; + if (!buf->buf.gem_id[i]) + return -ENOENT; + } + + /* pitch for additional planes must match */ + if (buf->format->num_planes > 2 && + buf->buf.pitch[1] != buf->buf.pitch[2]) + return -EINVAL; + + /* get GEM buffers and check their size */ + for (i = 0; i < buf->format->num_planes; i++) { + unsigned int height = (i == 0) ? buf->buf.height : + DIV_ROUND_UP(buf->buf.height, buf->format->vsub); + unsigned long size = height * buf->buf.pitch[i]; + struct drm_gem_object *obj = drm_gem_object_lookup(filp, + buf->buf.gem_id[i]); + if (!obj) { + ret = -ENOENT; + goto gem_free; + } + buf->exynos_gem[i] = to_exynos_gem(obj); + + if (size + buf->buf.offset[i] > buf->exynos_gem[i]->size) { + i++; + ret = -EINVAL; + goto gem_free; + } + buf->dma_addr[i] = buf->exynos_gem[i]->dma_addr + + buf->buf.offset[i]; + } + + return 0; +gem_free: + while (i--) { + drm_gem_object_put_unlocked(&buf->exynos_gem[i]->base); + buf->exynos_gem[i] = NULL; + } + return ret; +} + +static void exynos_drm_ipp_task_release_buf(struct exynos_drm_ipp_buffer *buf) +{ + int i; + + if (!buf->exynos_gem[0]) + return; + for (i = 0; i < buf->format->num_planes; i++) + drm_gem_object_put_unlocked(&buf->exynos_gem[i]->base); +} + +static void exynos_drm_ipp_task_free(struct exynos_drm_ipp *ipp, + struct exynos_drm_ipp_task *task) +{ + DRM_DEBUG_DRIVER("Freeing task %pK\n", task); + + exynos_drm_ipp_task_release_buf(&task->src); + exynos_drm_ipp_task_release_buf(&task->dst); + if (task->event) + drm_event_cancel_free(ipp->dev, &task->event->base); + kfree(task); +} + +struct drm_ipp_limit { + struct drm_exynos_ipp_limit_val h; + struct drm_exynos_ipp_limit_val v; +}; + +enum drm_ipp_size_id { + IPP_LIMIT_BUFFER, IPP_LIMIT_AREA, IPP_LIMIT_ROTATED, IPP_LIMIT_MAX +}; + +static const enum drm_ipp_size_id limit_id_fallback[IPP_LIMIT_MAX][4] = { + [IPP_LIMIT_BUFFER] = { DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER }, + [IPP_LIMIT_AREA] = { DRM_EXYNOS_IPP_LIMIT_SIZE_AREA, + DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER }, + [IPP_LIMIT_ROTATED] = { DRM_EXYNOS_IPP_LIMIT_SIZE_ROTATED, + DRM_EXYNOS_IPP_LIMIT_SIZE_AREA, + DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER }, +}; + +static inline void __limit_set_val(unsigned int *ptr, unsigned int val) +{ + if (!*ptr) + *ptr = val; +} + +static void __get_size_limit(const struct drm_exynos_ipp_limit *limits, + unsigned int num_limits, enum drm_ipp_size_id id, + struct drm_ipp_limit *res) +{ + const struct drm_exynos_ipp_limit *l = limits; + int i = 0; + + memset(res, 0, sizeof(*res)); + for (i = 0; limit_id_fallback[id][i]; i++) + for (l = limits; l - limits < num_limits; l++) { + if (((l->type & DRM_EXYNOS_IPP_LIMIT_TYPE_MASK) != + DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE) || + ((l->type & DRM_EXYNOS_IPP_LIMIT_SIZE_MASK) != + limit_id_fallback[id][i])) + continue; + __limit_set_val(&res->h.min, l->h.min); + __limit_set_val(&res->h.max, l->h.max); + __limit_set_val(&res->h.align, l->h.align); + __limit_set_val(&res->v.min, l->v.min); + __limit_set_val(&res->v.max, l->v.max); + __limit_set_val(&res->v.align, l->v.align); + } +} + +static inline bool __align_check(unsigned int val, unsigned int align) +{ + if (align && (val & (align - 1))) { + DRM_DEBUG_DRIVER("Value %d exceeds HW limits (align %d)\n", + val, align); + return false; + } + return true; +} + +static inline bool __size_limit_check(unsigned int val, + struct drm_exynos_ipp_limit_val *l) +{ + if ((l->min && val < l->min) || (l->max && val > l->max)) { + DRM_DEBUG_DRIVER("Value %d exceeds HW limits (min %d, max %d)\n", + val, l->min, l->max); + return false; + } + return __align_check(val, l->align); +} + +static int exynos_drm_ipp_check_size_limits(struct exynos_drm_ipp_buffer *buf, + const struct drm_exynos_ipp_limit *limits, unsigned int num_limits, + bool rotate, bool swap) +{ + enum drm_ipp_size_id id = rotate ? IPP_LIMIT_ROTATED : IPP_LIMIT_AREA; + struct drm_ipp_limit l; + struct drm_exynos_ipp_limit_val *lh = &l.h, *lv = &l.v; + + if (!limits) + return 0; + + __get_size_limit(limits, num_limits, IPP_LIMIT_BUFFER, &l); + if (!__size_limit_check(buf->buf.width, &l.h) || + !__size_limit_check(buf->buf.height, &l.v)) + return -EINVAL; + + if (swap) { + lv = &l.h; + lh = &l.v; + } + __get_size_limit(limits, num_limits, id, &l); + if (!__size_limit_check(buf->rect.w, lh) || + !__align_check(buf->rect.x, lh->align) || + !__size_limit_check(buf->rect.h, lv) || + !__align_check(buf->rect.y, lv->align)) + return -EINVAL; + + return 0; +} + +static inline bool __scale_limit_check(unsigned int src, unsigned int dst, + unsigned int min, unsigned int max) +{ + if ((max && (dst << 16) > src * max) || + (min && (dst << 16) < src * min)) { + DRM_DEBUG_DRIVER("Scale from %d to %d exceeds HW limits (ratio min %d.%05d, max %d.%05d)\n", + src, dst, + min >> 16, 100000 * (min & 0xffff) / (1 << 16), + max >> 16, 100000 * (max & 0xffff) / (1 << 16)); + return false; + } + return true; +} + +static int exynos_drm_ipp_check_scale_limits( + struct drm_exynos_ipp_task_rect *src, + struct drm_exynos_ipp_task_rect *dst, + const struct drm_exynos_ipp_limit *limits, + unsigned int num_limits, bool swap) +{ + const struct drm_exynos_ipp_limit_val *lh, *lv; + int dw, dh; + + for (; num_limits; limits++, num_limits--) + if ((limits->type & DRM_EXYNOS_IPP_LIMIT_TYPE_MASK) == + DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE) + break; + if (!num_limits) + return 0; + + lh = (!swap) ? &limits->h : &limits->v; + lv = (!swap) ? &limits->v : &limits->h; + dw = (!swap) ? dst->w : dst->h; + dh = (!swap) ? dst->h : dst->w; + + if (!__scale_limit_check(src->w, dw, lh->min, lh->max) || + !__scale_limit_check(src->h, dh, lv->min, lv->max)) + return -EINVAL; + + return 0; +} + +static int exynos_drm_ipp_task_check(struct exynos_drm_ipp_task *task) +{ + struct exynos_drm_ipp *ipp = task->ipp; + const struct exynos_drm_ipp_formats *src_fmt, *dst_fmt; + struct exynos_drm_ipp_buffer *src = &task->src, *dst = &task->dst; + unsigned int rotation = task->transform.rotation; + int ret = 0; + bool swap = drm_rotation_90_or_270(rotation); + bool rotate = (rotation != DRM_MODE_ROTATE_0); + bool scale = false; + + DRM_DEBUG_DRIVER("Checking task %pK\n", task); + + if (src->rect.w == UINT_MAX) + src->rect.w = src->buf.width; + if (src->rect.h == UINT_MAX) + src->rect.h = src->buf.height; + if (dst->rect.w == UINT_MAX) + dst->rect.w = dst->buf.width; + if (dst->rect.h == UINT_MAX) + dst->rect.h = dst->buf.height; + + if (src->rect.x + src->rect.w > (src->buf.width) || + src->rect.y + src->rect.h > (src->buf.height) || + dst->rect.x + dst->rect.w > (dst->buf.width) || + dst->rect.y + dst->rect.h > (dst->buf.height)) { + DRM_DEBUG_DRIVER("Task %pK: defined area is outside provided buffers\n", + task); + return -EINVAL; + } + + if ((!swap && (src->rect.w != dst->rect.w || + src->rect.h != dst->rect.h)) || + (swap && (src->rect.w != dst->rect.h || + src->rect.h != dst->rect.w))) + scale = true; + + if ((!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_CROP) && + (src->rect.x || src->rect.y || dst->rect.x || dst->rect.y)) || + (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_ROTATE) && rotate) || + (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_SCALE) && scale) || + (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_CONVERT) && + src->buf.fourcc != dst->buf.fourcc)) { + DRM_DEBUG_DRIVER("Task %pK: hw capabilities exceeded\n", task); + return -EINVAL; + } + + src_fmt = __ipp_format_get(ipp, src->buf.fourcc, src->buf.modifier, + DRM_EXYNOS_IPP_FORMAT_SOURCE); + if (!src_fmt) { + DRM_DEBUG_DRIVER("Task %pK: src format not supported\n", task); + return -EINVAL; + } + ret = exynos_drm_ipp_check_size_limits(src, src_fmt->limits, + src_fmt->num_limits, + rotate, false); + if (ret) + return ret; + ret = exynos_drm_ipp_check_scale_limits(&src->rect, &dst->rect, + src_fmt->limits, + src_fmt->num_limits, swap); + if (ret) + return ret; + + dst_fmt = __ipp_format_get(ipp, dst->buf.fourcc, dst->buf.modifier, + DRM_EXYNOS_IPP_FORMAT_DESTINATION); + if (!dst_fmt) { + DRM_DEBUG_DRIVER("Task %pK: dst format not supported\n", task); + return -EINVAL; + } + ret = exynos_drm_ipp_check_size_limits(dst, dst_fmt->limits, + dst_fmt->num_limits, + false, swap); + if (ret) + return ret; + ret = exynos_drm_ipp_check_scale_limits(&src->rect, &dst->rect, + dst_fmt->limits, + dst_fmt->num_limits, swap); + if (ret) + return ret; + + DRM_DEBUG_DRIVER("Task %pK: all checks done.\n", task); + + return ret; +} + +static int exynos_drm_ipp_task_setup_buffers(struct exynos_drm_ipp_task *task, + struct drm_file *filp) +{ + struct exynos_drm_ipp_buffer *src = &task->src, *dst = &task->dst; + int ret = 0; + + DRM_DEBUG_DRIVER("Setting buffer for task %pK\n", task); + + ret = exynos_drm_ipp_task_setup_buffer(src, filp); + if (ret) { + DRM_DEBUG_DRIVER("Task %pK: src buffer setup failed\n", task); + return ret; + } + ret = exynos_drm_ipp_task_setup_buffer(dst, filp); + if (ret) { + DRM_DEBUG_DRIVER("Task %pK: dst buffer setup failed\n", task); + return ret; + } + + DRM_DEBUG_DRIVER("Task %pK: buffers prepared.\n", task); + + return ret; +} + + +static int exynos_drm_ipp_event_create(struct exynos_drm_ipp_task *task, + struct drm_file *file_priv, uint64_t user_data) +{ + struct drm_pending_exynos_ipp_event *e = NULL; + int ret; + + e = kzalloc(sizeof(*e), GFP_KERNEL); + if (!e) + return -ENOMEM; + + e->event.base.type = DRM_EXYNOS_IPP_EVENT; + e->event.base.length = sizeof(e->event); + e->event.user_data = user_data; + + ret = drm_event_reserve_init(task->dev, file_priv, &e->base, + &e->event.base); + if (ret) + goto free; + + task->event = e; + return 0; +free: + kfree(e); + return ret; +} + +static void exynos_drm_ipp_event_send(struct exynos_drm_ipp_task *task) +{ + struct timespec64 now; + + ktime_get_ts64(&now); + task->event->event.tv_sec = now.tv_sec; + task->event->event.tv_usec = now.tv_nsec / NSEC_PER_USEC; + task->event->event.sequence = atomic_inc_return(&task->ipp->sequence); + + drm_send_event(task->dev, &task->event->base); +} + +static int exynos_drm_ipp_task_cleanup(struct exynos_drm_ipp_task *task) +{ + int ret = task->ret; + + if (ret == 0 && task->event) { + exynos_drm_ipp_event_send(task); + /* ensure event won't be canceled on task free */ + task->event = NULL; + } + + exynos_drm_ipp_task_free(task->ipp, task); + return ret; +} + +static void exynos_drm_ipp_cleanup_work(struct work_struct *work) +{ + struct exynos_drm_ipp_task *task = container_of(work, + struct exynos_drm_ipp_task, cleanup_work); + + exynos_drm_ipp_task_cleanup(task); +} + +static void exynos_drm_ipp_next_task(struct exynos_drm_ipp *ipp); + +/** + * exynos_drm_ipp_task_done - finish given task and set return code + * @task: ipp task to finish + * @ret: error code or 0 if operation has been performed successfully + */ +void exynos_drm_ipp_task_done(struct exynos_drm_ipp_task *task, int ret) +{ + struct exynos_drm_ipp *ipp = task->ipp; + unsigned long flags; + + DRM_DEBUG_DRIVER("ipp: %d, task %pK done: %d\n", ipp->id, task, ret); + + spin_lock_irqsave(&ipp->lock, flags); + if (ipp->task == task) + ipp->task = NULL; + task->flags |= DRM_EXYNOS_IPP_TASK_DONE; + task->ret = ret; + spin_unlock_irqrestore(&ipp->lock, flags); + + exynos_drm_ipp_next_task(ipp); + wake_up(&ipp->done_wq); + + if (task->flags & DRM_EXYNOS_IPP_TASK_ASYNC) { + INIT_WORK(&task->cleanup_work, exynos_drm_ipp_cleanup_work); + schedule_work(&task->cleanup_work); + } +} + +static void exynos_drm_ipp_next_task(struct exynos_drm_ipp *ipp) +{ + struct exynos_drm_ipp_task *task; + unsigned long flags; + int ret; + + DRM_DEBUG_DRIVER("ipp: %d, try to run new task\n", ipp->id); + + spin_lock_irqsave(&ipp->lock, flags); + + if (ipp->task || list_empty(&ipp->todo_list)) { + spin_unlock_irqrestore(&ipp->lock, flags); + return; + } + + task = list_first_entry(&ipp->todo_list, struct exynos_drm_ipp_task, + head); + list_del_init(&task->head); + ipp->task = task; + + spin_unlock_irqrestore(&ipp->lock, flags); + + DRM_DEBUG_DRIVER("ipp: %d, selected task %pK to run\n", ipp->id, task); + + ret = ipp->funcs->commit(ipp, task); + if (ret) + exynos_drm_ipp_task_done(task, ret); +} + +static void exynos_drm_ipp_schedule_task(struct exynos_drm_ipp *ipp, + struct exynos_drm_ipp_task *task) +{ + unsigned long flags; + + spin_lock_irqsave(&ipp->lock, flags); + list_add(&task->head, &ipp->todo_list); + spin_unlock_irqrestore(&ipp->lock, flags); + + exynos_drm_ipp_next_task(ipp); +} + +static void exynos_drm_ipp_task_abort(struct exynos_drm_ipp *ipp, + struct exynos_drm_ipp_task *task) +{ + unsigned long flags; + + spin_lock_irqsave(&ipp->lock, flags); + if (task->flags & DRM_EXYNOS_IPP_TASK_DONE) { + /* already completed task */ + exynos_drm_ipp_task_cleanup(task); + } else if (ipp->task != task) { + /* task has not been scheduled for execution yet */ + list_del_init(&task->head); + exynos_drm_ipp_task_cleanup(task); + } else { + /* + * currently processed task, call abort() and perform + * cleanup with async worker + */ + task->flags |= DRM_EXYNOS_IPP_TASK_ASYNC; + spin_unlock_irqrestore(&ipp->lock, flags); + if (ipp->funcs->abort) + ipp->funcs->abort(ipp, task); + return; + } + spin_unlock_irqrestore(&ipp->lock, flags); +} + +/** + * exynos_drm_ipp_commit_ioctl - perform image processing operation + * @dev: DRM device + * @data: ioctl data + * @file_priv: DRM file info + * + * Construct a ipp task from the set of properties provided from the user + * and try to schedule it to framebuffer processor hardware. + * + * Called by the user via ioctl. + * + * Returns: + * Zero on success, negative errno on failure. + */ +int exynos_drm_ipp_commit_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + struct drm_exynos_ioctl_ipp_commit *arg = data; + struct exynos_drm_ipp *ipp; + struct exynos_drm_ipp_task *task; + int ret = 0; + + if ((arg->flags & ~DRM_EXYNOS_IPP_FLAGS) || arg->reserved) + return -EINVAL; + + /* can't test and expect an event at the same time */ + if ((arg->flags & DRM_EXYNOS_IPP_FLAG_TEST_ONLY) && + (arg->flags & DRM_EXYNOS_IPP_FLAG_EVENT)) + return -EINVAL; + + ipp = __ipp_get(arg->ipp_id); + if (!ipp) + return -ENOENT; + + task = exynos_drm_ipp_task_alloc(ipp); + if (!task) + return -ENOMEM; + + ret = exynos_drm_ipp_task_set(task, arg); + if (ret) + goto free; + + ret = exynos_drm_ipp_task_check(task); + if (ret) + goto free; + + ret = exynos_drm_ipp_task_setup_buffers(task, file_priv); + if (ret || arg->flags & DRM_EXYNOS_IPP_FLAG_TEST_ONLY) + goto free; + + if (arg->flags & DRM_EXYNOS_IPP_FLAG_EVENT) { + ret = exynos_drm_ipp_event_create(task, file_priv, + arg->user_data); + if (ret) + goto free; + } + + /* + * Queue task for processing on the hardware. task object will be + * then freed after exynos_drm_ipp_task_done() + */ + if (arg->flags & DRM_EXYNOS_IPP_FLAG_NONBLOCK) { + DRM_DEBUG_DRIVER("ipp: %d, nonblocking processing task %pK\n", + ipp->id, task); + + task->flags |= DRM_EXYNOS_IPP_TASK_ASYNC; + exynos_drm_ipp_schedule_task(task->ipp, task); + ret = 0; + } else { + DRM_DEBUG_DRIVER("ipp: %d, processing task %pK\n", ipp->id, + task); + exynos_drm_ipp_schedule_task(ipp, task); + ret = wait_event_interruptible(ipp->done_wq, + task->flags & DRM_EXYNOS_IPP_TASK_DONE); + if (ret) + exynos_drm_ipp_task_abort(ipp, task); + else + ret = exynos_drm_ipp_task_cleanup(task); + } + return ret; +free: + exynos_drm_ipp_task_free(ipp, task); + + return ret; +} diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.h b/drivers/gpu/drm/exynos/exynos_drm_ipp.h new file mode 100644 index 000000000000..0b27d4a9bf94 --- /dev/null +++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.h @@ -0,0 +1,175 @@ +/* + * Copyright (c) 2017 Samsung Electronics Co., Ltd. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#ifndef _EXYNOS_DRM_IPP_H_ +#define _EXYNOS_DRM_IPP_H_ + +#include + +struct exynos_drm_ipp; +struct exynos_drm_ipp_task; + +/** + * struct exynos_drm_ipp_funcs - exynos_drm_ipp control functions + */ +struct exynos_drm_ipp_funcs { + /** + * @commit: + * + * This is the main entry point to start framebuffer processing + * in the hardware. The exynos_drm_ipp_task has been already validated. + * This function must not wait until the device finishes processing. + * When the driver finishes processing, it has to call + * exynos_exynos_drm_ipp_task_done() function. + * + * RETURNS: + * + * 0 on success or negative error codes in case of failure. + */ + int (*commit)(struct exynos_drm_ipp *ipp, + struct exynos_drm_ipp_task *task); + + /** + * @abort: + * + * Informs the driver that it has to abort the currently running + * task as soon as possible (i.e. as soon as it can stop the device + * safely), even if the task would not have been finished by then. + * After the driver performs the necessary steps, it has to call + * exynos_drm_ipp_task_done() (as if the task ended normally). + * This function does not have to (and will usually not) wait + * until the device enters a state when it can be stopped. + */ + void (*abort)(struct exynos_drm_ipp *ipp, + struct exynos_drm_ipp_task *task); +}; + +/** + * struct exynos_drm_ipp - central picture processor module structure + */ +struct exynos_drm_ipp { + struct drm_device *dev; + struct list_head head; + unsigned int id; + + const char *name; + const struct exynos_drm_ipp_funcs *funcs; + unsigned int capabilities; + const struct exynos_drm_ipp_formats *formats; + unsigned int num_formats; + atomic_t sequence; + + spinlock_t lock; + struct exynos_drm_ipp_task *task; + struct list_head todo_list; + wait_queue_head_t done_wq; +}; + +struct exynos_drm_ipp_buffer { + struct drm_exynos_ipp_task_buffer buf; + struct drm_exynos_ipp_task_rect rect; + + struct exynos_drm_gem *exynos_gem[MAX_FB_BUFFER]; + const struct drm_format_info *format; + dma_addr_t dma_addr[MAX_FB_BUFFER]; +}; + +/** + * struct exynos_drm_ipp_task - a structure describing transformation that + * has to be performed by the picture processor hardware module + */ +struct exynos_drm_ipp_task { + struct drm_device *dev; + struct exynos_drm_ipp *ipp; + struct list_head head; + + struct exynos_drm_ipp_buffer src; + struct exynos_drm_ipp_buffer dst; + + struct drm_exynos_ipp_task_transform transform; + struct drm_exynos_ipp_task_alpha alpha; + + struct work_struct cleanup_work; + unsigned int flags; + int ret; + + struct drm_pending_exynos_ipp_event *event; +}; + +#define DRM_EXYNOS_IPP_TASK_DONE (1 << 0) +#define DRM_EXYNOS_IPP_TASK_ASYNC (1 << 1) + +struct exynos_drm_ipp_formats { + uint32_t fourcc; + uint32_t type; + uint64_t modifier; + const struct drm_exynos_ipp_limit *limits; + unsigned int num_limits; +}; + +/* helper macros to set exynos_drm_ipp_formats structure and limits*/ +#define IPP_SRCDST_MFORMAT(f, m, l) \ + .fourcc = DRM_FORMAT_##f, .modifier = m, .limits = l, \ + .num_limits = ARRAY_SIZE(l), \ + .type = (DRM_EXYNOS_IPP_FORMAT_SOURCE | \ + DRM_EXYNOS_IPP_FORMAT_DESTINATION) + +#define IPP_SRCDST_FORMAT(f, l) IPP_SRCDST_MFORMAT(f, 0, l) + +#define IPP_SIZE_LIMIT(l, val...) \ + .type = (DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE | \ + DRM_EXYNOS_IPP_LIMIT_SIZE_##l), val + +#define IPP_SCALE_LIMIT(val...) \ + .type = (DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE), val + +int exynos_drm_ipp_register(struct drm_device *dev, struct exynos_drm_ipp *ipp, + const struct exynos_drm_ipp_funcs *funcs, unsigned int caps, + const struct exynos_drm_ipp_formats *formats, + unsigned int num_formats, const char *name); +void exynos_drm_ipp_unregister(struct drm_device *dev, + struct exynos_drm_ipp *ipp); + +void exynos_drm_ipp_task_done(struct exynos_drm_ipp_task *task, int ret); + +#ifdef CONFIG_DRM_EXYNOS_IPP +int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv); +int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv); +int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, void *data, + struct drm_file *file_priv); +int exynos_drm_ipp_commit_ioctl(struct drm_device *dev, + void *data, struct drm_file *file_priv); +#else +static inline int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, + void *data, struct drm_file *file_priv) +{ + struct drm_exynos_ioctl_ipp_get_res *resp = data; + + resp->count_ipps = 0; + return 0; +} +static inline int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, + void *data, struct drm_file *file_priv) +{ + return -ENODEV; +} +static inline int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, + void *data, struct drm_file *file_priv) +{ + return -ENODEV; +} +static inline int exynos_drm_ipp_commit_ioctl(struct drm_device *dev, + void *data, struct drm_file *file_priv) +{ + return -ENODEV; +} +#endif +#endif diff --git a/include/uapi/drm/exynos_drm.h b/include/uapi/drm/exynos_drm.h index 4a54305120e0..3e59b8382dd8 100644 --- a/include/uapi/drm/exynos_drm.h +++ b/include/uapi/drm/exynos_drm.h @@ -135,6 +135,219 @@ struct drm_exynos_g2d_exec { __u64 async; }; +/* Exynos DRM IPP v2 API */ + +/** + * Enumerate available IPP hardware modules. + * + * @count_ipps: size of ipp_id array / number of ipp modules (set by driver) + * @reserved: padding + * @ipp_id_ptr: pointer to ipp_id array or NULL + */ +struct drm_exynos_ioctl_ipp_get_res { + __u32 count_ipps; + __u32 reserved; + __u64 ipp_id_ptr; +}; + +enum drm_exynos_ipp_format_type { + DRM_EXYNOS_IPP_FORMAT_SOURCE = 0x01, + DRM_EXYNOS_IPP_FORMAT_DESTINATION = 0x02, +}; + +struct drm_exynos_ipp_format { + __u32 fourcc; + __u32 type; + __u64 modifier; +}; + +enum drm_exynos_ipp_capability { + DRM_EXYNOS_IPP_CAP_CROP = 0x01, + DRM_EXYNOS_IPP_CAP_ROTATE = 0x02, + DRM_EXYNOS_IPP_CAP_SCALE = 0x04, + DRM_EXYNOS_IPP_CAP_CONVERT = 0x08, +}; + +/** + * Get IPP hardware capabilities and supported image formats. + * + * @ipp_id: id of IPP module to query + * @capabilities: bitmask of drm_exynos_ipp_capability (set by driver) + * @reserved: padding + * @formats_count: size of formats array (in entries) / number of filled + * formats (set by driver) + * @formats_ptr: pointer to formats array or NULL + */ +struct drm_exynos_ioctl_ipp_get_caps { + __u32 ipp_id; + __u32 capabilities; + __u32 reserved; + __u32 formats_count; + __u64 formats_ptr; +}; + +enum drm_exynos_ipp_limit_type { + /* size (horizontal/vertial) limits, in pixels (min, max, alignment) */ + DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE = 0x0001, + /* scale ratio (horizonta/vertial), 16.16 fixed point (min, max) */ + DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE = 0x0002, + + /* image buffer area */ + DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER = 0x0001 << 16, + /* src/dst rectangle area */ + DRM_EXYNOS_IPP_LIMIT_SIZE_AREA = 0x0002 << 16, + /* src/dst rectangle area when rotation enabled */ + DRM_EXYNOS_IPP_LIMIT_SIZE_ROTATED = 0x0003 << 16, + + DRM_EXYNOS_IPP_LIMIT_TYPE_MASK = 0x000f, + DRM_EXYNOS_IPP_LIMIT_SIZE_MASK = 0x000f << 16, +}; + +struct drm_exynos_ipp_limit_val { + __u32 min; + __u32 max; + __u32 align; + __u32 reserved; +}; + +/** + * IPP module limitation. + * + * @type: limit type (see drm_exynos_ipp_limit_type enum) + * @reserved: padding + * @h: horizontal limits + * @v: vertical limits + */ +struct drm_exynos_ipp_limit { + __u32 type; + __u32 reserved; + struct drm_exynos_ipp_limit_val h; + struct drm_exynos_ipp_limit_val v; +}; + +/** + * Get IPP limits for given image format. + * + * @ipp_id: id of IPP module to query + * @fourcc: image format code (see DRM_FORMAT_* in drm_fourcc.h) + * @modifier: image format modifier (see DRM_FORMAT_MOD_* in drm_fourcc.h) + * @type: source/destination identifier (drm_exynos_ipp_format_flag enum) + * @limits_count: size of limits array (in entries) / number of filled entries + * (set by driver) + * @limits_ptr: pointer to limits array or NULL + */ +struct drm_exynos_ioctl_ipp_get_limits { + __u32 ipp_id; + __u32 fourcc; + __u64 modifier; + __u32 type; + __u32 limits_count; + __u64 limits_ptr; +}; + +enum drm_exynos_ipp_task_id { + /* buffer described by struct drm_exynos_ipp_task_buffer */ + DRM_EXYNOS_IPP_TASK_BUFFER = 0x0001, + /* rectangle described by struct drm_exynos_ipp_task_rect */ + DRM_EXYNOS_IPP_TASK_RECTANGLE = 0x0002, + /* transformation described by struct drm_exynos_ipp_task_transform */ + DRM_EXYNOS_IPP_TASK_TRANSFORM = 0x0003, + /* alpha configuration described by struct drm_exynos_ipp_task_alpha */ + DRM_EXYNOS_IPP_TASK_ALPHA = 0x0004, + + /* source image data (for buffer and rectangle chunks) */ + DRM_EXYNOS_IPP_TASK_TYPE_SOURCE = 0x0001 << 16, + /* destination image data (for buffer and rectangle chunks) */ + DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION = 0x0002 << 16, +}; + +/** + * Memory buffer with image data. + * + * @id: must be DRM_EXYNOS_IPP_TASK_BUFFER + * other parameters are same as for AddFB2 generic DRM ioctl + */ +struct drm_exynos_ipp_task_buffer { + __u32 id; + __u32 fourcc; + __u32 width, height; + __u32 gem_id[4]; + __u32 offset[4]; + __u32 pitch[4]; + __u64 modifier; +}; + +/** + * Rectangle for processing. + * + * @id: must be DRM_EXYNOS_IPP_TASK_RECTANGLE + * @reserved: padding + * @x,@y: left corner in pixels + * @w,@h: width/height in pixels + */ +struct drm_exynos_ipp_task_rect { + __u32 id; + __u32 reserved; + __u32 x; + __u32 y; + __u32 w; + __u32 h; +}; + +/** + * Image tranformation description. + * + * @id: must be DRM_EXYNOS_IPP_TASK_TRANSFORM + * @rotation: DRM_MODE_ROTATE_* and DRM_MODE_REFLECT_* values + */ +struct drm_exynos_ipp_task_transform { + __u32 id; + __u32 rotation; +}; + +/** + * Image global alpha configuration for formats without alpha values. + * + * @id: must be DRM_EXYNOS_IPP_TASK_ALPHA + * @value: global alpha value (0-255) + */ +struct drm_exynos_ipp_task_alpha { + __u32 id; + __u32 value; +}; + +enum drm_exynos_ipp_flag { + /* generate DRM event after processing */ + DRM_EXYNOS_IPP_FLAG_EVENT = 0x01, + /* dry run, only check task parameters */ + DRM_EXYNOS_IPP_FLAG_TEST_ONLY = 0x02, + /* non-blocking processing */ + DRM_EXYNOS_IPP_FLAG_NONBLOCK = 0x04, +}; + +#define DRM_EXYNOS_IPP_FLAGS (DRM_EXYNOS_IPP_FLAG_EVENT |\ + DRM_EXYNOS_IPP_FLAG_TEST_ONLY | DRM_EXYNOS_IPP_FLAG_NONBLOCK) + +/** + * Perform image processing described by array of drm_exynos_ipp_task_* + * structures (parameters array). + * + * @ipp_id: id of IPP module to run the task + * @flags: bitmask of drm_exynos_ipp_flag values + * @reserved: padding + * @params_size: size of parameters array (in bytes) + * @params_ptr: pointer to parameters array or NULL + * @user_data: (optional) data for drm event + */ +struct drm_exynos_ioctl_ipp_commit { + __u32 ipp_id; + __u32 flags; + __u32 reserved; + __u32 params_size; + __u64 params_ptr; + __u64 user_data; +}; + #define DRM_EXYNOS_GEM_CREATE 0x00 #define DRM_EXYNOS_GEM_MAP 0x01 /* Reserved 0x03 ~ 0x05 for exynos specific gem ioctl */ @@ -147,6 +360,11 @@ struct drm_exynos_g2d_exec { #define DRM_EXYNOS_G2D_EXEC 0x22 /* Reserved 0x30 ~ 0x33 for obsolete Exynos IPP ioctls */ +/* IPP - Image Post Processing */ +#define DRM_EXYNOS_IPP_GET_RESOURCES 0x40 +#define DRM_EXYNOS_IPP_GET_CAPS 0x41 +#define DRM_EXYNOS_IPP_GET_LIMITS 0x42 +#define DRM_EXYNOS_IPP_COMMIT 0x43 #define DRM_IOCTL_EXYNOS_GEM_CREATE DRM_IOWR(DRM_COMMAND_BASE + \ DRM_EXYNOS_GEM_CREATE, struct drm_exynos_gem_create) @@ -165,8 +383,20 @@ struct drm_exynos_g2d_exec { #define DRM_IOCTL_EXYNOS_G2D_EXEC DRM_IOWR(DRM_COMMAND_BASE + \ DRM_EXYNOS_G2D_EXEC, struct drm_exynos_g2d_exec) +#define DRM_IOCTL_EXYNOS_IPP_GET_RESOURCES DRM_IOWR(DRM_COMMAND_BASE + \ + DRM_EXYNOS_IPP_GET_RESOURCES, \ + struct drm_exynos_ioctl_ipp_get_res) +#define DRM_IOCTL_EXYNOS_IPP_GET_CAPS DRM_IOWR(DRM_COMMAND_BASE + \ + DRM_EXYNOS_IPP_GET_CAPS, struct drm_exynos_ioctl_ipp_get_caps) +#define DRM_IOCTL_EXYNOS_IPP_GET_LIMITS DRM_IOWR(DRM_COMMAND_BASE + \ + DRM_EXYNOS_IPP_GET_LIMITS, \ + struct drm_exynos_ioctl_ipp_get_limits) +#define DRM_IOCTL_EXYNOS_IPP_COMMIT DRM_IOWR(DRM_COMMAND_BASE + \ + DRM_EXYNOS_IPP_COMMIT, struct drm_exynos_ioctl_ipp_commit) + /* EXYNOS specific events */ #define DRM_EXYNOS_G2D_EVENT 0x80000000 +#define DRM_EXYNOS_IPP_EVENT 0x80000002 struct drm_exynos_g2d_event { struct drm_event base; @@ -177,6 +407,16 @@ struct drm_exynos_g2d_event { __u32 reserved; }; +struct drm_exynos_ipp_event { + struct drm_event base; + __u64 user_data; + __u32 tv_sec; + __u32 tv_usec; + __u32 ipp_id; + __u32 sequence; + __u64 reserved; +}; + #if defined(__cplusplus) } #endif -- cgit v1.2.3 From 0010f7052d6cb71c4b120238e28cd3fa413913d1 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 4 May 2018 16:57:30 +0200 Subject: libceph: add osd_req_op_extent_osd_data_bvecs() ... and store num_bvecs for client code's convenience. Signed-off-by: Ilya Dryomov Reviewed-by: Jeff Layton Reviewed-by: "Yan, Zheng" --- drivers/block/rbd.c | 4 +++- include/linux/ceph/osd_client.h | 12 ++++++++++-- net/ceph/osd_client.c | 27 +++++++++++++++++++++++---- 3 files changed, 36 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 8e8b04cc569a..33b36fea1d73 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2366,7 +2366,9 @@ static int rbd_obj_issue_copyup(struct rbd_obj_request *obj_req, u32 bytes) osd_req_op_cls_init(obj_req->osd_req, 0, CEPH_OSD_OP_CALL, "rbd", "copyup"); osd_req_op_cls_request_data_bvecs(obj_req->osd_req, 0, - obj_req->copyup_bvecs, bytes); + obj_req->copyup_bvecs, + obj_req->copyup_bvec_count, + bytes); switch (obj_req->img_request->op_type) { case OBJ_OP_WRITE: diff --git a/include/linux/ceph/osd_client.h b/include/linux/ceph/osd_client.h index 528ccc943cee..96bb32285989 100644 --- a/include/linux/ceph/osd_client.h +++ b/include/linux/ceph/osd_client.h @@ -77,7 +77,10 @@ struct ceph_osd_data { u32 bio_length; }; #endif /* CONFIG_BLOCK */ - struct ceph_bvec_iter bvec_pos; + struct { + struct ceph_bvec_iter bvec_pos; + u32 num_bvecs; + }; }; }; @@ -412,6 +415,10 @@ void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req, struct ceph_bio_iter *bio_pos, u32 bio_length); #endif /* CONFIG_BLOCK */ +void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req, + unsigned int which, + struct bio_vec *bvecs, u32 num_bvecs, + u32 bytes); void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req, unsigned int which, struct ceph_bvec_iter *bvec_pos); @@ -426,7 +433,8 @@ extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *, bool own_pages); void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req, unsigned int which, - struct bio_vec *bvecs, u32 bytes); + struct bio_vec *bvecs, u32 num_bvecs, + u32 bytes); extern void osd_req_op_cls_response_data_pages(struct ceph_osd_request *, unsigned int which, struct page **pages, u64 length, diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c index ea2a6c9fb7ce..d2667e5dddc3 100644 --- a/net/ceph/osd_client.c +++ b/net/ceph/osd_client.c @@ -157,10 +157,12 @@ static void ceph_osd_data_bio_init(struct ceph_osd_data *osd_data, #endif /* CONFIG_BLOCK */ static void ceph_osd_data_bvecs_init(struct ceph_osd_data *osd_data, - struct ceph_bvec_iter *bvec_pos) + struct ceph_bvec_iter *bvec_pos, + u32 num_bvecs) { osd_data->type = CEPH_OSD_DATA_TYPE_BVECS; osd_data->bvec_pos = *bvec_pos; + osd_data->num_bvecs = num_bvecs; } #define osd_req_op_data(oreq, whch, typ, fld) \ @@ -237,6 +239,22 @@ void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req, EXPORT_SYMBOL(osd_req_op_extent_osd_data_bio); #endif /* CONFIG_BLOCK */ +void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req, + unsigned int which, + struct bio_vec *bvecs, u32 num_bvecs, + u32 bytes) +{ + struct ceph_osd_data *osd_data; + struct ceph_bvec_iter it = { + .bvecs = bvecs, + .iter = { .bi_size = bytes }, + }; + + osd_data = osd_req_op_data(osd_req, which, extent, osd_data); + ceph_osd_data_bvecs_init(osd_data, &it, num_bvecs); +} +EXPORT_SYMBOL(osd_req_op_extent_osd_data_bvecs); + void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req, unsigned int which, struct ceph_bvec_iter *bvec_pos) @@ -244,7 +262,7 @@ void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req, struct ceph_osd_data *osd_data; osd_data = osd_req_op_data(osd_req, which, extent, osd_data); - ceph_osd_data_bvecs_init(osd_data, bvec_pos); + ceph_osd_data_bvecs_init(osd_data, bvec_pos, 0); } EXPORT_SYMBOL(osd_req_op_extent_osd_data_bvec_pos); @@ -287,7 +305,8 @@ EXPORT_SYMBOL(osd_req_op_cls_request_data_pages); void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req, unsigned int which, - struct bio_vec *bvecs, u32 bytes) + struct bio_vec *bvecs, u32 num_bvecs, + u32 bytes) { struct ceph_osd_data *osd_data; struct ceph_bvec_iter it = { @@ -296,7 +315,7 @@ void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req, }; osd_data = osd_req_op_data(osd_req, which, cls, request_data); - ceph_osd_data_bvecs_init(osd_data, &it); + ceph_osd_data_bvecs_init(osd_data, &it, num_bvecs); osd_req->r_ops[which].cls.indata_len += bytes; osd_req->r_ops[which].indata_len += bytes; } -- cgit v1.2.3 From 494337c918403db5743af0348d963cafb9b20cee Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 10 May 2018 23:26:01 +0100 Subject: rxrpc: Add a tracepoint to log ICMP/ICMP6 and error messages Add a tracepoint to log received ICMP/ICMP6 events and other error messages. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 30 +++++++++++++++++++++++++++++ net/rxrpc/peer_event.c | 46 ++++++++++++++++++++++---------------------- 2 files changed, 53 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 9e96c2fe2793..497d0b67f421 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -15,6 +15,7 @@ #define _TRACE_RXRPC_H #include +#include /* * Define enums for tracing information. @@ -1374,6 +1375,35 @@ TRACE_EVENT(rxrpc_resend, __entry->anno) ); +TRACE_EVENT(rxrpc_rx_icmp, + TP_PROTO(struct rxrpc_peer *peer, struct sock_extended_err *ee, + struct sockaddr_rxrpc *srx), + + TP_ARGS(peer, ee, srx), + + TP_STRUCT__entry( + __field(unsigned int, peer ) + __field_struct(struct sock_extended_err, ee ) + __field_struct(struct sockaddr_rxrpc, srx ) + ), + + TP_fast_assign( + __entry->peer = peer->debug_id; + memcpy(&__entry->ee, ee, sizeof(__entry->ee)); + memcpy(&__entry->srx, srx, sizeof(__entry->srx)); + ), + + TP_printk("P=%08x o=%u t=%u c=%u i=%u d=%u e=%d %pISp", + __entry->peer, + __entry->ee.ee_origin, + __entry->ee.ee_type, + __entry->ee.ee_code, + __entry->ee.ee_info, + __entry->ee.ee_data, + __entry->ee.ee_errno, + &__entry->srx.transport) + ); + #endif /* _TRACE_RXRPC_H */ /* This part must be outside protection */ diff --git a/net/rxrpc/peer_event.c b/net/rxrpc/peer_event.c index 78c2f95d1f22..0ed8b651cec2 100644 --- a/net/rxrpc/peer_event.c +++ b/net/rxrpc/peer_event.c @@ -28,39 +28,39 @@ static void rxrpc_store_error(struct rxrpc_peer *, struct sock_exterr_skb *); * Find the peer associated with an ICMP packet. */ static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local, - const struct sk_buff *skb) + const struct sk_buff *skb, + struct sockaddr_rxrpc *srx) { struct sock_exterr_skb *serr = SKB_EXT_ERR(skb); - struct sockaddr_rxrpc srx; _enter(""); - memset(&srx, 0, sizeof(srx)); - srx.transport_type = local->srx.transport_type; - srx.transport_len = local->srx.transport_len; - srx.transport.family = local->srx.transport.family; + memset(srx, 0, sizeof(*srx)); + srx->transport_type = local->srx.transport_type; + srx->transport_len = local->srx.transport_len; + srx->transport.family = local->srx.transport.family; /* Can we see an ICMP4 packet on an ICMP6 listening socket? and vice * versa? */ - switch (srx.transport.family) { + switch (srx->transport.family) { case AF_INET: - srx.transport.sin.sin_port = serr->port; + srx->transport.sin.sin_port = serr->port; switch (serr->ee.ee_origin) { case SO_EE_ORIGIN_ICMP: _net("Rx ICMP"); - memcpy(&srx.transport.sin.sin_addr, + memcpy(&srx->transport.sin.sin_addr, skb_network_header(skb) + serr->addr_offset, sizeof(struct in_addr)); break; case SO_EE_ORIGIN_ICMP6: _net("Rx ICMP6 on v4 sock"); - memcpy(&srx.transport.sin.sin_addr, + memcpy(&srx->transport.sin.sin_addr, skb_network_header(skb) + serr->addr_offset + 12, sizeof(struct in_addr)); break; default: - memcpy(&srx.transport.sin.sin_addr, &ip_hdr(skb)->saddr, + memcpy(&srx->transport.sin.sin_addr, &ip_hdr(skb)->saddr, sizeof(struct in_addr)); break; } @@ -68,25 +68,25 @@ static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local, #ifdef CONFIG_AF_RXRPC_IPV6 case AF_INET6: - srx.transport.sin6.sin6_port = serr->port; + srx->transport.sin6.sin6_port = serr->port; switch (serr->ee.ee_origin) { case SO_EE_ORIGIN_ICMP6: _net("Rx ICMP6"); - memcpy(&srx.transport.sin6.sin6_addr, + memcpy(&srx->transport.sin6.sin6_addr, skb_network_header(skb) + serr->addr_offset, sizeof(struct in6_addr)); break; case SO_EE_ORIGIN_ICMP: _net("Rx ICMP on v6 sock"); - srx.transport.sin6.sin6_addr.s6_addr32[0] = 0; - srx.transport.sin6.sin6_addr.s6_addr32[1] = 0; - srx.transport.sin6.sin6_addr.s6_addr32[2] = htonl(0xffff); - memcpy(srx.transport.sin6.sin6_addr.s6_addr + 12, + srx->transport.sin6.sin6_addr.s6_addr32[0] = 0; + srx->transport.sin6.sin6_addr.s6_addr32[1] = 0; + srx->transport.sin6.sin6_addr.s6_addr32[2] = htonl(0xffff); + memcpy(srx->transport.sin6.sin6_addr.s6_addr + 12, skb_network_header(skb) + serr->addr_offset, sizeof(struct in_addr)); break; default: - memcpy(&srx.transport.sin6.sin6_addr, + memcpy(&srx->transport.sin6.sin6_addr, &ipv6_hdr(skb)->saddr, sizeof(struct in6_addr)); break; @@ -98,7 +98,7 @@ static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local, BUG(); } - return rxrpc_lookup_peer_rcu(local, &srx); + return rxrpc_lookup_peer_rcu(local, srx); } /* @@ -146,6 +146,7 @@ static void rxrpc_adjust_mtu(struct rxrpc_peer *peer, struct sock_exterr_skb *se void rxrpc_error_report(struct sock *sk) { struct sock_exterr_skb *serr; + struct sockaddr_rxrpc srx; struct rxrpc_local *local = sk->sk_user_data; struct rxrpc_peer *peer; struct sk_buff *skb; @@ -166,7 +167,7 @@ void rxrpc_error_report(struct sock *sk) } rcu_read_lock(); - peer = rxrpc_lookup_peer_icmp_rcu(local, skb); + peer = rxrpc_lookup_peer_icmp_rcu(local, skb, &srx); if (peer && !rxrpc_get_peer_maybe(peer)) peer = NULL; if (!peer) { @@ -176,6 +177,8 @@ void rxrpc_error_report(struct sock *sk) return; } + trace_rxrpc_rx_icmp(peer, &serr->ee, &srx); + if ((serr->ee.ee_origin == SO_EE_ORIGIN_ICMP && serr->ee.ee_type == ICMP_DEST_UNREACH && serr->ee.ee_code == ICMP_FRAG_NEEDED)) { @@ -209,9 +212,6 @@ static void rxrpc_store_error(struct rxrpc_peer *peer, ee = &serr->ee; - _net("Rx Error o=%d t=%d c=%d e=%d", - ee->ee_origin, ee->ee_type, ee->ee_code, ee->ee_errno); - err = ee->ee_errno; switch (ee->ee_origin) { -- cgit v1.2.3 From 6b47fe1d1ca3aec3a1a8623439c22fbf51016cd8 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 10 May 2018 23:26:01 +0100 Subject: rxrpc: Trace UDP transmission failure Add a tracepoint to log transmission failure from the UDP transport socket being used by AF_RXRPC. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 55 ++++++++++++++++++++++++++++++++++++++++++++ net/rxrpc/conn_event.c | 11 ++++++--- net/rxrpc/local_event.c | 3 ++- net/rxrpc/output.c | 23 ++++++++++++++++-- net/rxrpc/rxkad.c | 6 +++-- 5 files changed, 90 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 497d0b67f421..077e664ac9a2 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -211,6 +211,20 @@ enum rxrpc_congest_change { rxrpc_cong_saw_nack, }; +enum rxrpc_tx_fail_trace { + rxrpc_tx_fail_call_abort, + rxrpc_tx_fail_call_ack, + rxrpc_tx_fail_call_data_frag, + rxrpc_tx_fail_call_data_nofrag, + rxrpc_tx_fail_call_final_resend, + rxrpc_tx_fail_conn_abort, + rxrpc_tx_fail_conn_challenge, + rxrpc_tx_fail_conn_response, + rxrpc_tx_fail_reject, + rxrpc_tx_fail_version_keepalive, + rxrpc_tx_fail_version_reply, +}; + #endif /* end __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY */ /* @@ -438,6 +452,19 @@ enum rxrpc_congest_change { EM(RXRPC_CALL_LOCAL_ERROR, "LocalError") \ E_(RXRPC_CALL_NETWORK_ERROR, "NetError") +#define rxrpc_tx_fail_traces \ + EM(rxrpc_tx_fail_call_abort, "CallAbort") \ + EM(rxrpc_tx_fail_call_ack, "CallAck") \ + EM(rxrpc_tx_fail_call_data_frag, "CallDataFrag") \ + EM(rxrpc_tx_fail_call_data_nofrag, "CallDataNofrag") \ + EM(rxrpc_tx_fail_call_final_resend, "CallFinalResend") \ + EM(rxrpc_tx_fail_conn_abort, "ConnAbort") \ + EM(rxrpc_tx_fail_conn_challenge, "ConnChall") \ + EM(rxrpc_tx_fail_conn_response, "ConnResp") \ + EM(rxrpc_tx_fail_reject, "Reject") \ + EM(rxrpc_tx_fail_version_keepalive, "VerKeepalive") \ + E_(rxrpc_tx_fail_version_reply, "VerReply") + /* * Export enum symbols via userspace. */ @@ -461,6 +488,7 @@ rxrpc_propose_ack_traces; rxrpc_propose_ack_outcomes; rxrpc_congest_modes; rxrpc_congest_changes; +rxrpc_tx_fail_traces; /* * Now redefine the EM() and E_() macros to map the enums to the strings that @@ -1404,6 +1432,33 @@ TRACE_EVENT(rxrpc_rx_icmp, &__entry->srx.transport) ); +TRACE_EVENT(rxrpc_tx_fail, + TP_PROTO(unsigned int debug_id, rxrpc_serial_t serial, int ret, + enum rxrpc_tx_fail_trace what), + + TP_ARGS(debug_id, serial, ret, what), + + TP_STRUCT__entry( + __field(unsigned int, debug_id ) + __field(rxrpc_serial_t, serial ) + __field(int, ret ) + __field(enum rxrpc_tx_fail_trace, what ) + ), + + TP_fast_assign( + __entry->debug_id = debug_id; + __entry->serial = serial; + __entry->ret = ret; + __entry->what = what; + ), + + TP_printk("c=%08x r=%x ret=%d %s", + __entry->debug_id, + __entry->serial, + __entry->ret, + __print_symbolic(__entry->what, rxrpc_tx_fail_traces)) + ); + #endif /* _TRACE_RXRPC_H */ /* This part must be outside protection */ diff --git a/net/rxrpc/conn_event.c b/net/rxrpc/conn_event.c index c717152070df..1350f1be8037 100644 --- a/net/rxrpc/conn_event.c +++ b/net/rxrpc/conn_event.c @@ -40,7 +40,7 @@ static void rxrpc_conn_retransmit_call(struct rxrpc_connection *conn, } __attribute__((packed)) pkt; struct rxrpc_ackinfo ack_info; size_t len; - int ioc; + int ret, ioc; u32 serial, mtu, call_id, padding; _enter("%d", conn->debug_id); @@ -135,10 +135,13 @@ static void rxrpc_conn_retransmit_call(struct rxrpc_connection *conn, break; } - kernel_sendmsg(conn->params.local->socket, &msg, iov, ioc, len); + ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, ioc, len); conn->params.peer->last_tx_at = ktime_get_real(); + if (ret < 0) + trace_rxrpc_tx_fail(conn->debug_id, serial, ret, + rxrpc_tx_fail_call_final_resend); + _leave(""); - return; } /* @@ -236,6 +239,8 @@ static int rxrpc_abort_connection(struct rxrpc_connection *conn, ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len); if (ret < 0) { + trace_rxrpc_tx_fail(conn->debug_id, serial, ret, + rxrpc_tx_fail_conn_abort); _debug("sendmsg failed: %d", ret); return -EAGAIN; } diff --git a/net/rxrpc/local_event.c b/net/rxrpc/local_event.c index 93b5d910b4a1..8325f1b86840 100644 --- a/net/rxrpc/local_event.c +++ b/net/rxrpc/local_event.c @@ -71,7 +71,8 @@ static void rxrpc_send_version_request(struct rxrpc_local *local, ret = kernel_sendmsg(local->socket, &msg, iov, 2, len); if (ret < 0) - _debug("sendmsg failed: %d", ret); + trace_rxrpc_tx_fail(local->debug_id, 0, ret, + rxrpc_tx_fail_version_reply); _leave(""); } diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c index 6b9d27f0d7ec..f03de1c59ba3 100644 --- a/net/rxrpc/output.c +++ b/net/rxrpc/output.c @@ -210,6 +210,9 @@ int rxrpc_send_ack_packet(struct rxrpc_call *call, bool ping, if (ping) call->ping_time = now; conn->params.peer->last_tx_at = ktime_get_real(); + if (ret < 0) + trace_rxrpc_tx_fail(call->debug_id, serial, ret, + rxrpc_tx_fail_call_ack); if (call->state < RXRPC_CALL_COMPLETE) { if (ret < 0) { @@ -294,6 +297,10 @@ int rxrpc_send_abort_packet(struct rxrpc_call *call) ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 1, sizeof(pkt)); conn->params.peer->last_tx_at = ktime_get_real(); + if (ret < 0) + trace_rxrpc_tx_fail(call->debug_id, serial, ret, + rxrpc_tx_fail_call_abort); + rxrpc_put_connection(conn); return ret; @@ -387,6 +394,9 @@ int rxrpc_send_data_packet(struct rxrpc_call *call, struct sk_buff *skb, conn->params.peer->last_tx_at = ktime_get_real(); up_read(&conn->params.local->defrag_sem); + if (ret < 0) + trace_rxrpc_tx_fail(call->debug_id, serial, ret, + rxrpc_tx_fail_call_data_nofrag); if (ret == -EMSGSIZE) goto send_fragmentable; @@ -476,6 +486,10 @@ send_fragmentable: #endif } + if (ret < 0) + trace_rxrpc_tx_fail(call->debug_id, serial, ret, + rxrpc_tx_fail_call_data_frag); + up_write(&conn->params.local->defrag_sem); goto done; } @@ -493,6 +507,7 @@ void rxrpc_reject_packets(struct rxrpc_local *local) struct kvec iov[2]; size_t size; __be32 code; + int ret; _enter("%d", local->debug_id); @@ -527,7 +542,10 @@ void rxrpc_reject_packets(struct rxrpc_local *local) whdr.flags ^= RXRPC_CLIENT_INITIATED; whdr.flags &= RXRPC_CLIENT_INITIATED; - kernel_sendmsg(local->socket, &msg, iov, 2, size); + ret = kernel_sendmsg(local->socket, &msg, iov, 2, size); + if (ret < 0) + trace_rxrpc_tx_fail(local->debug_id, 0, ret, + rxrpc_tx_fail_reject); } rxrpc_free_skb(skb, rxrpc_skb_rx_freed); @@ -578,7 +596,8 @@ void rxrpc_send_keepalive(struct rxrpc_peer *peer) ret = kernel_sendmsg(peer->local->socket, &msg, iov, 2, len); if (ret < 0) - _debug("sendmsg failed: %d", ret); + trace_rxrpc_tx_fail(peer->debug_id, 0, ret, + rxrpc_tx_fail_version_keepalive); peer->last_tx_at = ktime_get_real(); _leave(""); diff --git a/net/rxrpc/rxkad.c b/net/rxrpc/rxkad.c index 588fea0dd362..6c0ae27fff84 100644 --- a/net/rxrpc/rxkad.c +++ b/net/rxrpc/rxkad.c @@ -664,7 +664,8 @@ static int rxkad_issue_challenge(struct rxrpc_connection *conn) ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len); if (ret < 0) { - _debug("sendmsg failed: %d", ret); + trace_rxrpc_tx_fail(conn->debug_id, serial, ret, + rxrpc_tx_fail_conn_challenge); return -EAGAIN; } @@ -719,7 +720,8 @@ static int rxkad_send_response(struct rxrpc_connection *conn, ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 3, len); if (ret < 0) { - _debug("sendmsg failed: %d", ret); + trace_rxrpc_tx_fail(conn->debug_id, serial, ret, + rxrpc_tx_fail_conn_response); return -EAGAIN; } -- cgit v1.2.3 From 21706ee8a47d3ede7fdae0be6d7c0a0e31a83229 Mon Sep 17 00:00:00 2001 From: Debabrata Banerjee Date: Wed, 9 May 2018 19:32:11 -0400 Subject: bonding: send learning packets for vlans on slave There was a regression at some point from the intended functionality of commit f60c3704e87d ("bonding: Fix alb mode to only use first level vlans.") Given the return value vlan_get_encap_level() we need to store the nest level of the bond device, and then compare the vlan's encap level to this. Without this, this check always fails and learning packets are never sent. In addition, this same commit caused a regression in the behavior of balance_alb, which requires learning packets be sent for all interfaces using the slave's mac in order to load balance properly. For vlan's that have not set a user mac, we can send after checking one bit. Otherwise we need send the set mac, albeit defeating rx load balancing for that vlan. Signed-off-by: Debabrata Banerjee Signed-off-by: David S. Miller --- drivers/net/bonding/bond_alb.c | 13 ++++++++----- drivers/net/bonding/bond_main.c | 2 ++ include/net/bonding.h | 1 + 3 files changed, 11 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index 3f6faa657360..5eb0df2e5464 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -943,6 +943,10 @@ static void alb_send_lp_vid(struct slave *slave, u8 mac_addr[], skb->priority = TC_PRIO_CONTROL; skb->dev = slave->dev; + netdev_dbg(slave->bond->dev, + "Send learning packet: dev %s mac %pM vlan %d\n", + slave->dev->name, mac_addr, vid); + if (vid) __vlan_hwaccel_put_tag(skb, vlan_proto, vid); @@ -965,14 +969,13 @@ static int alb_upper_dev_walk(struct net_device *upper, void *_data) u8 *mac_addr = data->mac_addr; struct bond_vlan_tag *tags; - if (is_vlan_dev(upper) && vlan_get_encap_level(upper) == 0) { - if (strict_match && - ether_addr_equal_64bits(mac_addr, - upper->dev_addr)) { + if (is_vlan_dev(upper) && + bond->nest_level == vlan_get_encap_level(upper) - 1) { + if (upper->addr_assign_type == NET_ADDR_STOLEN) { alb_send_lp_vid(slave, mac_addr, vlan_dev_vlan_proto(upper), vlan_dev_vlan_id(upper)); - } else if (!strict_match) { + } else { alb_send_lp_vid(slave, upper->dev_addr, vlan_dev_vlan_proto(upper), vlan_dev_vlan_id(upper)); diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 718e4914e3a0..1f1e97b26f95 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1738,6 +1738,8 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev, if (bond_mode_uses_xmit_hash(bond)) bond_update_slave_arr(bond, NULL); + bond->nest_level = dev_get_nest_level(bond_dev); + netdev_info(bond_dev, "Enslaving %s as %s interface with %s link\n", slave_dev->name, bond_is_active_slave(new_slave) ? "an active" : "a backup", diff --git a/include/net/bonding.h b/include/net/bonding.h index f801fc940b29..b52235158836 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -198,6 +198,7 @@ struct bonding { struct slave __rcu *primary_slave; struct bond_up_slave __rcu *slave_arr; /* Array of usable slaves */ bool force_primary; + u32 nest_level; s32 slave_cnt; /* never change this value outside the attach/detach wrappers */ int (*recv_probe)(const struct sk_buff *, struct bonding *, struct slave *); -- cgit v1.2.3 From 27ae357fa82be5ab73b2ef8d39dcb8ca2563483a Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Fri, 11 May 2018 16:02:04 -0700 Subject: mm, oom: fix concurrent munlock and oom reaper unmap, v3 Since exit_mmap() is done without the protection of mm->mmap_sem, it is possible for the oom reaper to concurrently operate on an mm until MMF_OOM_SKIP is set. This allows munlock_vma_pages_all() to concurrently run while the oom reaper is operating on a vma. Since munlock_vma_pages_range() depends on clearing VM_LOCKED from vm_flags before actually doing the munlock to determine if any other vmas are locking the same memory, the check for VM_LOCKED in the oom reaper is racy. This is especially noticeable on architectures such as powerpc where clearing a huge pmd requires serialize_against_pte_lookup(). If the pmd is zapped by the oom reaper during follow_page_mask() after the check for pmd_none() is bypassed, this ends up deferencing a NULL ptl or a kernel oops. Fix this by manually freeing all possible memory from the mm before doing the munlock and then setting MMF_OOM_SKIP. The oom reaper can not run on the mm anymore so the munlock is safe to do in exit_mmap(). It also matches the logic that the oom reaper currently uses for determining when to set MMF_OOM_SKIP itself, so there's no new risk of excessive oom killing. This issue fixes CVE-2018-1000200. Link: http://lkml.kernel.org/r/alpine.DEB.2.21.1804241526320.238665@chino.kir.corp.google.com Fixes: 212925802454 ("mm: oom: let oom_reap_task and exit_mmap run concurrently") Signed-off-by: David Rientjes Suggested-by: Tetsuo Handa Acked-by: Michal Hocko Cc: Andrea Arcangeli Cc: [4.14+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/oom.h | 2 ++ mm/mmap.c | 44 +++++++++++++++++------------ mm/oom_kill.c | 81 ++++++++++++++++++++++++++++------------------------- 3 files changed, 71 insertions(+), 56 deletions(-) (limited to 'include') diff --git a/include/linux/oom.h b/include/linux/oom.h index 5bad038ac012..6adac113e96d 100644 --- a/include/linux/oom.h +++ b/include/linux/oom.h @@ -95,6 +95,8 @@ static inline int check_stable_address_space(struct mm_struct *mm) return 0; } +void __oom_reap_task_mm(struct mm_struct *mm); + extern unsigned long oom_badness(struct task_struct *p, struct mem_cgroup *memcg, const nodemask_t *nodemask, unsigned long totalpages); diff --git a/mm/mmap.c b/mm/mmap.c index 9d5968d1e8e3..d6836566e4e5 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -3024,6 +3024,32 @@ void exit_mmap(struct mm_struct *mm) /* mm's last user has gone, and its about to be pulled down */ mmu_notifier_release(mm); + if (unlikely(mm_is_oom_victim(mm))) { + /* + * Manually reap the mm to free as much memory as possible. + * Then, as the oom reaper does, set MMF_OOM_SKIP to disregard + * this mm from further consideration. Taking mm->mmap_sem for + * write after setting MMF_OOM_SKIP will guarantee that the oom + * reaper will not run on this mm again after mmap_sem is + * dropped. + * + * Nothing can be holding mm->mmap_sem here and the above call + * to mmu_notifier_release(mm) ensures mmu notifier callbacks in + * __oom_reap_task_mm() will not block. + * + * This needs to be done before calling munlock_vma_pages_all(), + * which clears VM_LOCKED, otherwise the oom reaper cannot + * reliably test it. + */ + mutex_lock(&oom_lock); + __oom_reap_task_mm(mm); + mutex_unlock(&oom_lock); + + set_bit(MMF_OOM_SKIP, &mm->flags); + down_write(&mm->mmap_sem); + up_write(&mm->mmap_sem); + } + if (mm->locked_vm) { vma = mm->mmap; while (vma) { @@ -3045,24 +3071,6 @@ void exit_mmap(struct mm_struct *mm) /* update_hiwater_rss(mm) here? but nobody should be looking */ /* Use -1 here to ensure all VMAs in the mm are unmapped */ unmap_vmas(&tlb, vma, 0, -1); - - if (unlikely(mm_is_oom_victim(mm))) { - /* - * Wait for oom_reap_task() to stop working on this - * mm. Because MMF_OOM_SKIP is already set before - * calling down_read(), oom_reap_task() will not run - * on this "mm" post up_write(). - * - * mm_is_oom_victim() cannot be set from under us - * either because victim->mm is already set to NULL - * under task_lock before calling mmput and oom_mm is - * set not NULL by the OOM killer only if victim->mm - * is found not NULL while holding the task_lock. - */ - set_bit(MMF_OOM_SKIP, &mm->flags); - down_write(&mm->mmap_sem); - up_write(&mm->mmap_sem); - } free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, USER_PGTABLES_CEILING); tlb_finish_mmu(&tlb, 0, -1); diff --git a/mm/oom_kill.c b/mm/oom_kill.c index ff992fa8760a..8ba6cb88cf58 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -469,7 +469,6 @@ bool process_shares_mm(struct task_struct *p, struct mm_struct *mm) return false; } - #ifdef CONFIG_MMU /* * OOM Reaper kernel thread which tries to reap the memory used by the OOM @@ -480,16 +479,54 @@ static DECLARE_WAIT_QUEUE_HEAD(oom_reaper_wait); static struct task_struct *oom_reaper_list; static DEFINE_SPINLOCK(oom_reaper_lock); -static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) +void __oom_reap_task_mm(struct mm_struct *mm) { - struct mmu_gather tlb; struct vm_area_struct *vma; + + /* + * Tell all users of get_user/copy_from_user etc... that the content + * is no longer stable. No barriers really needed because unmapping + * should imply barriers already and the reader would hit a page fault + * if it stumbled over a reaped memory. + */ + set_bit(MMF_UNSTABLE, &mm->flags); + + for (vma = mm->mmap ; vma; vma = vma->vm_next) { + if (!can_madv_dontneed_vma(vma)) + continue; + + /* + * Only anonymous pages have a good chance to be dropped + * without additional steps which we cannot afford as we + * are OOM already. + * + * We do not even care about fs backed pages because all + * which are reclaimable have already been reclaimed and + * we do not want to block exit_mmap by keeping mm ref + * count elevated without a good reason. + */ + if (vma_is_anonymous(vma) || !(vma->vm_flags & VM_SHARED)) { + const unsigned long start = vma->vm_start; + const unsigned long end = vma->vm_end; + struct mmu_gather tlb; + + tlb_gather_mmu(&tlb, mm, start, end); + mmu_notifier_invalidate_range_start(mm, start, end); + unmap_page_range(&tlb, vma, start, end, NULL); + mmu_notifier_invalidate_range_end(mm, start, end); + tlb_finish_mmu(&tlb, start, end); + } + } +} + +static bool oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) +{ bool ret = true; /* * We have to make sure to not race with the victim exit path * and cause premature new oom victim selection: - * __oom_reap_task_mm exit_mm + * oom_reap_task_mm exit_mm * mmget_not_zero * mmput * atomic_dec_and_test @@ -534,39 +571,8 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) trace_start_task_reaping(tsk->pid); - /* - * Tell all users of get_user/copy_from_user etc... that the content - * is no longer stable. No barriers really needed because unmapping - * should imply barriers already and the reader would hit a page fault - * if it stumbled over a reaped memory. - */ - set_bit(MMF_UNSTABLE, &mm->flags); - - for (vma = mm->mmap ; vma; vma = vma->vm_next) { - if (!can_madv_dontneed_vma(vma)) - continue; + __oom_reap_task_mm(mm); - /* - * Only anonymous pages have a good chance to be dropped - * without additional steps which we cannot afford as we - * are OOM already. - * - * We do not even care about fs backed pages because all - * which are reclaimable have already been reclaimed and - * we do not want to block exit_mmap by keeping mm ref - * count elevated without a good reason. - */ - if (vma_is_anonymous(vma) || !(vma->vm_flags & VM_SHARED)) { - const unsigned long start = vma->vm_start; - const unsigned long end = vma->vm_end; - - tlb_gather_mmu(&tlb, mm, start, end); - mmu_notifier_invalidate_range_start(mm, start, end); - unmap_page_range(&tlb, vma, start, end, NULL); - mmu_notifier_invalidate_range_end(mm, start, end); - tlb_finish_mmu(&tlb, start, end); - } - } pr_info("oom_reaper: reaped process %d (%s), now anon-rss:%lukB, file-rss:%lukB, shmem-rss:%lukB\n", task_pid_nr(tsk), tsk->comm, K(get_mm_counter(mm, MM_ANONPAGES)), @@ -587,14 +593,13 @@ static void oom_reap_task(struct task_struct *tsk) struct mm_struct *mm = tsk->signal->oom_mm; /* Retry the down_read_trylock(mmap_sem) a few times */ - while (attempts++ < MAX_OOM_REAP_RETRIES && !__oom_reap_task_mm(tsk, mm)) + while (attempts++ < MAX_OOM_REAP_RETRIES && !oom_reap_task_mm(tsk, mm)) schedule_timeout_idle(HZ/10); if (attempts <= MAX_OOM_REAP_RETRIES || test_bit(MMF_OOM_SKIP, &mm->flags)) goto done; - pr_info("oom_reaper: unable to reap pid:%d (%s)\n", task_pid_nr(tsk), tsk->comm); debug_show_all_locks(); -- cgit v1.2.3 From 2075b16e32c26e4031b9fd3cbe26c54676a8fcb5 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Fri, 11 May 2018 16:02:14 -0700 Subject: rbtree: include rcu.h Since commit c1adf20052d8 ("Introduce rb_replace_node_rcu()") rbtree_augmented.h uses RCU related data structures but does not include the header file. It works as long as it gets somehow included before that and fails otherwise. Link: http://lkml.kernel.org/r/20180504103159.19938-1-bigeasy@linutronix.de Signed-off-by: Sebastian Andrzej Siewior Reviewed-by: Andrew Morton Cc: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/rbtree_augmented.h | 1 + include/linux/rbtree_latch.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/rbtree_augmented.h b/include/linux/rbtree_augmented.h index 6bfd2b581f75..af8a61be2d8d 100644 --- a/include/linux/rbtree_augmented.h +++ b/include/linux/rbtree_augmented.h @@ -26,6 +26,7 @@ #include #include +#include /* * Please note - only struct rb_augment_callbacks and the prototypes for diff --git a/include/linux/rbtree_latch.h b/include/linux/rbtree_latch.h index ece43e882b56..7d012faa509a 100644 --- a/include/linux/rbtree_latch.h +++ b/include/linux/rbtree_latch.h @@ -35,6 +35,7 @@ #include #include +#include struct latch_tree_node { struct rb_node node[2]; -- cgit v1.2.3 From 8ee3a52e3f35e064a3bf82f21dc74ddaf9843648 Mon Sep 17 00:00:00 2001 From: Emily Deng Date: Mon, 16 Apr 2018 10:07:02 +0800 Subject: drm/gpu-sched: fix force APP kill hang(v4) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit issue: there are VMC page fault occurred if force APP kill during 3dmark test, the cause is in entity_fini we manually signal all those jobs in entity's queue which confuse the sync/dep mechanism: 1)page fault occurred in sdma's clear job which operate on shadow buffer, and shadow buffer's Gart table is cleaned by ttm_bo_release since the fence in its reservation was fake signaled by entity_fini() under the case of SIGKILL received. 2)page fault occurred in gfx' job because during the lifetime of gfx job we manually fake signal all jobs from its entity in entity_fini(), thus the unmapping/clear PTE job depend on those result fence is satisfied and sdma start clearing the PTE and lead to GFX page fault. fix: 1)should at least wait all jobs already scheduled complete in entity_fini() if SIGKILL is the case. 2)if a fence signaled and try to clear some entity's dependency, should set this entity guilty to prevent its job really run since the dependency is fake signaled. v2: splitting drm_sched_entity_fini() into two functions: 1)The first one is does the waiting, removes the entity from the runqueue and returns an error when the process was killed. 2)The second one then goes over the entity, install it as completion signal for the remaining jobs and signals all jobs with an error code. v3: 1)Replace the fini1 and fini2 with better name 2)Call the first part before the VM teardown in amdgpu_driver_postclose_kms() and the second part after the VM teardown 3)Keep the original function drm_sched_entity_fini to refine the code. v4: 1)Rename entity->finished to entity->last_scheduled; 2)Rename drm_sched_entity_fini_job_cb() to drm_sched_entity_kill_jobs_cb(); 3)Pass NULL to drm_sched_entity_fini_job_cb() if -ENOENT; 4)Replace the type of entity->fini_status with "int"; 5)Remove the check about entity->finished. Signed-off-by: Monk Liu Signed-off-by: Emily Deng Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 + drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 64 ++++++++++++++++++++++++---- drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 5 ++- drivers/gpu/drm/scheduler/gpu_scheduler.c | 71 ++++++++++++++++++++++++++----- include/drm/gpu_scheduler.h | 7 +++ 5 files changed, 128 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index c25ee750c362..ea1b28536bfc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -681,6 +681,8 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data, int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id); void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr); +void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr); +void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr); void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index 09d35051fdd6..eb80edfb1b0a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -111,8 +111,9 @@ failed: return r; } -static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx) +static void amdgpu_ctx_fini(struct kref *ref) { + struct amdgpu_ctx *ctx = container_of(ref, struct amdgpu_ctx, refcount); struct amdgpu_device *adev = ctx->adev; unsigned i, j; @@ -125,13 +126,11 @@ static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx) kfree(ctx->fences); ctx->fences = NULL; - for (i = 0; i < adev->num_rings; i++) - drm_sched_entity_fini(&adev->rings[i]->sched, - &ctx->rings[i].entity); - amdgpu_queue_mgr_fini(adev, &ctx->queue_mgr); mutex_destroy(&ctx->lock); + + kfree(ctx); } static int amdgpu_ctx_alloc(struct amdgpu_device *adev, @@ -170,12 +169,15 @@ static int amdgpu_ctx_alloc(struct amdgpu_device *adev, static void amdgpu_ctx_do_release(struct kref *ref) { struct amdgpu_ctx *ctx; + u32 i; ctx = container_of(ref, struct amdgpu_ctx, refcount); - amdgpu_ctx_fini(ctx); + for (i = 0; i < ctx->adev->num_rings; i++) + drm_sched_entity_fini(&ctx->adev->rings[i]->sched, + &ctx->rings[i].entity); - kfree(ctx); + amdgpu_ctx_fini(ref); } static int amdgpu_ctx_free(struct amdgpu_fpriv *fpriv, uint32_t id) @@ -435,16 +437,62 @@ void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr) idr_init(&mgr->ctx_handles); } +void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) +{ + struct amdgpu_ctx *ctx; + struct idr *idp; + uint32_t id, i; + + idp = &mgr->ctx_handles; + + idr_for_each_entry(idp, ctx, id) { + + if (!ctx->adev) + return; + + for (i = 0; i < ctx->adev->num_rings; i++) + if (kref_read(&ctx->refcount) == 1) + drm_sched_entity_do_release(&ctx->adev->rings[i]->sched, + &ctx->rings[i].entity); + else + DRM_ERROR("ctx %p is still alive\n", ctx); + } +} + +void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr) +{ + struct amdgpu_ctx *ctx; + struct idr *idp; + uint32_t id, i; + + idp = &mgr->ctx_handles; + + idr_for_each_entry(idp, ctx, id) { + + if (!ctx->adev) + return; + + for (i = 0; i < ctx->adev->num_rings; i++) + if (kref_read(&ctx->refcount) == 1) + drm_sched_entity_cleanup(&ctx->adev->rings[i]->sched, + &ctx->rings[i].entity); + else + DRM_ERROR("ctx %p is still alive\n", ctx); + } +} + void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr) { struct amdgpu_ctx *ctx; struct idr *idp; uint32_t id; + amdgpu_ctx_mgr_entity_cleanup(mgr); + idp = &mgr->ctx_handles; idr_for_each_entry(idp, ctx, id) { - if (kref_put(&ctx->refcount, amdgpu_ctx_do_release) != 1) + if (kref_put(&ctx->refcount, amdgpu_ctx_fini) != 1) DRM_ERROR("ctx %p is still alive\n", ctx); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index bd9e723dbb2b..1ed379524117 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -913,8 +913,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, return; pm_runtime_get_sync(dev->dev); - - amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr); + amdgpu_ctx_mgr_entity_fini(&fpriv->ctx_mgr); if (adev->asic_type != CHIP_RAVEN) { amdgpu_uvd_free_handles(adev, file_priv); @@ -935,6 +934,8 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, pd = amdgpu_bo_ref(fpriv->vm.root.base.bo); amdgpu_vm_fini(adev, &fpriv->vm); + amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr); + if (pasid) amdgpu_pasid_free_delayed(pd->tbo.resv, pasid); amdgpu_bo_unref(&pd); diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c index 310275eaf128..44d21981bf3b 100644 --- a/drivers/gpu/drm/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c @@ -136,6 +136,8 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched, entity->rq = rq; entity->sched = sched; entity->guilty = guilty; + entity->fini_status = 0; + entity->last_scheduled = NULL; spin_lock_init(&entity->rq_lock); spin_lock_init(&entity->queue_lock); @@ -197,19 +199,30 @@ static bool drm_sched_entity_is_ready(struct drm_sched_entity *entity) return true; } +static void drm_sched_entity_kill_jobs_cb(struct dma_fence *f, + struct dma_fence_cb *cb) +{ + struct drm_sched_job *job = container_of(cb, struct drm_sched_job, + finish_cb); + 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); +} + + /** * Destroy a context entity * * @sched Pointer to scheduler instance * @entity The pointer to a valid scheduler entity * - * Cleanup and free the allocated resources. + * Splitting drm_sched_entity_fini() into two functions, The first one is does the waiting, + * removes the entity from the runqueue and returns an error when the process was killed. */ -void drm_sched_entity_fini(struct drm_gpu_scheduler *sched, +void drm_sched_entity_do_release(struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity) { - int r; - if (!drm_sched_entity_is_initialized(sched, entity)) return; /** @@ -217,13 +230,28 @@ void drm_sched_entity_fini(struct drm_gpu_scheduler *sched, * queued IBs or discard them on SIGKILL */ if ((current->flags & PF_SIGNALED) && current->exit_code == SIGKILL) - r = -ERESTARTSYS; + entity->fini_status = -ERESTARTSYS; else - r = wait_event_killable(sched->job_scheduled, + entity->fini_status = wait_event_killable(sched->job_scheduled, drm_sched_entity_is_idle(entity)); drm_sched_entity_set_rq(entity, NULL); - if (r) { +} +EXPORT_SYMBOL(drm_sched_entity_do_release); + +/** + * Destroy a context entity + * + * @sched Pointer to scheduler instance + * @entity The pointer to a valid scheduler entity + * + * The second one then goes over the entity and signals all jobs with an error code. + */ +void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched, + struct drm_sched_entity *entity) +{ + if (entity->fini_status) { struct drm_sched_job *job; + int r; /* Park the kernel for a moment to make sure it isn't processing * our enity. @@ -241,13 +269,26 @@ void drm_sched_entity_fini(struct drm_gpu_scheduler *sched, struct drm_sched_fence *s_fence = job->s_fence; drm_sched_fence_scheduled(s_fence); dma_fence_set_error(&s_fence->finished, -ESRCH); - drm_sched_fence_finished(s_fence); - WARN_ON(s_fence->parent); - dma_fence_put(&s_fence->finished); - sched->ops->free_job(job); + r = dma_fence_add_callback(entity->last_scheduled, &job->finish_cb, + drm_sched_entity_kill_jobs_cb); + if (r == -ENOENT) + drm_sched_entity_kill_jobs_cb(NULL, &job->finish_cb); + else if (r) + DRM_ERROR("fence add callback failed (%d)\n", r); } + + dma_fence_put(entity->last_scheduled); + entity->last_scheduled = NULL; } } +EXPORT_SYMBOL(drm_sched_entity_cleanup); + +void drm_sched_entity_fini(struct drm_gpu_scheduler *sched, + struct drm_sched_entity *entity) +{ + drm_sched_entity_do_release(sched, entity); + drm_sched_entity_cleanup(sched, entity); +} EXPORT_SYMBOL(drm_sched_entity_fini); static void drm_sched_entity_wakeup(struct dma_fence *f, struct dma_fence_cb *cb) @@ -530,6 +571,10 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched) spin_unlock(&sched->job_list_lock); fence = sched->ops->run_job(s_job); atomic_inc(&sched->hw_rq_count); + + dma_fence_put(s_job->entity->last_scheduled); + s_job->entity->last_scheduled = dma_fence_get(&s_fence->finished); + if (fence) { s_fence->parent = dma_fence_get(fence); r = dma_fence_add_callback(fence, &s_fence->cb, @@ -556,6 +601,7 @@ int drm_sched_job_init(struct drm_sched_job *job, void *owner) { job->sched = sched; + job->entity = entity; job->s_priority = entity->rq - sched->sched_rq; job->s_fence = drm_sched_fence_create(entity, owner); if (!job->s_fence) @@ -669,6 +715,9 @@ static int drm_sched_main(void *param) fence = sched->ops->run_job(sched_job); drm_sched_fence_scheduled(s_fence); + dma_fence_put(entity->last_scheduled); + entity->last_scheduled = dma_fence_get(&s_fence->finished); + if (fence) { s_fence->parent = dma_fence_get(fence); r = dma_fence_add_callback(fence, &s_fence->cb, diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index c053a32341bf..350a62c26b29 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -65,6 +65,8 @@ struct drm_sched_entity { struct dma_fence *dependency; struct dma_fence_cb cb; atomic_t *guilty; /* points to ctx's guilty */ + int fini_status; + struct dma_fence *last_scheduled; }; /** @@ -119,6 +121,7 @@ struct drm_sched_job { uint64_t id; atomic_t karma; enum drm_sched_priority s_priority; + struct drm_sched_entity *entity; }; static inline bool drm_sched_invalidate_job(struct drm_sched_job *s_job, @@ -186,6 +189,10 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity, struct drm_sched_rq *rq, uint32_t jobs, atomic_t *guilty); +void drm_sched_entity_do_release(struct drm_gpu_scheduler *sched, + struct drm_sched_entity *entity); +void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched, + struct drm_sched_entity *entity); void drm_sched_entity_fini(struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity); void drm_sched_entity_push_job(struct drm_sched_job *sched_job, -- cgit v1.2.3 From 1afd30efeddbb1b32cf35d3bf6477b35690eeca6 Mon Sep 17 00:00:00 2001 From: Christian König Date: Tue, 10 Apr 2018 13:42:29 +0200 Subject: drm/amdgpu: revert "add new bo flag that indicates BOs don't need fallback (v2)" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit 6f51d28bfe8e1a676de5cd877639245bed3cc818. Makes fallback handling to complicated. This is just a feature for the GEM interface and shouldn't leak into the core BO create function. Signed-off-by: Christian König Acked-by: Chunming Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 3 +-- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 +---- include/uapi/drm/amdgpu_drm.h | 2 -- 3 files changed, 2 insertions(+), 8 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 68af2f878bc9..e1756b68a17b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -385,8 +385,7 @@ retry: amdgpu_bo_in_cpu_visible_vram(bo)) p->bytes_moved_vis += ctx.bytes_moved; - if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains && - !(bo->flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) { + if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { domain = bo->allowed_domains; goto retry; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index a160ef0332d6..1de6864da717 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -388,8 +388,6 @@ retry: drm_gem_private_object_init(adev->ddev, &bo->gem_base, size); INIT_LIST_HEAD(&bo->shadow_list); INIT_LIST_HEAD(&bo->va); - bo->preferred_domains = preferred_domains; - bo->allowed_domains = allowed_domains; bo->flags = flags; @@ -426,8 +424,7 @@ retry: r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type, &bo->placement, page_align, &ctx, acc_size, NULL, resv, &amdgpu_ttm_bo_destroy); - if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device && - !(flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) { + if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device) { if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) { flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; goto retry; diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index 4f5a27d64c54..c363b67f2d0a 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -95,8 +95,6 @@ extern "C" { #define AMDGPU_GEM_CREATE_VM_ALWAYS_VALID (1 << 6) /* Flag that BO sharing will be explicitly synchronized */ #define AMDGPU_GEM_CREATE_EXPLICIT_SYNC (1 << 7) -/* Flag that BO doesn't need fallback */ -#define AMDGPU_GEM_CREATE_NO_FALLBACK (1 << 8) struct drm_amdgpu_gem_create_in { /** the requested memory size */ -- cgit v1.2.3 From 3f188453faf7ba5b59e8064df4afffbc946e25ec Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Tue, 17 Apr 2018 18:34:40 +0800 Subject: drm/amdgpu: handle domain mask checking v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit if domain is illegal, we should return error. v2: remove duplicated domain checking. Signed-off-by: Chunming Zhou Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 7 +------ drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 9 +-------- include/uapi/drm/amdgpu_drm.h | 6 ++++++ 3 files changed, 8 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index ff606ce88837..c62c3dd4dcc6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -229,12 +229,7 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data, return -EINVAL; /* reject invalid gem domains */ - if (args->in.domains & ~(AMDGPU_GEM_DOMAIN_CPU | - AMDGPU_GEM_DOMAIN_GTT | - AMDGPU_GEM_DOMAIN_VRAM | - AMDGPU_GEM_DOMAIN_GDS | - AMDGPU_GEM_DOMAIN_GWS | - AMDGPU_GEM_DOMAIN_OA)) + if (args->in.domains & ~AMDGPU_GEM_DOMAIN_MASK) return -EINVAL; /* create a gem object to contain this object in */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 9258f0694922..feece0a491a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -360,7 +360,6 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, }; struct amdgpu_bo *bo; unsigned long page_align, size = bp->size; - u32 preferred_domains; size_t acc_size; int r; @@ -381,14 +380,8 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, drm_gem_private_object_init(adev->ddev, &bo->gem_base, size); INIT_LIST_HEAD(&bo->shadow_list); INIT_LIST_HEAD(&bo->va); - preferred_domains = bp->preferred_domain ? bp->preferred_domain : + bo->preferred_domains = bp->preferred_domain ? bp->preferred_domain : bp->domain; - bo->preferred_domains = preferred_domains & (AMDGPU_GEM_DOMAIN_VRAM | - AMDGPU_GEM_DOMAIN_GTT | - AMDGPU_GEM_DOMAIN_CPU | - AMDGPU_GEM_DOMAIN_GDS | - AMDGPU_GEM_DOMAIN_GWS | - AMDGPU_GEM_DOMAIN_OA); bo->allowed_domains = bo->preferred_domains; if (bp->type != ttm_bo_type_kernel && bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM) diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index c363b67f2d0a..b193e95f1f24 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -78,6 +78,12 @@ extern "C" { #define AMDGPU_GEM_DOMAIN_GDS 0x8 #define AMDGPU_GEM_DOMAIN_GWS 0x10 #define AMDGPU_GEM_DOMAIN_OA 0x20 +#define AMDGPU_GEM_DOMAIN_MASK (AMDGPU_GEM_DOMAIN_CPU | \ + AMDGPU_GEM_DOMAIN_GTT | \ + AMDGPU_GEM_DOMAIN_VRAM | \ + AMDGPU_GEM_DOMAIN_GDS | \ + AMDGPU_GEM_DOMAIN_GWS | \ + AMDGPU_GEM_DOMAIN_OA) /* Flag that CPU access will be required for the case of VRAM domain */ #define AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED (1 << 0) -- cgit v1.2.3 From d240cd9eddd943dbe0267d081697195ff1e90b65 Mon Sep 17 00:00:00 2001 From: Marek Olšák Date: Tue, 3 Apr 2018 13:05:03 -0400 Subject: drm/amdgpu: optionally do a writeback but don't invalidate TC for IB fences MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is a new IB flag that enables this new behavior. Full invalidation is unnecessary for RELEASE_MEM and doesn't make sense when draw calls from two adjacent gfx IBs run in parallel. This will be the new default for Mesa. v2: bump the version Signed-off-by: Marek Olšák Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 5 +++-- drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 8 ++++++-- drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h | 4 +++- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 11 +++++++---- drivers/gpu/drm/amd/amdgpu/soc15d.h | 1 + include/uapi/drm/amdgpu_drm.h | 4 ++++ 8 files changed, 27 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 5c0567ad1ba7..7c17a0bc2cd2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -75,9 +75,10 @@ * - 3.23.0 - Add query for VRAM lost counter * - 3.24.0 - Add high priority compute support for gfx9 * - 3.25.0 - Add support for sensor query info (stable pstate sclk/mclk). + * - 3.26.0 - GFX9: Process AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE. */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 25 +#define KMS_DRIVER_MINOR 26 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 97449e06a242..d09fcab2398f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -131,7 +131,8 @@ static u32 amdgpu_fence_read(struct amdgpu_ring *ring) * Emits a fence command on the requested ring (all asics). * Returns 0 on success, -ENOMEM on failure. */ -int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f) +int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f, + unsigned flags) { struct amdgpu_device *adev = ring->adev; struct amdgpu_fence *fence; @@ -149,7 +150,7 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f) adev->fence_context + ring->idx, seq); amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr, - seq, AMDGPU_FENCE_FLAG_INT); + seq, flags | AMDGPU_FENCE_FLAG_INT); ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask]; /* This function can't be called concurrently anyway, otherwise diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 311589e02d17..f70eeed9ed76 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -127,6 +127,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, struct amdgpu_vm *vm; uint64_t fence_ctx; uint32_t status = 0, alloc_size; + unsigned fence_flags = 0; unsigned i; int r = 0; @@ -227,7 +228,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, #endif amdgpu_asic_invalidate_hdp(adev, ring); - r = amdgpu_fence_emit(ring, f); + if (ib->flags & AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE) + fence_flags |= AMDGPU_FENCE_FLAG_TC_WB_ONLY; + + r = amdgpu_fence_emit(ring, f, fence_flags); if (r) { dev_err(adev->dev, "failed to emit fence (%d)\n", r); if (job && job->vmid) @@ -242,7 +246,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, /* wrap the last IB with fence */ if (job && job->uf_addr) { amdgpu_ring_emit_fence(ring, job->uf_addr, job->uf_sequence, - AMDGPU_FENCE_FLAG_64BIT); + fence_flags | AMDGPU_FENCE_FLAG_64BIT); } if (patch_offset != ~0 && ring->funcs->patch_cond_exec) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index 08fcdf6f7b53..4f8dac2d36a5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -42,6 +42,7 @@ #define AMDGPU_FENCE_FLAG_64BIT (1 << 0) #define AMDGPU_FENCE_FLAG_INT (1 << 1) +#define AMDGPU_FENCE_FLAG_TC_WB_ONLY (1 << 2) enum amdgpu_ring_type { AMDGPU_RING_TYPE_GFX, @@ -90,7 +91,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, unsigned irq_type); void amdgpu_fence_driver_suspend(struct amdgpu_device *adev); void amdgpu_fence_driver_resume(struct amdgpu_device *adev); -int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence); +int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence, + unsigned flags); int amdgpu_fence_emit_polling(struct amdgpu_ring *ring, uint32_t *s); void amdgpu_fence_process(struct amdgpu_ring *ring); int amdgpu_fence_wait_empty(struct amdgpu_ring *ring); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 9ec7c1041df2..9c2195a2896d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -633,7 +633,7 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_ amdgpu_gmc_emit_pasid_mapping(ring, job->vmid, job->pasid); if (vm_flush_needed || pasid_mapping_needed) { - r = amdgpu_fence_emit(ring, &fence); + r = amdgpu_fence_emit(ring, &fence, 0); if (r) return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 6a19e0311a9c..05b2d34110b7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -3775,13 +3775,16 @@ static void gfx_v9_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, { bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT; bool int_sel = flags & AMDGPU_FENCE_FLAG_INT; + bool writeback = flags & AMDGPU_FENCE_FLAG_TC_WB_ONLY; /* RELEASE_MEM - flush caches, send int */ amdgpu_ring_write(ring, PACKET3(PACKET3_RELEASE_MEM, 6)); - amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN | - EOP_TC_ACTION_EN | - EOP_TC_WB_ACTION_EN | - EOP_TC_MD_ACTION_EN | + amdgpu_ring_write(ring, ((writeback ? (EOP_TC_WB_ACTION_EN | + EOP_TC_NC_ACTION_EN) : + (EOP_TCL1_ACTION_EN | + EOP_TC_ACTION_EN | + EOP_TC_WB_ACTION_EN | + EOP_TC_MD_ACTION_EN)) | EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | EVENT_INDEX(5))); amdgpu_ring_write(ring, DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0)); diff --git a/drivers/gpu/drm/amd/amdgpu/soc15d.h b/drivers/gpu/drm/amd/amdgpu/soc15d.h index 7f408f85fdb6..839a144c1645 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15d.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15d.h @@ -159,6 +159,7 @@ #define EOP_TC_WB_ACTION_EN (1 << 15) /* L2 */ #define EOP_TCL1_ACTION_EN (1 << 16) #define EOP_TC_ACTION_EN (1 << 17) /* L2 */ +#define EOP_TC_NC_ACTION_EN (1 << 19) #define EOP_TC_MD_ACTION_EN (1 << 21) /* L2 metadata */ #define DATA_SEL(x) ((x) << 29) diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index b193e95f1f24..78fe828f2f79 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -526,6 +526,10 @@ union drm_amdgpu_cs { /* Preempt flag, IB should set Pre_enb bit if PREEMPT flag detected */ #define AMDGPU_IB_FLAG_PREEMPT (1<<2) +/* The IB fence should do the L2 writeback but not invalidate any shader + * caches (L2/vL1/sL1/I$). */ +#define AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE (1 << 3) + struct drm_amdgpu_cs_chunk_ib { __u32 _pad; /** AMDGPU_IB_FLAG_* */ -- cgit v1.2.3 From 621a6318adea69b08a3652c64bc7cc0cb4dacfb4 Mon Sep 17 00:00:00 2001 From: Huang Rui Date: Mon, 22 Jan 2018 20:48:14 +0800 Subject: drm/amdgpu: add save restore list cntl gpm and srm firmware support MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit RLC save/restore list cntl/gpm_mem/srm_mem ucodes are used for CGPG and gfxoff function. Signed-off-by: Huang Rui Acked-by: Alex Deucher Acked-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 15 +++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 36 ++++++++++++++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 17 +++++++++- drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h | 3 ++ drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 55 +++++++++++++++++++++++++++++-- drivers/gpu/drm/amd/amdgpu/psp_v10_0.c | 9 +++++ include/uapi/drm/amdgpu_drm.h | 6 ++++ 7 files changed, 138 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index d64ef30fed47..5ad893915a85 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -774,9 +774,18 @@ struct amdgpu_rlc { u32 starting_offsets_start; u32 reg_list_format_size_bytes; u32 reg_list_size_bytes; + u32 reg_list_format_direct_reg_list_length; + u32 save_restore_list_cntl_size_bytes; + u32 save_restore_list_gpm_size_bytes; + u32 save_restore_list_srm_size_bytes; u32 *register_list_format; u32 *register_restore; + u8 *save_restore_list_cntl; + u8 *save_restore_list_gpm; + u8 *save_restore_list_srm; + + bool is_rlc_v2_1; }; #define AMDGPU_MAX_COMPUTE_QUEUES KGD_MAX_QUEUES @@ -943,6 +952,12 @@ struct amdgpu_gfx { uint32_t ce_feature_version; uint32_t pfp_feature_version; uint32_t rlc_feature_version; + uint32_t rlc_srlc_fw_version; + uint32_t rlc_srlc_feature_version; + uint32_t rlc_srlg_fw_version; + uint32_t rlc_srlg_feature_version; + uint32_t rlc_srls_fw_version; + uint32_t rlc_srls_feature_version; uint32_t mec_feature_version; uint32_t mec2_feature_version; struct amdgpu_ring gfx_ring[AMDGPU_MAX_GFX_RINGS]; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index d602f8b14c58..eb4785e51573 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -215,6 +215,18 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info, fw_info->ver = adev->gfx.rlc_fw_version; fw_info->feature = adev->gfx.rlc_feature_version; break; + case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL: + fw_info->ver = adev->gfx.rlc_srlc_fw_version; + fw_info->feature = adev->gfx.rlc_srlc_feature_version; + break; + case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM: + fw_info->ver = adev->gfx.rlc_srlg_fw_version; + fw_info->feature = adev->gfx.rlc_srlg_feature_version; + break; + case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM: + fw_info->ver = adev->gfx.rlc_srls_fw_version; + fw_info->feature = adev->gfx.rlc_srls_feature_version; + break; case AMDGPU_INFO_FW_GFX_MEC: if (query_fw->index == 0) { fw_info->ver = adev->gfx.mec_fw_version; @@ -1149,6 +1161,30 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data) seq_printf(m, "RLC feature version: %u, firmware version: 0x%08x\n", fw_info.feature, fw_info.ver); + /* RLC SAVE RESTORE LIST CNTL */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "RLC SRLC feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* RLC SAVE RESTORE LIST GPM MEM */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "RLC SRLG feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* RLC SAVE RESTORE LIST SRM MEM */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "RLC SRLS feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + /* MEC */ query_fw.fw_type = AMDGPU_INFO_FW_GFX_MEC; query_fw.index = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index 84d652599d5b..0c74c09ef3b0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -337,7 +337,10 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev, (ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1 && ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2 && ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1_JT && - ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT)) { + ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT && + ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL && + ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM && + ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM)) { ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes); memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data + @@ -359,6 +362,18 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev, le32_to_cpu(header->ucode_array_offset_bytes) + le32_to_cpu(cp_hdr->jt_offset) * 4), ucode->ucode_size); + } else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL) { + ucode->ucode_size = adev->gfx.rlc.save_restore_list_cntl_size_bytes; + memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_cntl, + ucode->ucode_size); + } else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM) { + ucode->ucode_size = adev->gfx.rlc.save_restore_list_gpm_size_bytes; + memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_gpm, + ucode->ucode_size); + } else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM) { + ucode->ucode_size = adev->gfx.rlc.save_restore_list_srm_size_bytes; + memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_srm, + ucode->ucode_size); } return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index 0b262f4bb4fc..08e38579af24 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -187,6 +187,9 @@ enum AMDGPU_UCODE_ID { AMDGPU_UCODE_ID_CP_MEC2, AMDGPU_UCODE_ID_CP_MEC2_JT, AMDGPU_UCODE_ID_RLC_G, + AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL, + AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM, + AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM, AMDGPU_UCODE_ID_STORAGE, AMDGPU_UCODE_ID_SMC, AMDGPU_UCODE_ID_UVD, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 587a8731fa31..73b76fa29bad 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -41,7 +41,6 @@ #define GFX9_MEC_HPD_SIZE 2048 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L #define RLC_SAVE_RESTORE_ADDR_STARTING_OFFSET 0x00000000L -#define GFX9_RLC_FORMAT_DIRECT_REG_LIST_LENGTH 34 #define mmPWR_MISC_CNTL_STATUS 0x0183 #define mmPWR_MISC_CNTL_STATUS_BASE_IDX 0 @@ -401,6 +400,27 @@ static void gfx_v9_0_free_microcode(struct amdgpu_device *adev) kfree(adev->gfx.rlc.register_list_format); } +static void gfx_v9_0_init_rlc_ext_microcode(struct amdgpu_device *adev) +{ + const struct rlc_firmware_header_v2_1 *rlc_hdr; + + rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data; + adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver); + adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver); + adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes); + adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes); + adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver); + adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver); + adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes); + adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes); + adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver); + adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver); + adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes); + adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes); + adev->gfx.rlc.reg_list_format_direct_reg_list_length = + le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length); +} + static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) { const char *chip_name; @@ -412,6 +432,8 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) const struct rlc_firmware_header_v2_0 *rlc_hdr; unsigned int *tmp = NULL; unsigned int i = 0; + uint16_t version_major; + uint16_t version_minor; DRM_DEBUG("\n"); @@ -468,6 +490,12 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) goto out; err = amdgpu_ucode_validate(adev->gfx.rlc_fw); rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data; + + version_major = le16_to_cpu(rlc_hdr->header.header_version_major); + version_minor = le16_to_cpu(rlc_hdr->header.header_version_minor); + if (version_major == 2 && version_minor == 1) + adev->gfx.rlc.is_rlc_v2_1 = true; + adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version); adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version); adev->gfx.rlc.save_and_restore_offset = @@ -508,6 +536,9 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) for (i = 0 ; i < (rlc_hdr->reg_list_size_bytes >> 2); i++) adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]); + if (adev->gfx.rlc.is_rlc_v2_1) + gfx_v9_0_init_rlc_ext_microcode(adev); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", chip_name); err = request_firmware(&adev->gfx.mec_fw, fw_name, adev->dev); if (err) @@ -566,6 +597,26 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) adev->firmware.fw_size += ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); + if (adev->gfx.rlc.is_rlc_v2_1) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.save_restore_list_cntl_size_bytes, PAGE_SIZE); + + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE); + + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE); + } + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1]; info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1; info->fw = adev->gfx.mec_fw; @@ -1781,7 +1832,7 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev) /* setup unique_indirect_regs array and indirect_start_offsets array */ gfx_v9_0_parse_ind_reg_list(register_list_format, - GFX9_RLC_FORMAT_DIRECT_REG_LIST_LENGTH, + adev->gfx.rlc.reg_list_format_direct_reg_list_length, adev->gfx.rlc.reg_list_format_size_bytes >> 2, unique_indirect_regs, &unique_indirect_reg_count, diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c index 8873d833a7f7..0ff136d02d9b 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c @@ -70,6 +70,15 @@ psp_v10_0_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type * case AMDGPU_UCODE_ID_RLC_G: *type = GFX_FW_TYPE_RLC_G; break; + case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL: + *type = GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL; + break; + case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM: + *type = GFX_FW_TYPE_RLC_RESTORE_LIST_GPM_MEM; + break; + case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM: + *type = GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_MEM; + break; case AMDGPU_UCODE_ID_SMC: *type = GFX_FW_TYPE_SMU; break; diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index 78fe828f2f79..081d25640b64 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -630,6 +630,12 @@ struct drm_amdgpu_cs_chunk_data { #define AMDGPU_INFO_FW_ASD 0x0d /* Subquery id: Query VCN firmware version */ #define AMDGPU_INFO_FW_VCN 0x0e + /* Subquery id: Query GFX RLC SRLC firmware version */ + #define AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL 0x0f + /* Subquery id: Query GFX RLC SRLG firmware version */ + #define AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM 0x10 + /* Subquery id: Query GFX RLC SRLS firmware version */ + #define AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM 0x11 /* number of bytes moved for TTM migration */ #define AMDGPU_INFO_NUM_BYTES_MOVED 0x0f /* the used VRAM size */ -- cgit v1.2.3 From 48ff108d9dc42bf92256484c50cdb3697f5ccb04 Mon Sep 17 00:00:00 2001 From: Leo Liu Date: Thu, 9 Nov 2017 13:18:24 -0500 Subject: drm/amdgpu: add VEGAM ASIC type Signed-off-by: Leo Liu Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 + include/drm/amd_asic_type.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 9e917f53f357..8ce60e6e2614 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -83,6 +83,7 @@ static const char *amdgpu_asic_name[] = { "POLARIS10", "POLARIS11", "POLARIS12", + "VEGAM", "VEGA10", "VEGA12", "RAVEN", diff --git a/include/drm/amd_asic_type.h b/include/drm/amd_asic_type.h index 6c731c52c071..695bde7eb055 100644 --- a/include/drm/amd_asic_type.h +++ b/include/drm/amd_asic_type.h @@ -44,6 +44,7 @@ enum amd_asic_type { CHIP_POLARIS10, CHIP_POLARIS11, CHIP_POLARIS12, + CHIP_VEGAM, CHIP_VEGA10, CHIP_VEGA12, CHIP_RAVEN, -- cgit v1.2.3 From 959a2091fae0fa498c79e095a4f6cbbb202a1194 Mon Sep 17 00:00:00 2001 From: Yong Zhao Date: Mon, 14 May 2018 12:15:27 -0400 Subject: drm/amdgpu: Add support to change mtype for 2nd part of gart BOs on GFX9 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This change prepares for a workaround in amdkfd for a GFX9 HW bug. It requires the control stack memory of compute queues, which is allocated from the second page of MQD gart BOs, to have mtype NC, rather than the default UC. Signed-off-by: Yong Zhao Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 54 +++++++++++++++++++++++++++------ drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 5 +-- include/uapi/drm/amdgpu_drm.h | 4 +++ 3 files changed, 51 insertions(+), 12 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 dfd22db13fb1..cc3b067e1ec6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -834,6 +834,45 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm) sg_free_table(ttm->sg); } +int amdgpu_ttm_gart_bind(struct amdgpu_device *adev, + struct ttm_buffer_object *tbo, + uint64_t flags) +{ + struct amdgpu_bo *abo = ttm_to_amdgpu_bo(tbo); + struct ttm_tt *ttm = tbo->ttm; + struct amdgpu_ttm_tt *gtt = (void *)ttm; + int r; + + if (abo->flags & AMDGPU_GEM_CREATE_MQD_GFX9) { + uint64_t page_idx = 1; + + r = amdgpu_gart_bind(adev, gtt->offset, page_idx, + ttm->pages, gtt->ttm.dma_address, flags); + if (r) + goto gart_bind_fail; + + /* Patch mtype of the second part BO */ + flags &= ~AMDGPU_PTE_MTYPE_MASK; + flags |= AMDGPU_PTE_MTYPE(AMDGPU_MTYPE_NC); + + r = amdgpu_gart_bind(adev, + gtt->offset + (page_idx << PAGE_SHIFT), + ttm->num_pages - page_idx, + &ttm->pages[page_idx], + &(gtt->ttm.dma_address[page_idx]), flags); + } else { + r = amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages, + ttm->pages, gtt->ttm.dma_address, flags); + } + +gart_bind_fail: + if (r) + DRM_ERROR("failed to bind %lu pages at 0x%08llX\n", + ttm->num_pages, gtt->offset); + + return r; +} + static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) { @@ -907,8 +946,7 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo) flags = amdgpu_ttm_tt_pte_flags(adev, bo->ttm, &tmp); gtt->offset = (u64)tmp.start << PAGE_SHIFT; - r = amdgpu_gart_bind(adev, gtt->offset, bo->ttm->num_pages, - bo->ttm->pages, gtt->ttm.dma_address, flags); + r = amdgpu_ttm_gart_bind(adev, bo, flags); if (unlikely(r)) { ttm_bo_mem_put(bo, &tmp); return r; @@ -925,19 +963,15 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo) int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo) { struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev); - struct amdgpu_ttm_tt *gtt = (void *)tbo->ttm; uint64_t flags; int r; - if (!gtt) + if (!tbo->ttm) return 0; - flags = amdgpu_ttm_tt_pte_flags(adev, >t->ttm.ttm, &tbo->mem); - r = amdgpu_gart_bind(adev, gtt->offset, gtt->ttm.ttm.num_pages, - gtt->ttm.ttm.pages, gtt->ttm.dma_address, flags); - if (r) - DRM_ERROR("failed to bind %lu pages at 0x%08llX\n", - gtt->ttm.ttm.num_pages, gtt->offset); + flags = amdgpu_ttm_tt_pte_flags(adev, tbo->ttm, &tbo->mem); + r = amdgpu_ttm_gart_bind(adev, tbo, flags); + return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 30f080364c97..4cf678684a12 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -75,11 +75,12 @@ struct amdgpu_bo_list_entry; /* PDE Block Fragment Size for VEGA10 */ #define AMDGPU_PDE_BFS(a) ((uint64_t)a << 59) -/* VEGA10 only */ + +/* For GFX9 */ #define AMDGPU_PTE_MTYPE(a) ((uint64_t)a << 57) #define AMDGPU_PTE_MTYPE_MASK AMDGPU_PTE_MTYPE(3ULL) -/* For Raven */ +#define AMDGPU_MTYPE_NC 0 #define AMDGPU_MTYPE_CC 2 #define AMDGPU_PTE_DEFAULT_ATC (AMDGPU_PTE_SYSTEM \ diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h index 081d25640b64..78b4dd89fcb4 100644 --- a/include/uapi/drm/amdgpu_drm.h +++ b/include/uapi/drm/amdgpu_drm.h @@ -101,6 +101,10 @@ extern "C" { #define AMDGPU_GEM_CREATE_VM_ALWAYS_VALID (1 << 6) /* Flag that BO sharing will be explicitly synchronized */ #define AMDGPU_GEM_CREATE_EXPLICIT_SYNC (1 << 7) +/* Flag that indicates allocating MQD gart on GFX9, where the mtype + * for the second page onward should be set to NC. + */ +#define AMDGPU_GEM_CREATE_MQD_GFX9 (1 << 8) struct drm_amdgpu_gem_create_in { /** the requested memory size */ -- cgit v1.2.3 From 8344c53f57057b42a5da87e9557c40fcda18fb7a Mon Sep 17 00:00:00 2001 From: Nayan Deshmukh Date: Thu, 29 Mar 2018 22:36:32 +0530 Subject: drm/scheduler: remove unused parameter MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit this patch also effect the amdgpu and etnaviv drivers which use the function drm_sched_entity_init Signed-off-by: Nayan Deshmukh Suggested-by: Christian König Acked-by: Lucas Stach Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 4 ++-- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 +- drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 2 +- drivers/gpu/drm/etnaviv/etnaviv_drv.c | 2 +- drivers/gpu/drm/scheduler/gpu_scheduler.c | 3 +-- include/drm/gpu_scheduler.h | 2 +- 11 files changed, 12 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index 6741a62a7d15..a8e531d604fa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -91,7 +91,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev, continue; r = drm_sched_entity_init(&ring->sched, &ctx->rings[i].entity, - rq, amdgpu_sched_jobs, &ctx->guilty); + rq, &ctx->guilty); if (r) goto failed; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index cc3b067e1ec6..5e9fd256faad 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -111,7 +111,7 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev) ring = adev->mman.buffer_funcs_ring; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL]; r = drm_sched_entity_init(&ring->sched, &adev->mman.entity, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r) { DRM_ERROR("Failed setting up TTM BO move run queue.\n"); goto error_entity; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index d8dd4028c2bb..de4d77af02ae 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -242,7 +242,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) ring = &adev->uvd.ring; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r != 0) { DRM_ERROR("Failed setting up UVD run queue.\n"); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index e2186eda3271..a86322f5164f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -186,7 +186,7 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size) ring = &adev->vce.ring[0]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; r = drm_sched_entity_init(&ring->sched, &adev->vce.entity, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r != 0) { DRM_ERROR("Failed setting up VCE run queue.\n"); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 58e495330b38..e5d234cf804f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -105,7 +105,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) ring = &adev->vcn.ring_dec; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_dec, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r != 0) { DRM_ERROR("Failed setting up VCN dec run queue.\n"); return r; @@ -114,7 +114,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) ring = &adev->vcn.ring_enc[0]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_enc, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r != 0) { DRM_ERROR("Failed setting up VCN enc run queue.\n"); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 8e71d3984016..1a8f4e0dd023 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -2404,7 +2404,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, ring = adev->vm_manager.vm_pte_rings[ring_instance]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL]; r = drm_sched_entity_init(&ring->sched, &vm->entity, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 8041b26a7a21..ca6ab56357b5 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -429,7 +429,7 @@ static int uvd_v6_0_sw_init(void *handle) ring = &adev->uvd.ring_enc[0]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r) { DRM_ERROR("Failed setting up UVD ENC run queue.\n"); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index b0de1e04093b..0ca63d588670 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -418,7 +418,7 @@ static int uvd_v7_0_sw_init(void *handle) ring = &adev->uvd.ring_enc[0]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc, - rq, amdgpu_sched_jobs, NULL); + rq, NULL); if (r) { DRM_ERROR("Failed setting up UVD ENC run queue.\n"); return r; diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c index ab50090d066c..23e73c2a19f4 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c @@ -116,7 +116,7 @@ static int etnaviv_open(struct drm_device *dev, struct drm_file *file) drm_sched_entity_init(&gpu->sched, &ctx->sched_entity[i], &gpu->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL], - 32, NULL); + NULL); } } diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c index 1f1dd70125a7..a364fc0b38c3 100644 --- a/drivers/gpu/drm/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c @@ -117,7 +117,6 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq) * @sched The pointer to the scheduler * @entity The pointer to a valid drm_sched_entity * @rq The run queue this entity belongs - * @jobs The max number of jobs in the job queue * @guilty atomic_t set to 1 when a job on this queue * is found to be guilty causing a timeout * @@ -126,7 +125,7 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq) int drm_sched_entity_init(struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity, struct drm_sched_rq *rq, - uint32_t jobs, atomic_t *guilty) + atomic_t *guilty) { if (!(sched && entity && rq)) return -EINVAL; diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index 350a62c26b29..52380067a43f 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -188,7 +188,7 @@ void drm_sched_fini(struct drm_gpu_scheduler *sched); int drm_sched_entity_init(struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity, struct drm_sched_rq *rq, - uint32_t jobs, atomic_t *guilty); + atomic_t *guilty); void drm_sched_entity_do_release(struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity); void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched, -- cgit v1.2.3 From 956fcddc0b2a7430b6ee4783827f57cb7c823c7d Mon Sep 17 00:00:00 2001 From: Feifei Xu Date: Fri, 20 Apr 2018 12:27:54 +0800 Subject: drm/amdgpu: Add vega20 to asic_type enum. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add vega20 to amd_asic_type enum and amdgpu_asic_name[]. Reviewed-by: Christian König Signed-off-by: Feifei Xu Reviewed-by: Alex Deucher Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 + include/drm/amd_asic_type.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 9fb20a53d5b2..f84fc560c797 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -86,6 +86,7 @@ static const char *amdgpu_asic_name[] = { "VEGAM", "VEGA10", "VEGA12", + "VEGA20", "RAVEN", "LAST", }; diff --git a/include/drm/amd_asic_type.h b/include/drm/amd_asic_type.h index 695bde7eb055..dd63d08cc54e 100644 --- a/include/drm/amd_asic_type.h +++ b/include/drm/amd_asic_type.h @@ -47,6 +47,7 @@ enum amd_asic_type { CHIP_VEGAM, CHIP_VEGA10, CHIP_VEGA12, + CHIP_VEGA20, CHIP_RAVEN, CHIP_LAST, }; -- cgit v1.2.3 From 24c94e166dfe89839129b8e0fae208b6af60d6f1 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Sat, 5 May 2018 08:45:47 +0200 Subject: gpu: host1x: Remove wait check support The job submission userspace ABI doesn't support this and there are no plans to implement it, so all of this code is dead and can be removed. Reviewed-by: Dmitry Osipenko Tested-by: Dmitry Osipenko Signed-off-by: Thierry Reding --- drivers/gpu/drm/tegra/drm.c | 62 +------------------ drivers/gpu/host1x/dev.h | 8 --- drivers/gpu/host1x/hw/channel_hw.c | 3 +- drivers/gpu/host1x/hw/syncpt_hw.c | 11 ---- drivers/gpu/host1x/job.c | 124 +------------------------------------ drivers/gpu/host1x/syncpt.c | 6 -- drivers/gpu/host1x/syncpt.h | 3 - include/linux/host1x.h | 15 +---- include/trace/events/host1x.h | 16 +++-- 9 files changed, 14 insertions(+), 234 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c index 3cdef659cd39..204b10e33f16 100644 --- a/drivers/gpu/drm/tegra/drm.c +++ b/drivers/gpu/drm/tegra/drm.c @@ -321,46 +321,14 @@ static int host1x_reloc_copy_from_user(struct host1x_reloc *dest, return 0; } -static int host1x_waitchk_copy_from_user(struct host1x_waitchk *dest, - struct drm_tegra_waitchk __user *src, - struct drm_file *file) -{ - u32 cmdbuf; - int err; - - err = get_user(cmdbuf, &src->handle); - if (err < 0) - return err; - - err = get_user(dest->offset, &src->offset); - if (err < 0) - return err; - - err = get_user(dest->syncpt_id, &src->syncpt); - if (err < 0) - return err; - - err = get_user(dest->thresh, &src->thresh); - if (err < 0) - return err; - - dest->bo = host1x_bo_lookup(file, cmdbuf); - if (!dest->bo) - return -ENOENT; - - return 0; -} - int tegra_drm_submit(struct tegra_drm_context *context, struct drm_tegra_submit *args, struct drm_device *drm, struct drm_file *file) { unsigned int num_cmdbufs = args->num_cmdbufs; unsigned int num_relocs = args->num_relocs; - unsigned int num_waitchks = args->num_waitchks; struct drm_tegra_cmdbuf __user *user_cmdbufs; struct drm_tegra_reloc __user *user_relocs; - struct drm_tegra_waitchk __user *user_waitchks; struct drm_tegra_syncpt __user *user_syncpt; struct drm_tegra_syncpt syncpt; struct host1x *host1x = dev_get_drvdata(drm->dev->parent); @@ -372,7 +340,6 @@ int tegra_drm_submit(struct tegra_drm_context *context, user_cmdbufs = u64_to_user_ptr(args->cmdbufs); user_relocs = u64_to_user_ptr(args->relocs); - user_waitchks = u64_to_user_ptr(args->waitchks); user_syncpt = u64_to_user_ptr(args->syncpts); /* We don't yet support other than one syncpt_incr struct per submit */ @@ -384,12 +351,11 @@ int tegra_drm_submit(struct tegra_drm_context *context, return -EINVAL; job = host1x_job_alloc(context->channel, args->num_cmdbufs, - args->num_relocs, args->num_waitchks); + args->num_relocs); if (!job) return -ENOMEM; job->num_relocs = args->num_relocs; - job->num_waitchk = args->num_waitchks; job->client = (u32)args->context; job->class = context->client->base.class; job->serialize = true; @@ -398,7 +364,7 @@ int tegra_drm_submit(struct tegra_drm_context *context, * Track referenced BOs so that they can be unreferenced after the * submission is complete. */ - num_refs = num_cmdbufs + num_relocs * 2 + num_waitchks; + num_refs = num_cmdbufs + num_relocs * 2; refs = kmalloc_array(num_refs, sizeof(*refs), GFP_KERNEL); if (!refs) { @@ -489,30 +455,6 @@ int tegra_drm_submit(struct tegra_drm_context *context, } } - /* copy and resolve waitchks from submit */ - while (num_waitchks--) { - struct host1x_waitchk *wait = &job->waitchk[num_waitchks]; - struct tegra_bo *obj; - - err = host1x_waitchk_copy_from_user( - wait, &user_waitchks[num_waitchks], file); - if (err < 0) - goto fail; - - obj = host1x_to_tegra_bo(wait->bo); - refs[num_refs++] = &obj->gem; - - /* - * The unaligned offset will cause an unaligned write during - * of the waitchks patching, corrupting the commands stream. - */ - if (wait->offset & 3 || - wait->offset >= obj->gem.size) { - err = -EINVAL; - goto fail; - } - } - if (copy_from_user(&syncpt, user_syncpt, sizeof(syncpt))) { err = -EFAULT; goto fail; diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h index 43e9fabb43a1..36f44ffebe73 100644 --- a/drivers/gpu/host1x/dev.h +++ b/drivers/gpu/host1x/dev.h @@ -78,7 +78,6 @@ struct host1x_syncpt_ops { void (*load_wait_base)(struct host1x_syncpt *syncpt); u32 (*load)(struct host1x_syncpt *syncpt); int (*cpu_incr)(struct host1x_syncpt *syncpt); - int (*patch_wait)(struct host1x_syncpt *syncpt, void *patch_addr); void (*assign_to_channel)(struct host1x_syncpt *syncpt, struct host1x_channel *channel); void (*enable_protection)(struct host1x *host); @@ -183,13 +182,6 @@ static inline int host1x_hw_syncpt_cpu_incr(struct host1x *host, return host->syncpt_op->cpu_incr(sp); } -static inline int host1x_hw_syncpt_patch_wait(struct host1x *host, - struct host1x_syncpt *sp, - void *patch_addr) -{ - return host->syncpt_op->patch_wait(sp, patch_addr); -} - static inline void host1x_hw_syncpt_assign_to_channel( struct host1x *host, struct host1x_syncpt *sp, struct host1x_channel *ch) diff --git a/drivers/gpu/host1x/hw/channel_hw.c b/drivers/gpu/host1x/hw/channel_hw.c index 9af758785a11..4c9555038a95 100644 --- a/drivers/gpu/host1x/hw/channel_hw.c +++ b/drivers/gpu/host1x/hw/channel_hw.c @@ -104,8 +104,7 @@ static int channel_submit(struct host1x_job *job) sp = host->syncpt + job->syncpt_id; trace_host1x_channel_submit(dev_name(ch->dev), job->num_gathers, job->num_relocs, - job->num_waitchk, job->syncpt_id, - job->syncpt_incrs); + job->syncpt_id, job->syncpt_incrs); /* before error checks, return current max */ prev_max = job->syncpt_end = host1x_syncpt_read_max(sp); diff --git a/drivers/gpu/host1x/hw/syncpt_hw.c b/drivers/gpu/host1x/hw/syncpt_hw.c index 7dfd47d74f89..a23bb3352d02 100644 --- a/drivers/gpu/host1x/hw/syncpt_hw.c +++ b/drivers/gpu/host1x/hw/syncpt_hw.c @@ -96,16 +96,6 @@ static int syncpt_cpu_incr(struct host1x_syncpt *sp) return 0; } -/* remove a wait pointed to by patch_addr */ -static int syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr) -{ - u32 override = host1x_class_host_wait_syncpt(HOST1X_SYNCPT_RESERVED, 0); - - *((u32 *)patch_addr) = override; - - return 0; -} - /** * syncpt_assign_to_channel() - Assign syncpoint to channel * @sp: syncpoint @@ -156,7 +146,6 @@ static const struct host1x_syncpt_ops host1x_syncpt_ops = { .load_wait_base = syncpt_read_wait_base, .load = syncpt_load, .cpu_incr = syncpt_cpu_incr, - .patch_wait = syncpt_patch_wait, .assign_to_channel = syncpt_assign_to_channel, .enable_protection = syncpt_enable_protection, }; diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c index db509ab8874e..3cbfc6e37668 100644 --- a/drivers/gpu/host1x/job.c +++ b/drivers/gpu/host1x/job.c @@ -34,8 +34,7 @@ #define HOST1X_WAIT_SYNCPT_OFFSET 0x8 struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, - u32 num_cmdbufs, u32 num_relocs, - u32 num_waitchks) + u32 num_cmdbufs, u32 num_relocs) { struct host1x_job *job = NULL; unsigned int num_unpins = num_cmdbufs + num_relocs; @@ -46,7 +45,6 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, total = sizeof(struct host1x_job) + (u64)num_relocs * sizeof(struct host1x_reloc) + (u64)num_unpins * sizeof(struct host1x_job_unpin_data) + - (u64)num_waitchks * sizeof(struct host1x_waitchk) + (u64)num_cmdbufs * sizeof(struct host1x_job_gather) + (u64)num_unpins * sizeof(dma_addr_t) + (u64)num_unpins * sizeof(u32 *); @@ -66,8 +64,6 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, mem += num_relocs * sizeof(struct host1x_reloc); job->unpins = num_unpins ? mem : NULL; mem += num_unpins * sizeof(struct host1x_job_unpin_data); - job->waitchk = num_waitchks ? mem : NULL; - mem += num_waitchks * sizeof(struct host1x_waitchk); job->gathers = num_cmdbufs ? mem : NULL; mem += num_cmdbufs * sizeof(struct host1x_job_gather); job->addr_phys = num_unpins ? mem : NULL; @@ -111,73 +107,6 @@ void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo, } EXPORT_SYMBOL(host1x_job_add_gather); -/* - * NULL an already satisfied WAIT_SYNCPT host method, by patching its - * args in the command stream. The method data is changed to reference - * a reserved (never given out or incr) HOST1X_SYNCPT_RESERVED syncpt - * with a matching threshold value of 0, so is guaranteed to be popped - * by the host HW. - */ -static void host1x_syncpt_patch_offset(struct host1x_syncpt *sp, - struct host1x_bo *h, u32 offset) -{ - void *patch_addr = NULL; - - /* patch the wait */ - patch_addr = host1x_bo_kmap(h, offset >> PAGE_SHIFT); - if (patch_addr) { - host1x_syncpt_patch_wait(sp, - patch_addr + (offset & ~PAGE_MASK)); - host1x_bo_kunmap(h, offset >> PAGE_SHIFT, patch_addr); - } else - pr_err("Could not map cmdbuf for wait check\n"); -} - -/* - * Check driver supplied waitchk structs for syncpt thresholds - * that have already been satisfied and NULL the comparison (to - * avoid a wrap condition in the HW). - */ -static int do_waitchks(struct host1x_job *job, struct host1x *host, - struct host1x_job_gather *g) -{ - struct host1x_bo *patch = g->bo; - int i; - - /* compare syncpt vs wait threshold */ - for (i = 0; i < job->num_waitchk; i++) { - struct host1x_waitchk *wait = &job->waitchk[i]; - struct host1x_syncpt *sp = - host1x_syncpt_get(host, wait->syncpt_id); - - /* validate syncpt id */ - if (wait->syncpt_id > host1x_syncpt_nb_pts(host)) - continue; - - /* skip all other gathers */ - if (patch != wait->bo) - continue; - - trace_host1x_syncpt_wait_check(wait->bo, wait->offset, - wait->syncpt_id, wait->thresh, - host1x_syncpt_read_min(sp)); - - if (host1x_syncpt_is_expired(sp, wait->thresh)) { - dev_dbg(host->dev, - "drop WAIT id %u (%s) thresh 0x%x, min 0x%x\n", - wait->syncpt_id, sp->name, wait->thresh, - host1x_syncpt_read_min(sp)); - - host1x_syncpt_patch_offset(sp, patch, - g->offset + wait->offset); - } - - wait->bo = NULL; - } - - return 0; -} - static unsigned int pin_job(struct host1x *host, struct host1x_job *job) { unsigned int i; @@ -331,17 +260,6 @@ static bool check_reloc(struct host1x_reloc *reloc, struct host1x_bo *cmdbuf, return true; } -static bool check_wait(struct host1x_waitchk *wait, struct host1x_bo *cmdbuf, - unsigned int offset) -{ - offset *= sizeof(u32); - - if (wait->bo != cmdbuf || wait->offset != offset) - return false; - - return true; -} - struct host1x_firewall { struct host1x_job *job; struct device *dev; @@ -349,9 +267,6 @@ struct host1x_firewall { unsigned int num_relocs; struct host1x_reloc *reloc; - unsigned int num_waitchks; - struct host1x_waitchk *waitchk; - struct host1x_bo *cmdbuf; unsigned int offset; @@ -378,20 +293,6 @@ static int check_register(struct host1x_firewall *fw, unsigned long offset) fw->reloc++; } - if (offset == HOST1X_WAIT_SYNCPT_OFFSET) { - if (fw->class != HOST1X_CLASS_HOST1X) - return -EINVAL; - - if (!fw->num_waitchks) - return -EINVAL; - - if (!check_wait(fw->waitchk, fw->cmdbuf, fw->offset)) - return -EINVAL; - - fw->num_waitchks--; - fw->waitchk++; - } - return 0; } @@ -556,8 +457,6 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev) fw.dev = dev; fw.reloc = job->relocarray; fw.num_relocs = job->num_relocs; - fw.waitchk = job->waitchk; - fw.num_waitchks = job->num_waitchk; fw.class = job->class; for (i = 0; i < job->num_gathers; i++) { @@ -604,8 +503,8 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev) offset += g->words * sizeof(u32); } - /* No relocs and waitchks should remain at this point */ - if (fw.num_relocs || fw.num_waitchks) + /* No relocs should remain at this point */ + if (fw.num_relocs) return -EINVAL; return 0; @@ -616,19 +515,6 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev) int err; unsigned int i, j; struct host1x *host = dev_get_drvdata(dev->parent); - DECLARE_BITMAP(waitchk_mask, host1x_syncpt_nb_pts(host)); - - bitmap_zero(waitchk_mask, host1x_syncpt_nb_pts(host)); - for (i = 0; i < job->num_waitchk; i++) { - u32 syncpt_id = job->waitchk[i].syncpt_id; - - if (syncpt_id < host1x_syncpt_nb_pts(host)) - set_bit(syncpt_id, waitchk_mask); - } - - /* get current syncpt values for waitchk */ - for_each_set_bit(i, waitchk_mask, host1x_syncpt_nb_pts(host)) - host1x_syncpt_load(host->syncpt + i); /* pin memory */ err = pin_job(host, job); @@ -663,10 +549,6 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev) err = do_relocs(job, g); if (err) break; - - err = do_waitchks(job, host, g); - if (err) - break; } out: diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c index a2a952adc136..a108669188e8 100644 --- a/drivers/gpu/host1x/syncpt.c +++ b/drivers/gpu/host1x/syncpt.c @@ -373,12 +373,6 @@ bool host1x_syncpt_is_expired(struct host1x_syncpt *sp, u32 thresh) return (s32)(current_val - thresh) >= 0; } -/* remove a wait pointed to by patch_addr */ -int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr) -{ - return host1x_hw_syncpt_patch_wait(sp->host, sp, patch_addr); -} - int host1x_syncpt_init(struct host1x *host) { struct host1x_syncpt_base *bases; diff --git a/drivers/gpu/host1x/syncpt.h b/drivers/gpu/host1x/syncpt.h index 9d88d37c2397..d98e22325e9d 100644 --- a/drivers/gpu/host1x/syncpt.h +++ b/drivers/gpu/host1x/syncpt.h @@ -124,7 +124,4 @@ static inline int host1x_syncpt_is_valid(struct host1x_syncpt *sp) return sp->id < host1x_syncpt_nb_pts(sp->host); } -/* Patch a wait by replacing it with a wait for syncpt 0 value 0 */ -int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr); - #endif diff --git a/include/linux/host1x.h b/include/linux/host1x.h index ddf7f9ca86cc..f66bece1e1b7 100644 --- a/include/linux/host1x.h +++ b/include/linux/host1x.h @@ -192,13 +192,6 @@ struct host1x_reloc { unsigned long shift; }; -struct host1x_waitchk { - struct host1x_bo *bo; - u32 offset; - u32 syncpt_id; - u32 thresh; -}; - struct host1x_job { /* When refcount goes to zero, job can be freed */ struct kref ref; @@ -215,11 +208,6 @@ struct host1x_job { struct host1x_job_gather *gathers; unsigned int num_gathers; - /* Wait checks to be processed at submit time */ - struct host1x_waitchk *waitchk; - unsigned int num_waitchk; - u32 waitchk_mask; - /* Array of handles to be pinned & unpinned */ struct host1x_reloc *relocarray; unsigned int num_relocs; @@ -261,8 +249,7 @@ struct host1x_job { }; struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, - u32 num_cmdbufs, u32 num_relocs, - u32 num_waitchks); + u32 num_cmdbufs, u32 num_relocs); void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *mem_id, u32 words, u32 offset); struct host1x_job *host1x_job_get(struct host1x_job *job); diff --git a/include/trace/events/host1x.h b/include/trace/events/host1x.h index 63116362543c..a37ef73092e5 100644 --- a/include/trace/events/host1x.h +++ b/include/trace/events/host1x.h @@ -115,16 +115,15 @@ TRACE_EVENT(host1x_cdma_push_gather, ); TRACE_EVENT(host1x_channel_submit, - TP_PROTO(const char *name, u32 cmdbufs, u32 relocs, u32 waitchks, - u32 syncpt_id, u32 syncpt_incrs), + TP_PROTO(const char *name, u32 cmdbufs, u32 relocs, u32 syncpt_id, + u32 syncpt_incrs), - TP_ARGS(name, cmdbufs, relocs, waitchks, syncpt_id, syncpt_incrs), + TP_ARGS(name, cmdbufs, relocs, syncpt_id, syncpt_incrs), TP_STRUCT__entry( __field(const char *, name) __field(u32, cmdbufs) __field(u32, relocs) - __field(u32, waitchks) __field(u32, syncpt_id) __field(u32, syncpt_incrs) ), @@ -133,15 +132,14 @@ TRACE_EVENT(host1x_channel_submit, __entry->name = name; __entry->cmdbufs = cmdbufs; __entry->relocs = relocs; - __entry->waitchks = waitchks; __entry->syncpt_id = syncpt_id; __entry->syncpt_incrs = syncpt_incrs; ), - TP_printk("name=%s, cmdbufs=%u, relocs=%u, waitchks=%d," - "syncpt_id=%u, syncpt_incrs=%u", - __entry->name, __entry->cmdbufs, __entry->relocs, __entry->waitchks, - __entry->syncpt_id, __entry->syncpt_incrs) + TP_printk("name=%s, cmdbufs=%u, relocs=%u, syncpt_id=%u, " + "syncpt_incrs=%u", + __entry->name, __entry->cmdbufs, __entry->relocs, + __entry->syncpt_id, __entry->syncpt_incrs) ); TRACE_EVENT(host1x_channel_submitted, -- cgit v1.2.3 From bf3d41ccabb53c57e19fcfc8b81d790043ac2bed Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Wed, 16 May 2018 14:12:33 +0200 Subject: gpu: host1x: Store pointer to client in jobs Rather than storing some identifier derived from the application context that can't be used concretely anywhere, store a pointer to the client directly so that accesses can be made directly through that client object. Reviewed-by: Dmitry Osipenko Tested-by: Dmitry Osipenko Signed-off-by: Thierry Reding --- drivers/gpu/drm/tegra/drm.c | 5 +++-- drivers/gpu/host1x/cdma.c | 2 +- drivers/gpu/host1x/cdma.h | 2 +- include/linux/host1x.h | 3 ++- 4 files changed, 7 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c index 204b10e33f16..8f29323611dd 100644 --- a/drivers/gpu/drm/tegra/drm.c +++ b/drivers/gpu/drm/tegra/drm.c @@ -325,6 +325,7 @@ int tegra_drm_submit(struct tegra_drm_context *context, struct drm_tegra_submit *args, struct drm_device *drm, struct drm_file *file) { + struct host1x_client *client = &context->client->base; unsigned int num_cmdbufs = args->num_cmdbufs; unsigned int num_relocs = args->num_relocs; struct drm_tegra_cmdbuf __user *user_cmdbufs; @@ -356,8 +357,8 @@ int tegra_drm_submit(struct tegra_drm_context *context, return -ENOMEM; job->num_relocs = args->num_relocs; - job->client = (u32)args->context; - job->class = context->client->base.class; + job->client = client; + job->class = client->class; job->serialize = true; /* diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c index 69bb77372ed9..91df51e631b2 100644 --- a/drivers/gpu/host1x/cdma.c +++ b/drivers/gpu/host1x/cdma.c @@ -247,7 +247,7 @@ static void cdma_start_timer_locked(struct host1x_cdma *cdma, static void stop_cdma_timer_locked(struct host1x_cdma *cdma) { cancel_delayed_work(&cdma->timeout.wq); - cdma->timeout.client = 0; + cdma->timeout.client = NULL; } /* diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h index 446ee1a84969..e97e17b82370 100644 --- a/drivers/gpu/host1x/cdma.h +++ b/drivers/gpu/host1x/cdma.h @@ -58,7 +58,7 @@ struct buffer_timeout { u32 syncpt_val; /* syncpt value when completed */ ktime_t start_ktime; /* starting time */ /* context timeout information */ - int client; + struct host1x_client *client; }; enum cdma_event { diff --git a/include/linux/host1x.h b/include/linux/host1x.h index f66bece1e1b7..0632010f47fb 100644 --- a/include/linux/host1x.h +++ b/include/linux/host1x.h @@ -202,7 +202,8 @@ struct host1x_job { /* Channel where job is submitted to */ struct host1x_channel *channel; - u32 client; + /* client where the job originated */ + struct host1x_client *client; /* Gathers and their memory */ struct host1x_job_gather *gathers; -- cgit v1.2.3 From 06490bb99e1840ab2b6814af7356e8b4ab0e3ee6 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Wed, 16 May 2018 16:58:44 +0200 Subject: gpu: host1x: Rename relocarray -> relocs for consistency All other array variables use a plural, and this is the only one using the *array suffix. This is confusing, so rename it for consistency. Reviewed-by: Dmitry Osipenko Tested-by: Dmitry Osipenko Signed-off-by: Thierry Reding --- drivers/gpu/drm/tegra/drm.c | 4 ++-- drivers/gpu/host1x/job.c | 8 ++++---- include/linux/host1x.h | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c index 8f29323611dd..bfbd3a89c26f 100644 --- a/drivers/gpu/drm/tegra/drm.c +++ b/drivers/gpu/drm/tegra/drm.c @@ -426,13 +426,13 @@ int tegra_drm_submit(struct tegra_drm_context *context, struct host1x_reloc *reloc; struct tegra_bo *obj; - err = host1x_reloc_copy_from_user(&job->relocarray[num_relocs], + err = host1x_reloc_copy_from_user(&job->relocs[num_relocs], &user_relocs[num_relocs], drm, file); if (err < 0) goto fail; - reloc = &job->relocarray[num_relocs]; + reloc = &job->relocs[num_relocs]; obj = host1x_to_tegra_bo(reloc->cmdbuf.bo); refs[num_refs++] = &obj->gem; diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c index 2be0bcaf8288..9d6d3e151291 100644 --- a/drivers/gpu/host1x/job.c +++ b/drivers/gpu/host1x/job.c @@ -60,7 +60,7 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, /* Redistribute memory to the structs */ mem += sizeof(struct host1x_job); - job->relocarray = num_relocs ? mem : NULL; + job->relocs = num_relocs ? mem : NULL; mem += num_relocs * sizeof(struct host1x_reloc); job->unpins = num_unpins ? mem : NULL; mem += num_unpins * sizeof(struct host1x_job_unpin_data); @@ -115,7 +115,7 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job) job->num_unpins = 0; for (i = 0; i < job->num_relocs; i++) { - struct host1x_reloc *reloc = &job->relocarray[i]; + struct host1x_reloc *reloc = &job->relocs[i]; struct sg_table *sgt; dma_addr_t phys_addr; @@ -203,7 +203,7 @@ static int do_relocs(struct host1x_job *job, struct host1x_job_gather *g) /* pin & patch the relocs for one gather */ for (i = 0; i < job->num_relocs; i++) { - struct host1x_reloc *reloc = &job->relocarray[i]; + struct host1x_reloc *reloc = &job->relocs[i]; u32 reloc_addr = (job->reloc_addr_phys[i] + reloc->target.offset) >> reloc->shift; u32 *target; @@ -455,7 +455,7 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev) fw.job = job; fw.dev = dev; - fw.reloc = job->relocarray; + fw.reloc = job->relocs; fw.num_relocs = job->num_relocs; fw.class = job->class; diff --git a/include/linux/host1x.h b/include/linux/host1x.h index 0632010f47fb..dcb6140d39d7 100644 --- a/include/linux/host1x.h +++ b/include/linux/host1x.h @@ -210,7 +210,7 @@ struct host1x_job { unsigned int num_gathers; /* Array of handles to be pinned & unpinned */ - struct host1x_reloc *relocarray; + struct host1x_reloc *relocs; unsigned int num_relocs; struct host1x_job_unpin_data *unpins; unsigned int num_unpins; -- cgit v1.2.3 From 326bbd79fd61716841585a52d5b68f48f4e6644e Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Wed, 16 May 2018 17:01:43 +0200 Subject: gpu: host1x: Use not explicitly sized types The number of words and the offset in a gather don't need to be explicitly sized, so make them unsigned int instead. Reviewed-by: Dmitry Osipenko Tested-by: Dmitry Osipenko Signed-off-by: Thierry Reding --- drivers/gpu/host1x/job.c | 11 ++++++----- drivers/gpu/host1x/job.h | 4 ++-- include/linux/host1x.h | 4 ++-- 3 files changed, 10 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c index 9d6d3e151291..e2f4a4d93d20 100644 --- a/drivers/gpu/host1x/job.c +++ b/drivers/gpu/host1x/job.c @@ -96,13 +96,14 @@ void host1x_job_put(struct host1x_job *job) EXPORT_SYMBOL(host1x_job_put); void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo, - u32 words, u32 offset) + unsigned int words, unsigned int offset) { - struct host1x_job_gather *cur_gather = &job->gathers[job->num_gathers]; + struct host1x_job_gather *gather = &job->gathers[job->num_gathers]; + + gather->words = words; + gather->bo = bo; + gather->offset = offset; - cur_gather->words = words; - cur_gather->bo = bo; - cur_gather->offset = offset; job->num_gathers++; } EXPORT_SYMBOL(host1x_job_add_gather); diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h index 4bda51d503ec..188400e00192 100644 --- a/drivers/gpu/host1x/job.h +++ b/drivers/gpu/host1x/job.h @@ -20,10 +20,10 @@ #define __HOST1X_JOB_H struct host1x_job_gather { - u32 words; + unsigned int words; dma_addr_t base; struct host1x_bo *bo; - u32 offset; + unsigned int offset; bool handled; }; diff --git a/include/linux/host1x.h b/include/linux/host1x.h index dcb6140d39d7..89110d896d72 100644 --- a/include/linux/host1x.h +++ b/include/linux/host1x.h @@ -251,8 +251,8 @@ struct host1x_job { struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, u32 num_cmdbufs, u32 num_relocs); -void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *mem_id, - u32 words, u32 offset); +void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo, + unsigned int words, unsigned int offset); struct host1x_job *host1x_job_get(struct host1x_job *job); void host1x_job_put(struct host1x_job *job); int host1x_job_pin(struct host1x_job *job, struct device *dev); -- cgit v1.2.3 From c850ece71f71c2a68a9921c52fb5fd8d3ec2b8d7 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Sat, 5 May 2018 08:12:53 +0200 Subject: drm/tegra: Use proper arguments for DRM_TEGRA_CLOSE_CHANNEL IOCTL A separate data structure exists for the DRM_TEGRA_CLOSE_CHANNEL IOCTL, but it is currently unused. The IOCTL was using the data structure for the DRM_TEGRA_OPEN_CHANNEL IOCTL. Reviewed-by: Dmitry Osipenko Tested-by: Dmitry Osipenko Signed-off-by: Thierry Reding --- include/uapi/drm/tegra_drm.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/uapi/drm/tegra_drm.h b/include/uapi/drm/tegra_drm.h index d954f8c33321..99e15d82d1e9 100644 --- a/include/uapi/drm/tegra_drm.h +++ b/include/uapi/drm/tegra_drm.h @@ -193,7 +193,7 @@ struct drm_tegra_gem_get_flags { #define DRM_IOCTL_TEGRA_SYNCPT_INCR DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_SYNCPT_INCR, struct drm_tegra_syncpt_incr) #define DRM_IOCTL_TEGRA_SYNCPT_WAIT DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_SYNCPT_WAIT, struct drm_tegra_syncpt_wait) #define DRM_IOCTL_TEGRA_OPEN_CHANNEL DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_OPEN_CHANNEL, struct drm_tegra_open_channel) -#define DRM_IOCTL_TEGRA_CLOSE_CHANNEL DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_CLOSE_CHANNEL, struct drm_tegra_open_channel) +#define DRM_IOCTL_TEGRA_CLOSE_CHANNEL DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_CLOSE_CHANNEL, struct drm_tegra_close_channel) #define DRM_IOCTL_TEGRA_GET_SYNCPT DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_GET_SYNCPT, struct drm_tegra_get_syncpt) #define DRM_IOCTL_TEGRA_SUBMIT DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_SUBMIT, struct drm_tegra_submit) #define DRM_IOCTL_TEGRA_GET_SYNCPT_BASE DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_GET_SYNCPT_BASE, struct drm_tegra_get_syncpt_base) -- cgit v1.2.3 From 563e1e664d27292a3b55ca08366dc8c32db52450 Mon Sep 17 00:00:00 2001 From: Andrey Grodzovsky Date: Tue, 15 May 2018 14:42:20 -0400 Subject: drm/scheduler: Remove obsolete spinlock. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This spinlock is superfluous, any call to drm_sched_entity_push_job should already be under a lock together with matching drm_sched_job_init to match the order of insertion into queue with job's fence seqence number. v2: Improve patch description. Add functions documentation describing the locking considerations Signed-off-by: Andrey Grodzovsky Acked-by: Chunming Zhou Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/scheduler/gpu_scheduler.c | 15 ++++++++++----- include/drm/gpu_scheduler.h | 1 - 2 files changed, 10 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c index a364fc0b38c3..df1578d6f42e 100644 --- a/drivers/gpu/drm/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c @@ -139,7 +139,6 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched, entity->last_scheduled = NULL; spin_lock_init(&entity->rq_lock); - spin_lock_init(&entity->queue_lock); spsc_queue_init(&entity->job_queue); atomic_set(&entity->fence_seq, 0); @@ -413,6 +412,10 @@ drm_sched_entity_pop_job(struct drm_sched_entity *entity) * * @sched_job The pointer to job required to submit * + * Note: To guarantee that the order of insertion to queue matches + * the job's fence sequence number this function should be + * called with drm_sched_job_init under common lock. + * * Returns 0 for success, negative error code otherwise. */ void drm_sched_entity_push_job(struct drm_sched_job *sched_job, @@ -423,11 +426,8 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job, trace_drm_sched_job(sched_job, entity); - spin_lock(&entity->queue_lock); first = spsc_queue_push(&entity->job_queue, &sched_job->queue_node); - spin_unlock(&entity->queue_lock); - /* first job wakes up scheduler */ if (first) { /* Add the entity to the run queue */ @@ -593,7 +593,12 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched) } EXPORT_SYMBOL(drm_sched_job_recovery); -/* init a sched_job with basic field */ +/** + * Init a sched_job with basic field + * + * Note: Refer to drm_sched_entity_push_job documentation + * for locking considerations. + */ int drm_sched_job_init(struct drm_sched_job *job, struct drm_gpu_scheduler *sched, struct drm_sched_entity *entity, diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index 52380067a43f..dec655894d08 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -56,7 +56,6 @@ struct drm_sched_entity { spinlock_t rq_lock; struct drm_gpu_scheduler *sched; - spinlock_t queue_lock; struct spsc_queue job_queue; atomic_t fence_seq; -- cgit v1.2.3 From 6134534ca24f42043cacdd7108026803577f6c59 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Wed, 16 May 2018 16:43:11 +0200 Subject: drm/tegra: Add kerneldoc for UAPI Document the userspace ABI with kerneldoc to provide some information on how to use it. v3: - reword description of arrays and array lengths v2: - keep GEM object creation flags for ABI compatibility - fix typo in struct drm_tegra_syncpt_incr kerneldoc - fix typos in struct drm_tegra_submit kerneldoc - reworded some descriptions as suggested Reviewed-by: Dmitry Osipenko Signed-off-by: Thierry Reding --- include/uapi/drm/tegra_drm.h | 490 ++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 481 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/uapi/drm/tegra_drm.h b/include/uapi/drm/tegra_drm.h index 99e15d82d1e9..c4df3c3668b3 100644 --- a/include/uapi/drm/tegra_drm.h +++ b/include/uapi/drm/tegra_drm.h @@ -32,143 +32,615 @@ extern "C" { #define DRM_TEGRA_GEM_CREATE_TILED (1 << 0) #define DRM_TEGRA_GEM_CREATE_BOTTOM_UP (1 << 1) +/** + * struct drm_tegra_gem_create - parameters for the GEM object creation IOCTL + */ struct drm_tegra_gem_create { + /** + * @size: + * + * The size, in bytes, of the buffer object to be created. + */ __u64 size; + + /** + * @flags: + * + * A bitmask of flags that influence the creation of GEM objects: + * + * DRM_TEGRA_GEM_CREATE_TILED + * Use the 16x16 tiling format for this buffer. + * + * DRM_TEGRA_GEM_CREATE_BOTTOM_UP + * The buffer has a bottom-up layout. + */ __u32 flags; + + /** + * @handle: + * + * The handle of the created GEM object. Set by the kernel upon + * successful completion of the IOCTL. + */ __u32 handle; }; +/** + * struct drm_tegra_gem_mmap - parameters for the GEM mmap IOCTL + */ struct drm_tegra_gem_mmap { + /** + * @handle: + * + * Handle of the GEM object to obtain an mmap offset for. + */ __u32 handle; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; + + /** + * @offset: + * + * The mmap offset for the given GEM object. Set by the kernel upon + * successful completion of the IOCTL. + */ __u64 offset; }; +/** + * struct drm_tegra_syncpt_read - parameters for the read syncpoint IOCTL + */ struct drm_tegra_syncpt_read { + /** + * @id: + * + * ID of the syncpoint to read the current value from. + */ __u32 id; + + /** + * @value: + * + * The current syncpoint value. Set by the kernel upon successful + * completion of the IOCTL. + */ __u32 value; }; +/** + * struct drm_tegra_syncpt_incr - parameters for the increment syncpoint IOCTL + */ struct drm_tegra_syncpt_incr { + /** + * @id: + * + * ID of the syncpoint to increment. + */ __u32 id; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; }; +/** + * struct drm_tegra_syncpt_wait - parameters for the wait syncpoint IOCTL + */ struct drm_tegra_syncpt_wait { + /** + * @id: + * + * ID of the syncpoint to wait on. + */ __u32 id; + + /** + * @thresh: + * + * Threshold value for which to wait. + */ __u32 thresh; + + /** + * @timeout: + * + * Timeout, in milliseconds, to wait. + */ __u32 timeout; + + /** + * @value: + * + * The new syncpoint value after the wait. Set by the kernel upon + * successful completion of the IOCTL. + */ __u32 value; }; #define DRM_TEGRA_NO_TIMEOUT (0xffffffff) +/** + * struct drm_tegra_open_channel - parameters for the open channel IOCTL + */ struct drm_tegra_open_channel { + /** + * @client: + * + * The client ID for this channel. + */ __u32 client; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; + + /** + * @context: + * + * The application context of this channel. Set by the kernel upon + * successful completion of the IOCTL. This context needs to be passed + * to the DRM_TEGRA_CHANNEL_CLOSE or the DRM_TEGRA_SUBMIT IOCTLs. + */ __u64 context; }; +/** + * struct drm_tegra_close_channel - parameters for the close channel IOCTL + */ struct drm_tegra_close_channel { + /** + * @context: + * + * The application context of this channel. This is obtained from the + * DRM_TEGRA_OPEN_CHANNEL IOCTL. + */ __u64 context; }; +/** + * struct drm_tegra_get_syncpt - parameters for the get syncpoint IOCTL + */ struct drm_tegra_get_syncpt { + /** + * @context: + * + * The application context identifying the channel for which to obtain + * the syncpoint ID. + */ __u64 context; + + /** + * @index: + * + * Index of the client syncpoint for which to obtain the ID. + */ __u32 index; + + /** + * @id: + * + * The ID of the given syncpoint. Set by the kernel upon successful + * completion of the IOCTL. + */ __u32 id; }; +/** + * struct drm_tegra_get_syncpt_base - parameters for the get wait base IOCTL + */ struct drm_tegra_get_syncpt_base { + /** + * @context: + * + * The application context identifying for which channel to obtain the + * wait base. + */ __u64 context; + + /** + * @syncpt: + * + * ID of the syncpoint for which to obtain the wait base. + */ __u32 syncpt; + + /** + * @id: + * + * The ID of the wait base corresponding to the client syncpoint. Set + * by the kernel upon successful completion of the IOCTL. + */ __u32 id; }; +/** + * struct drm_tegra_syncpt - syncpoint increment operation + */ struct drm_tegra_syncpt { + /** + * @id: + * + * ID of the syncpoint to operate on. + */ __u32 id; + + /** + * @incrs: + * + * Number of increments to perform for the syncpoint. + */ __u32 incrs; }; +/** + * struct drm_tegra_cmdbuf - structure describing a command buffer + */ struct drm_tegra_cmdbuf { + /** + * @handle: + * + * Handle to a GEM object containing the command buffer. + */ __u32 handle; + + /** + * @offset: + * + * Offset, in bytes, into the GEM object identified by @handle at + * which the command buffer starts. + */ __u32 offset; + + /** + * @words: + * + * Number of 32-bit words in this command buffer. + */ __u32 words; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; }; +/** + * struct drm_tegra_reloc - GEM object relocation structure + */ struct drm_tegra_reloc { struct { + /** + * @cmdbuf.handle: + * + * Handle to the GEM object containing the command buffer for + * which to perform this GEM object relocation. + */ __u32 handle; + + /** + * @cmdbuf.offset: + * + * Offset, in bytes, into the command buffer at which to + * insert the relocated address. + */ __u32 offset; } cmdbuf; struct { + /** + * @target.handle: + * + * Handle to the GEM object to be relocated. + */ __u32 handle; + + /** + * @target.offset: + * + * Offset, in bytes, into the target GEM object at which the + * relocated data starts. + */ __u32 offset; } target; + + /** + * @shift: + * + * The number of bits by which to shift relocated addresses. + */ __u32 shift; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; }; +/** + * struct drm_tegra_waitchk - wait check structure + */ struct drm_tegra_waitchk { + /** + * @handle: + * + * Handle to the GEM object containing a command stream on which to + * perform the wait check. + */ __u32 handle; + + /** + * @offset: + * + * Offset, in bytes, of the location in the command stream to perform + * the wait check on. + */ __u32 offset; + + /** + * @syncpt: + * + * ID of the syncpoint to wait check. + */ __u32 syncpt; + + /** + * @thresh: + * + * Threshold value for which to check. + */ __u32 thresh; }; +/** + * struct drm_tegra_submit - job submission structure + */ struct drm_tegra_submit { + /** + * @context: + * + * The application context identifying the channel to use for the + * execution of this job. + */ __u64 context; + + /** + * @num_syncpts: + * + * The number of syncpoints operated on by this job. This defines the + * length of the array pointed to by @syncpts. + */ __u32 num_syncpts; + + /** + * @num_cmdbufs: + * + * The number of command buffers to execute as part of this job. This + * defines the length of the array pointed to by @cmdbufs. + */ __u32 num_cmdbufs; + + /** + * @num_relocs: + * + * The number of relocations to perform before executing this job. + * This defines the length of the array pointed to by @relocs. + */ __u32 num_relocs; + + /** + * @num_waitchks: + * + * The number of wait checks to perform as part of this job. This + * defines the length of the array pointed to by @waitchks. + */ __u32 num_waitchks; + + /** + * @waitchk_mask: + * + * Bitmask of valid wait checks. + */ __u32 waitchk_mask; + + /** + * @timeout: + * + * Timeout, in milliseconds, before this job is cancelled. + */ __u32 timeout; + + /** + * @syncpts: + * + * A pointer to an array of &struct drm_tegra_syncpt structures that + * specify the syncpoint operations performed as part of this job. + * The number of elements in the array must be equal to the value + * given by @num_syncpts. + */ __u64 syncpts; + + /** + * @cmdbufs: + * + * A pointer to an array of &struct drm_tegra_cmdbuf structures that + * define the command buffers to execute as part of this job. The + * number of elements in the array must be equal to the value given + * by @num_syncpts. + */ __u64 cmdbufs; + + /** + * @relocs: + * + * A pointer to an array of &struct drm_tegra_reloc structures that + * specify the relocations that need to be performed before executing + * this job. The number of elements in the array must be equal to the + * value given by @num_relocs. + */ __u64 relocs; + + /** + * @waitchks: + * + * A pointer to an array of &struct drm_tegra_waitchk structures that + * specify the wait checks to be performed while executing this job. + * The number of elements in the array must be equal to the value + * given by @num_waitchks. + */ __u64 waitchks; - __u32 fence; /* Return value */ - __u32 reserved[5]; /* future expansion */ + /** + * @fence: + * + * The threshold of the syncpoint associated with this job after it + * has been completed. Set by the kernel upon successful completion of + * the IOCTL. This can be used with the DRM_TEGRA_SYNCPT_WAIT IOCTL to + * wait for this job to be finished. + */ + __u32 fence; + + /** + * @reserved: + * + * This field is reserved for future use. Must be 0. + */ + __u32 reserved[5]; }; #define DRM_TEGRA_GEM_TILING_MODE_PITCH 0 #define DRM_TEGRA_GEM_TILING_MODE_TILED 1 #define DRM_TEGRA_GEM_TILING_MODE_BLOCK 2 +/** + * struct drm_tegra_gem_set_tiling - parameters for the set tiling IOCTL + */ struct drm_tegra_gem_set_tiling { - /* input */ + /** + * @handle: + * + * Handle to the GEM object for which to set the tiling parameters. + */ __u32 handle; + + /** + * @mode: + * + * The tiling mode to set. Must be one of: + * + * DRM_TEGRA_GEM_TILING_MODE_PITCH + * pitch linear format + * + * DRM_TEGRA_GEM_TILING_MODE_TILED + * 16x16 tiling format + * + * DRM_TEGRA_GEM_TILING_MODE_BLOCK + * 16Bx2 tiling format + */ __u32 mode; + + /** + * @value: + * + * The value to set for the tiling mode parameter. + */ __u32 value; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; }; +/** + * struct drm_tegra_gem_get_tiling - parameters for the get tiling IOCTL + */ struct drm_tegra_gem_get_tiling { - /* input */ + /** + * @handle: + * + * Handle to the GEM object for which to query the tiling parameters. + */ __u32 handle; - /* output */ + + /** + * @mode: + * + * The tiling mode currently associated with the GEM object. Set by + * the kernel upon successful completion of the IOCTL. + */ __u32 mode; + + /** + * @value: + * + * The tiling mode parameter currently associated with the GEM object. + * Set by the kernel upon successful completion of the IOCTL. + */ __u32 value; + + /** + * @pad: + * + * Structure padding that may be used in the future. Must be 0. + */ __u32 pad; }; #define DRM_TEGRA_GEM_BOTTOM_UP (1 << 0) #define DRM_TEGRA_GEM_FLAGS (DRM_TEGRA_GEM_BOTTOM_UP) +/** + * struct drm_tegra_gem_set_flags - parameters for the set flags IOCTL + */ struct drm_tegra_gem_set_flags { - /* input */ + /** + * @handle: + * + * Handle to the GEM object for which to set the flags. + */ __u32 handle; - /* output */ + + /** + * @flags: + * + * The flags to set for the GEM object. + */ __u32 flags; }; +/** + * struct drm_tegra_gem_get_flags - parameters for the get flags IOCTL + */ struct drm_tegra_gem_get_flags { - /* input */ + /** + * @handle: + * + * Handle to the GEM object for which to query the flags. + */ __u32 handle; - /* output */ + + /** + * @flags: + * + * The flags currently associated with the GEM object. Set by the + * kernel upon successful completion of the IOCTL. + */ __u32 flags; }; -- cgit v1.2.3