From 0b7018aae7e1798f55f736b9a77c201708aa0e33 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 11 Apr 2006 23:42:03 -0400 Subject: Input: ads7846 - debouncing and rudimentary sample filtering MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Some touchscreens seem to oscillate heavily for a while after touching the screen.  Implement support for sampling the screen until we get two consecutive values that are close enough. Signed-off-by: Imre Deak Signed-off-by: Juha Yrjola Signed-off-by: David Brownell Signed-off-by: Dmitry Torokhov --- include/linux/spi/ads7846.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include') diff --git a/include/linux/spi/ads7846.h b/include/linux/spi/ads7846.h index 72261e0f2ac1..3f7664951256 100644 --- a/include/linux/spi/ads7846.h +++ b/include/linux/spi/ads7846.h @@ -14,5 +14,8 @@ struct ads7846_platform_data { u16 x_min, x_max; u16 y_min, y_max; u16 pressure_min, pressure_max; + + u16 debounce_max; /* max number of readings per sample */ + u16 debounce_tol; /* tolerance used for filtering */ }; -- cgit v1.2.3 From c9e617a563ad646239270fa2222cdb06966cf1fa Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 11 Apr 2006 23:44:05 -0400 Subject: Input: ads7846 - handle IRQs that were latched during disabled IRQs The pen down IRQ will toggle during each X,Y,Z measurement cycle. Even though the IRQ is disabled it will be latched and delivered when after enable_irq. Thus in the IRQ handler we must avoid starting a new measurement cycle when such an "unwanted" IRQ happens. Add a get_pendown_state platform function, which will probably determine this by reading the current GPIO level of the pen IRQ pin. Move the IRQ reenabling from the SPI RX function to the timer. After the last power down message the pen IRQ pin is still active for a while and get_pendown_state would report incorrectly a pen down state. When suspending we should check the ts->pending flag instead of ts->pendown, since the timer can be pending regardless of ts->pendown. Also if ts->pending is set we can be sure that the timer is running, so no need to rearm it. Similarly if ts->pending is not set we can be sure that the IRQ is enabled (and the timer is not). Signed-off-by: Imre Deak Signed-off-by: David Brownell Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/ads7846.c | 82 ++++++++++++++++++++++--------------- include/linux/spi/ads7846.h | 2 + 2 files changed, 50 insertions(+), 34 deletions(-) (limited to 'include') diff --git a/drivers/input/touchscreen/ads7846.c b/drivers/input/touchscreen/ads7846.c index fec3b9b22309..e7cabf12c8dc 100644 --- a/drivers/input/touchscreen/ads7846.c +++ b/drivers/input/touchscreen/ads7846.c @@ -102,6 +102,8 @@ struct ads7846 { // FIXME remove "irq_disabled" unsigned irq_disabled:1; /* P: lock */ unsigned disabled:1; + + int (*get_pendown_state)(void); }; /* leave chip selected when we're done, for quicker re-select? */ @@ -175,6 +177,12 @@ struct ser_req { static void ads7846_enable(struct ads7846 *ts); static void ads7846_disable(struct ads7846 *ts); +static int device_suspended(struct device *dev) +{ + struct ads7846 *ts = dev_get_drvdata(dev); + return dev->power.power_state.event != PM_EVENT_ON || ts->disabled; +} + static int ads7846_read12_ser(struct device *dev, unsigned command) { struct spi_device *spi = to_spi_device(dev); @@ -227,8 +235,10 @@ static int ads7846_read12_ser(struct device *dev, unsigned command) for (i = 0; i < 6; i++) spi_message_add_tail(&req->xfer[i], &req->msg); + ts->irq_disabled = 1; disable_irq(spi->irq); status = spi_sync(spi, &req->msg); + ts->irq_disabled = 0; enable_irq(spi->irq); if (req->msg.status) @@ -333,7 +343,7 @@ static void ads7846_rx(void *ads) if (x == MAX_12BIT) x = 0; - if (x && z1 && ts->spi->dev.power.power_state.event == PM_EVENT_ON) { + if (likely(x && z1 && !device_suspended(&ts->spi->dev))) { /* compute touch pressure resistance using equation #2 */ Rt = z2; Rt -= z1; @@ -377,20 +387,10 @@ static void ads7846_rx(void *ads) x, y, Rt, Rt ? "" : " UP"); #endif - /* don't retrigger while we're suspended */ spin_lock_irqsave(&ts->lock, flags); ts->pendown = (Rt != 0); - ts->pending = 0; - - if (ts->spi->dev.power.power_state.event == PM_EVENT_ON) { - if (ts->pendown) - mod_timer(&ts->timer, jiffies + TS_POLL_PERIOD); - else if (ts->irq_disabled) { - ts->irq_disabled = 0; - enable_irq(ts->spi->irq); - } - } + mod_timer(&ts->timer, jiffies + TS_POLL_PERIOD); spin_unlock_irqrestore(&ts->lock, flags); } @@ -431,10 +431,25 @@ static void ads7846_timer(unsigned long handle) struct ads7846 *ts = (void *)handle; int status = 0; - ts->msg_idx = 0; - status = spi_async(ts->spi, &ts->msg[0]); - if (status) - dev_err(&ts->spi->dev, "spi_async --> %d\n", status); + spin_lock_irq(&ts->lock); + + if (unlikely(ts->msg_idx && !ts->pendown)) { + /* measurment cycle ended */ + if (!device_suspended(&ts->spi->dev)) { + ts->irq_disabled = 0; + enable_irq(ts->spi->irq); + } + ts->pending = 0; + ts->msg_idx = 0; + } else { + /* pen is still down, continue with the measurement */ + ts->msg_idx = 0; + status = spi_async(ts->spi, &ts->msg[0]); + if (status) + dev_err(&ts->spi->dev, "spi_async --> %d\n", status); + } + + spin_unlock_irq(&ts->lock); } static irqreturn_t ads7846_irq(int irq, void *handle, struct pt_regs *regs) @@ -443,7 +458,7 @@ static irqreturn_t ads7846_irq(int irq, void *handle, struct pt_regs *regs) unsigned long flags; spin_lock_irqsave(&ts->lock, flags); - if (likely(!ts->irq_disabled && !ts->disabled)) { + if (likely(ts->get_pendown_state())) { if (!ts->irq_disabled) { /* REVISIT irq logic for many ARM chips has cloned a * bug wherein disabling an irq in its handler won't @@ -452,10 +467,7 @@ static irqreturn_t ads7846_irq(int irq, void *handle, struct pt_regs *regs) * that state here. */ ts->irq_disabled = 1; - disable_irq(ts->spi->irq); - } - if (!ts->pending) { ts->pending = 1; mod_timer(&ts->timer, jiffies); } @@ -473,20 +485,17 @@ static void ads7846_disable(struct ads7846 *ts) if (ts->disabled) return; + ts->disabled = 1; + /* are we waiting for IRQ, or polling? */ - if (!ts->pendown) { - if (!ts->irq_disabled) { - ts->irq_disabled = 1; - disable_irq(ts->spi->irq); - } + if (!ts->pending) { + ts->irq_disabled = 1; + disable_irq(ts->spi->irq); } else { - /* polling; force a final SPI completion; - * that will clean things up neatly + /* the timer will run at least once more, and + * leave everything in a clean state, IRQ disabled */ - if (!ts->pending) - mod_timer(&ts->timer, jiffies); - - while (ts->pendown || ts->pending) { + while (ts->pending) { spin_unlock_irq(&ts->lock); msleep(1); spin_lock_irq(&ts->lock); @@ -497,7 +506,6 @@ static void ads7846_disable(struct ads7846 *ts) * leave it that way after every request */ - ts->disabled = 1; } /* Must be called with ts->lock held */ @@ -566,6 +574,11 @@ static int __devinit ads7846_probe(struct spi_device *spi) return -EINVAL; } + if (pdata->get_pendown_state == NULL) { + dev_dbg(&spi->dev, "no get_pendown_state function?\n"); + return -EINVAL; + } + /* We'd set the wordsize to 12 bits ... except that some controllers * will then treat the 8 bit command words as 12 bits (and drop the * four MSBs of the 12 bit result). Result: inputs must be shifted @@ -596,6 +609,7 @@ static int __devinit ads7846_probe(struct spi_device *spi) ts->x_plate_ohms = pdata->x_plate_ohms ? : 400; ts->debounce_max = pdata->debounce_max ? : 1; ts->debounce_tol = pdata->debounce_tol ? : 10; + ts->get_pendown_state = pdata->get_pendown_state; snprintf(ts->phys, sizeof(ts->phys), "%s/input0", spi->dev.bus_id); @@ -786,8 +800,8 @@ static int __devexit ads7846_remove(struct spi_device *spi) device_remove_file(&spi->dev, &dev_attr_vaux); free_irq(ts->spi->irq, ts); - if (ts->irq_disabled) - enable_irq(ts->spi->irq); + /* suspend left the IRQ disabled */ + enable_irq(ts->spi->irq); kfree(ts); diff --git a/include/linux/spi/ads7846.h b/include/linux/spi/ads7846.h index 3f7664951256..d8823e237e0b 100644 --- a/include/linux/spi/ads7846.h +++ b/include/linux/spi/ads7846.h @@ -17,5 +17,7 @@ struct ads7846_platform_data { u16 debounce_max; /* max number of readings per sample */ u16 debounce_tol; /* tolerance used for filtering */ + + int (*get_pendown_state)(void); }; -- cgit v1.2.3 From 308a878210cde6ab19df9f392c24db53ad6f56bf Mon Sep 17 00:00:00 2001 From: Russ Anderson Date: Tue, 18 Apr 2006 11:26:34 -0500 Subject: [IA64] Remove unused variable in sn_sal.h cnodeid was being set but not used. The dead code was left over from a previous version that grabbed a per node lock. Signed-off-by: Russ Anderson (rja@sgi.com) Signed-off-by: Tony Luck --- include/asm-ia64/sn/sn_sal.h | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'include') diff --git a/include/asm-ia64/sn/sn_sal.h b/include/asm-ia64/sn/sn_sal.h index bf4cc867a698..7ddce80e1918 100644 --- a/include/asm-ia64/sn/sn_sal.h +++ b/include/asm-ia64/sn/sn_sal.h @@ -8,7 +8,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2000-2005 Silicon Graphics, Inc. All rights reserved. + * Copyright (c) 2000-2006 Silicon Graphics, Inc. All rights reserved. */ @@ -705,10 +705,8 @@ static inline int sn_change_memprotect(u64 paddr, u64 len, u64 perms, u64 *nasid_array) { struct ia64_sal_retval ret_stuff; - int cnodeid; unsigned long irq_flags; - cnodeid = nasid_to_cnodeid(get_node_number(paddr)); local_irq_save(irq_flags); ia64_sal_oemcall_nolock(&ret_stuff, SN_SAL_MEMPROTECT, paddr, len, (u64)nasid_array, perms, 0, 0, 0); -- cgit v1.2.3 From 0d9adec525b87d8ab7e64efeabffb5b3f293056e Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Tue, 18 Apr 2006 15:00:45 -0500 Subject: [IA64] - Fix MAX_PXM_DOMAINS for systems with > 256 nodes Correctly size the PXM-related arrays for systems that have more than 256 nodes. Signed-off-by: Jack Steiner Signed-off-by: Tony Luck --- include/asm-ia64/acpi.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/asm-ia64/acpi.h b/include/asm-ia64/acpi.h index d734585a23cf..09a5dd0e44a8 100644 --- a/include/asm-ia64/acpi.h +++ b/include/asm-ia64/acpi.h @@ -110,9 +110,8 @@ extern void prefill_possible_map(void); extern int additional_cpus; #ifdef CONFIG_ACPI_NUMA -/* Proximity bitmap length; _PXM is at most 255 (8 bit)*/ -#ifdef CONFIG_IA64_NR_NODES -#define MAX_PXM_DOMAINS CONFIG_IA64_NR_NODES +#if MAX_NUMNODES > 256 +#define MAX_PXM_DOMAINS MAX_NUMNODES #else #define MAX_PXM_DOMAINS (256) #endif -- cgit v1.2.3 From 86db2f4239e2556cd37b853c2307aa9d43041458 Mon Sep 17 00:00:00 2001 From: Russ Anderson Date: Thu, 20 Apr 2006 17:05:43 -0700 Subject: [IA64-SGI] SN SAL call to inject memory errors The SGI Altix SAL provides an interface for modifying the ECC on memory to create memory errors. The SAL call can be used to inject memory errors for testing MCA recovery code. Signed-off-by: Russ Anderson (rja@sgi.com) Signed-off-by: Tony Luck --- include/asm-ia64/sn/sn_sal.h | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'include') diff --git a/include/asm-ia64/sn/sn_sal.h b/include/asm-ia64/sn/sn_sal.h index 7ddce80e1918..51aca022cf39 100644 --- a/include/asm-ia64/sn/sn_sal.h +++ b/include/asm-ia64/sn/sn_sal.h @@ -85,6 +85,7 @@ #define SN_SAL_GET_PROM_FEATURE_SET 0x02000065 #define SN_SAL_SET_OS_FEATURE_SET 0x02000066 +#define SN_SAL_INJECT_ERROR 0x02000067 /* * Service-specific constants @@ -1138,4 +1139,16 @@ ia64_sn_set_os_feature(int feature) return rv.status; } +static inline int +sn_inject_error(u64 paddr, u64 *data, u64 *ecc) +{ + struct ia64_sal_retval ret_stuff; + unsigned long irq_flags; + + local_irq_save(irq_flags); + ia64_sal_oemcall_nolock(&ret_stuff, SN_SAL_INJECT_ERROR, paddr, (u64)data, + (u64)ecc, 0, 0, 0, 0); + local_irq_restore(irq_flags); + return ret_stuff.status; +} #endif /* _ASM_IA64_SN_SN_SAL_H */ -- cgit v1.2.3 From a72391e42f0a13116995045b3d492d660f96697d Mon Sep 17 00:00:00 2001 From: Satoru Takeuchi Date: Thu, 20 Apr 2006 18:49:48 +0900 Subject: [IA64] eliminate compile time warnings This patch removes following compile time warnings: drivers/pci/pci-sysfs.c: In function `pci_read_legacy_io': drivers/pci/pci-sysfs.c:257: warning: implicit declaration of function `ia64_pci_legacy_read' drivers/pci/pci-sysfs.c: In function `pci_write_legacy_io': drivers/pci/pci-sysfs.c:280: warning: implicit declaration of function `ia64_pci_legacy_write' It also fixes wrong definition of ia64_pci_legacy_write (type of `bus' is not `pci_dev', but `pci_bus'). Signed-Off-By: Satoru Takeuchi Signed-off-by: Tony Luck --- arch/ia64/pci/pci.c | 3 +-- include/asm-ia64/machvec.h | 2 ++ 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c index 9ba32b2d96d0..ab829a22f8a4 100644 --- a/arch/ia64/pci/pci.c +++ b/arch/ia64/pci/pci.c @@ -31,7 +31,6 @@ #include #include - /* * Low-level SAL-based PCI configuration access functions. Note that SAL * calls are already serialized (via sal_lock), so we don't need another @@ -707,7 +706,7 @@ int ia64_pci_legacy_read(struct pci_bus *bus, u16 port, u32 *val, u8 size) * * Simply writes @size bytes of @val to @port. */ -int ia64_pci_legacy_write(struct pci_dev *bus, u16 port, u32 val, u8 size) +int ia64_pci_legacy_write(struct pci_bus *bus, u16 port, u32 val, u8 size) { int ret = size; diff --git a/include/asm-ia64/machvec.h b/include/asm-ia64/machvec.h index c3e4ed8a3e17..a9c995a86c21 100644 --- a/include/asm-ia64/machvec.h +++ b/include/asm-ia64/machvec.h @@ -347,9 +347,11 @@ extern ia64_mv_dma_supported swiotlb_dma_supported; #endif #ifndef platform_pci_legacy_read # define platform_pci_legacy_read ia64_pci_legacy_read +extern int ia64_pci_legacy_read(struct pci_bus *bus, u16 port, u32 *val, u8 size); #endif #ifndef platform_pci_legacy_write # define platform_pci_legacy_write ia64_pci_legacy_write +extern int ia64_pci_legacy_write(struct pci_bus *bus, u16 port, u32 val, u8 size); #endif #ifndef platform_inb # define platform_inb __ia64_inb -- cgit v1.2.3 From e5ecc192dfc5e0b325dd8c99ce4c755714c9acbf Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 13 Apr 2006 18:23:53 -0700 Subject: [IA64] Setup an IA64 specific reclaim distance RECLAIM_DISTANCE is checked on bootup against the SLIT table distances. Zone reclaim is important for system that have higher latencies but not for systems that have multiple nodes on one motherboard and therefore low latencies. We found that on motherboard latencies are typically 1 to 1.4 of local memory access speed whereas multinode systems which benefit from zone reclaim have usually more than 1.5 times the latency of a local access. Set the reclaim distance for IA64 to 1.5 times. Signed-off-by: Christoph Lameter Signed-off-by: Tony Luck --- include/asm-ia64/topology.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/asm-ia64/topology.h b/include/asm-ia64/topology.h index 3ee19dfa46df..616b5ed2aa72 100644 --- a/include/asm-ia64/topology.h +++ b/include/asm-ia64/topology.h @@ -22,6 +22,11 @@ /* Nodes w/o CPUs are preferred for memory allocations, see build_zonelists */ #define PENALTY_FOR_NODE_WITH_CPUS 255 +/* + * Distance above which we begin to use zone reclaim + */ +#define RECLAIM_DISTANCE 15 + /* * Returns the number of the node containing CPU 'cpu' */ -- cgit v1.2.3 From d5b415c95f0e6510451f1446cea832c1f77bd7ea Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Wed, 26 Apr 2006 00:13:18 -0400 Subject: Input: ads7846 - improve filtering for thumb press accuracy Providing more accurate coordinates for thumb press requires additional steps in the filtering logic: - Ignore samples found invalid by the debouncing logic, or the ones that have out of bound pressure value. - Add a parameter to repeat debouncing, so that more then two consecutive good readings are required for a valid sample. Signed-off-by: Imre Deak Acked-by: Juha Yrjola Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/ads7846.c | 72 +++++++++++++++++++++++++++++-------- include/linux/spi/ads7846.h | 6 ++-- 2 files changed, 61 insertions(+), 17 deletions(-) (limited to 'include') diff --git a/drivers/input/touchscreen/ads7846.c b/drivers/input/touchscreen/ads7846.c index 1aaa153a2774..1494175ac6fe 100644 --- a/drivers/input/touchscreen/ads7846.c +++ b/drivers/input/touchscreen/ads7846.c @@ -71,6 +71,7 @@ struct ts_event { __be16 x; __be16 y; __be16 z1, z2; + int ignore; }; struct ads7846 { @@ -81,6 +82,7 @@ struct ads7846 { u16 model; u16 vref_delay_usecs; u16 x_plate_ohms; + u16 pressure_max; u8 read_x, read_y, read_z1, read_z2, pwrdown; u16 dummy; /* for the pwrdown read */ @@ -88,12 +90,15 @@ struct ads7846 { struct spi_transfer xfer[10]; struct spi_message msg[5]; + struct spi_message *last_msg; int msg_idx; int read_cnt; + int read_rep; int last_read; u16 debounce_max; u16 debounce_tol; + u16 debounce_rep; spinlock_t lock; struct timer_list timer; /* P: lock */ @@ -354,6 +359,14 @@ static void ads7846_rx(void *ads) } else Rt = 0; + /* Sample found inconsistent by debouncing or pressure is beyond + * the maximum. Don't report it to user space, repeat at least + * once more the measurement */ + if (ts->tc.ignore || Rt > ts->pressure_max) { + mod_timer(&ts->timer, jiffies + TS_POLL_PERIOD); + return; + } + /* NOTE: "pendown" is inferred from pressure; we don't rely on * being able to check nPENIRQ status, or "friendly" trigger modes * (both-edges is much better than just-falling or low-level). @@ -402,25 +415,45 @@ static void ads7846_debounce(void *ads) struct ads7846 *ts = ads; struct spi_message *m; struct spi_transfer *t; - u16 val; + int val; int status; m = &ts->msg[ts->msg_idx]; t = list_entry(m->transfers.prev, struct spi_transfer, transfer_list); val = (*(u16 *)t->rx_buf) >> 3; - - if (!ts->read_cnt || (abs(ts->last_read - val) > ts->debounce_tol - && ts->read_cnt < ts->debounce_max)) { - /* Repeat it, if this was the first read or the read wasn't - * consistent enough - */ - ts->read_cnt++; - ts->last_read = val; + if (!ts->read_cnt || (abs(ts->last_read - val) > ts->debounce_tol)) { + /* Repeat it, if this was the first read or the read + * wasn't consistent enough. */ + if (ts->read_cnt < ts->debounce_max) { + ts->last_read = val; + ts->read_cnt++; + } else { + /* Maximum number of debouncing reached and still + * not enough number of consistent readings. Abort + * the whole sample, repeat it in the next sampling + * period. + */ + ts->tc.ignore = 1; + ts->read_cnt = 0; + /* Last message will contain ads7846_rx() as the + * completion function. + */ + m = ts->last_msg; + } + /* Start over collecting consistent readings. */ + ts->read_rep = 0; } else { - /* Go for the next read */ - ts->msg_idx++; - ts->read_cnt = 0; - m++; + if (++ts->read_rep > ts->debounce_rep) { + /* Got a good reading for this coordinate, + * go for the next one. */ + ts->tc.ignore = 0; + ts->msg_idx++; + ts->read_cnt = 0; + ts->read_rep = 0; + m++; + } else + /* Read more values that are consistent. */ + ts->read_cnt++; } status = spi_async(ts->spi, m); if (status) @@ -609,8 +642,15 @@ static int __devinit ads7846_probe(struct spi_device *spi) ts->model = pdata->model ? : 7846; ts->vref_delay_usecs = pdata->vref_delay_usecs ? : 100; ts->x_plate_ohms = pdata->x_plate_ohms ? : 400; - ts->debounce_max = pdata->debounce_max ? : 1; - ts->debounce_tol = pdata->debounce_tol ? : 10; + ts->pressure_max = pdata->pressure_max ? : ~0; + if (pdata->debounce_max) { + ts->debounce_max = pdata->debounce_max; + ts->debounce_tol = pdata->debounce_tol; + ts->debounce_rep = pdata->debounce_rep; + if (ts->debounce_rep > ts->debounce_max + 1) + ts->debounce_rep = ts->debounce_max - 1; + } else + ts->debounce_tol = ~0; ts->get_pendown_state = pdata->get_pendown_state; snprintf(ts->phys, sizeof(ts->phys), "%s/input0", spi->dev.bus_id); @@ -728,6 +768,8 @@ static int __devinit ads7846_probe(struct spi_device *spi) m->complete = ads7846_rx; m->context = ts; + ts->last_msg = m; + if (request_irq(spi->irq, ads7846_irq, SA_SAMPLE_RANDOM | SA_TRIGGER_FALLING, spi->dev.bus_id, ts)) { diff --git a/include/linux/spi/ads7846.h b/include/linux/spi/ads7846.h index d8823e237e0b..adb3dafd33e9 100644 --- a/include/linux/spi/ads7846.h +++ b/include/linux/spi/ads7846.h @@ -15,9 +15,11 @@ struct ads7846_platform_data { u16 y_min, y_max; u16 pressure_min, pressure_max; - u16 debounce_max; /* max number of readings per sample */ + u16 debounce_max; /* max number of additional readings + * per sample */ u16 debounce_tol; /* tolerance used for filtering */ - + u16 debounce_rep; /* additional consecutive good readings + * required after the first two */ int (*get_pendown_state)(void); }; -- cgit v1.2.3 From 1a0ccece05efb8a9c04b1130c24a2652239f3bea Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 26 Apr 2006 00:13:57 -0400 Subject: Input: allow passing NULL to input_free_device() Many drivers rely on input_free_device() behaving like kfree(). Signed-off-by: Dmitry Torokhov --- include/linux/input.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/input.h b/include/linux/input.h index b0e612dda0cf..16c19d710a4d 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -1016,7 +1016,8 @@ static inline void input_put_device(struct input_dev *dev) static inline void input_free_device(struct input_dev *dev) { - input_put_device(dev); + if (dev) + input_put_device(dev); } int input_register_device(struct input_dev *); -- cgit v1.2.3 From ddc5d3414593e4d7ad7fbd33e7f7517fcc234544 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 26 Apr 2006 00:14:19 -0400 Subject: Input: move input_device_id to mod_devicetable.h Signed-off-by: Dmitry Torokhov --- drivers/input/input.c | 8 ++-- include/linux/input.h | 104 ++++++++++++++++++++-------------------- include/linux/mod_devicetable.h | 48 +++++++++++++++++++ scripts/mod/file2alias.c | 36 +++++++------- 4 files changed, 122 insertions(+), 74 deletions(-) (limited to 'include') diff --git a/drivers/input/input.c b/drivers/input/input.c index a935abeffffc..591c70d80cd8 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -286,19 +286,19 @@ static struct input_device_id *input_match_device(struct input_device_id *id, st for (; id->flags || id->driver_info; id++) { if (id->flags & INPUT_DEVICE_ID_MATCH_BUS) - if (id->id.bustype != dev->id.bustype) + if (id->bustype != dev->id.bustype) continue; if (id->flags & INPUT_DEVICE_ID_MATCH_VENDOR) - if (id->id.vendor != dev->id.vendor) + if (id->vendor != dev->id.vendor) continue; if (id->flags & INPUT_DEVICE_ID_MATCH_PRODUCT) - if (id->id.product != dev->id.product) + if (id->product != dev->id.product) continue; if (id->flags & INPUT_DEVICE_ID_MATCH_VERSION) - if (id->id.version != dev->id.version) + if (id->version != dev->id.version) continue; MATCH_BIT(evbit, EV_MAX); diff --git a/include/linux/input.h b/include/linux/input.h index 16c19d710a4d..8298b4bf5a07 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -12,8 +12,6 @@ #ifdef __KERNEL__ #include #include -#include -#include #else #include #include @@ -577,15 +575,15 @@ struct input_absinfo { * Switch events */ -#define SW_0 0x00 -#define SW_1 0x01 -#define SW_2 0x02 -#define SW_3 0x03 -#define SW_4 0x04 -#define SW_5 0x05 -#define SW_6 0x06 -#define SW_7 0x07 -#define SW_MAX 0x0f +#define SW_0 0x00 +#define SW_1 0x01 +#define SW_2 0x02 +#define SW_3 0x03 +#define SW_4 0x04 +#define SW_5 0x05 +#define SW_6 0x06 +#define SW_7 0x07 +#define SW_MAX 0x0f /* * Misc events @@ -805,52 +803,16 @@ struct ff_effect { #define FF_MAX 0x7f -struct input_device_id { - - kernel_ulong_t flags; - - struct input_id id; - - kernel_ulong_t evbit[EV_MAX/BITS_PER_LONG+1]; - kernel_ulong_t keybit[KEY_MAX/BITS_PER_LONG+1]; - kernel_ulong_t relbit[REL_MAX/BITS_PER_LONG+1]; - kernel_ulong_t absbit[ABS_MAX/BITS_PER_LONG+1]; - kernel_ulong_t mscbit[MSC_MAX/BITS_PER_LONG+1]; - kernel_ulong_t ledbit[LED_MAX/BITS_PER_LONG+1]; - kernel_ulong_t sndbit[SND_MAX/BITS_PER_LONG+1]; - kernel_ulong_t ffbit[FF_MAX/BITS_PER_LONG+1]; - kernel_ulong_t swbit[SW_MAX/BITS_PER_LONG+1]; - - kernel_ulong_t driver_info; -}; - -/* - * Structure for hotplug & device<->driver matching. - */ - -#define INPUT_DEVICE_ID_MATCH_BUS 1 -#define INPUT_DEVICE_ID_MATCH_VENDOR 2 -#define INPUT_DEVICE_ID_MATCH_PRODUCT 4 -#define INPUT_DEVICE_ID_MATCH_VERSION 8 - -#define INPUT_DEVICE_ID_MATCH_EVBIT 0x010 -#define INPUT_DEVICE_ID_MATCH_KEYBIT 0x020 -#define INPUT_DEVICE_ID_MATCH_RELBIT 0x040 -#define INPUT_DEVICE_ID_MATCH_ABSBIT 0x080 -#define INPUT_DEVICE_ID_MATCH_MSCIT 0x100 -#define INPUT_DEVICE_ID_MATCH_LEDBIT 0x200 -#define INPUT_DEVICE_ID_MATCH_SNDBIT 0x400 -#define INPUT_DEVICE_ID_MATCH_FFBIT 0x800 -#define INPUT_DEVICE_ID_MATCH_SWBIT 0x1000 - #ifdef __KERNEL__ /* * In-kernel definitions. */ +#include #include #include +#include #define NBITS(x) (((x)/BITS_PER_LONG)+1) #define BIT(x) (1UL<<((x)%BITS_PER_LONG)) @@ -951,9 +913,49 @@ struct input_dev { }; #define to_input_dev(d) container_of(d, struct input_dev, cdev) -#define INPUT_DEVICE_ID_MATCH_DEVICE\ +/* + * Verify that we are in sync with input_device_id mod_devicetable.h #defines + */ + +#if EV_MAX != INPUT_DEVICE_ID_EV_MAX +#error "EV_MAX and INPUT_DEVICE_ID_EV_MAX do not match" +#endif + +#if KEY_MAX != INPUT_DEVICE_ID_KEY_MAX +#error "KEY_MAX and INPUT_DEVICE_ID_KEY_MAX do not match" +#endif + +#if REL_MAX != INPUT_DEVICE_ID_REL_MAX +#error "REL_MAX and INPUT_DEVICE_ID_REL_MAX do not match" +#endif + +#if ABS_MAX != INPUT_DEVICE_ID_ABS_MAX +#error "ABS_MAX and INPUT_DEVICE_ID_ABS_MAX do not match" +#endif + +#if MSC_MAX != INPUT_DEVICE_ID_MSC_MAX +#error "MSC_MAX and INPUT_DEVICE_ID_MSC_MAX do not match" +#endif + +#if LED_MAX != INPUT_DEVICE_ID_LED_MAX +#error "LED_MAX and INPUT_DEVICE_ID_LED_MAX do not match" +#endif + +#if SND_MAX != INPUT_DEVICE_ID_SND_MAX +#error "SND_MAX and INPUT_DEVICE_ID_SND_MAX do not match" +#endif + +#if FF_MAX != INPUT_DEVICE_ID_FF_MAX +#error "FF_MAX and INPUT_DEVICE_ID_FF_MAX do not match" +#endif + +#if SW_MAX != INPUT_DEVICE_ID_SW_MAX +#error "SW_MAX and INPUT_DEVICE_ID_SW_MAX do not match" +#endif + +#define INPUT_DEVICE_ID_MATCH_DEVICE \ (INPUT_DEVICE_ID_MATCH_BUS | INPUT_DEVICE_ID_MATCH_VENDOR | INPUT_DEVICE_ID_MATCH_PRODUCT) -#define INPUT_DEVICE_ID_MATCH_DEVICE_AND_VERSION\ +#define INPUT_DEVICE_ID_MATCH_DEVICE_AND_VERSION \ (INPUT_DEVICE_ID_MATCH_DEVICE | INPUT_DEVICE_ID_MATCH_VERSION) struct input_handle; diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h index 7b08c11ec4cc..f6977708585c 100644 --- a/include/linux/mod_devicetable.h +++ b/include/linux/mod_devicetable.h @@ -249,4 +249,52 @@ struct i2c_device_id { __u16 id; }; +/* Input */ +#define INPUT_DEVICE_ID_EV_MAX 0x1f +#define INPUT_DEVICE_ID_KEY_MAX 0x1ff +#define INPUT_DEVICE_ID_REL_MAX 0x0f +#define INPUT_DEVICE_ID_ABS_MAX 0x3f +#define INPUT_DEVICE_ID_MSC_MAX 0x07 +#define INPUT_DEVICE_ID_LED_MAX 0x0f +#define INPUT_DEVICE_ID_SND_MAX 0x07 +#define INPUT_DEVICE_ID_FF_MAX 0x7f +#define INPUT_DEVICE_ID_SW_MAX 0x0f + +#define INPUT_DEVICE_ID_MATCH_BUS 1 +#define INPUT_DEVICE_ID_MATCH_VENDOR 2 +#define INPUT_DEVICE_ID_MATCH_PRODUCT 4 +#define INPUT_DEVICE_ID_MATCH_VERSION 8 + +#define INPUT_DEVICE_ID_MATCH_EVBIT 0x0010 +#define INPUT_DEVICE_ID_MATCH_KEYBIT 0x0020 +#define INPUT_DEVICE_ID_MATCH_RELBIT 0x0040 +#define INPUT_DEVICE_ID_MATCH_ABSBIT 0x0080 +#define INPUT_DEVICE_ID_MATCH_MSCIT 0x0100 +#define INPUT_DEVICE_ID_MATCH_LEDBIT 0x0200 +#define INPUT_DEVICE_ID_MATCH_SNDBIT 0x0400 +#define INPUT_DEVICE_ID_MATCH_FFBIT 0x0800 +#define INPUT_DEVICE_ID_MATCH_SWBIT 0x1000 + +struct input_device_id { + + kernel_ulong_t flags; + + __u16 bustype; + __u16 vendor; + __u16 product; + __u16 version; + + kernel_ulong_t evbit[INPUT_DEVICE_ID_EV_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t keybit[INPUT_DEVICE_ID_KEY_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t relbit[INPUT_DEVICE_ID_REL_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t absbit[INPUT_DEVICE_ID_ABS_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t mscbit[INPUT_DEVICE_ID_MSC_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t ledbit[INPUT_DEVICE_ID_LED_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t sndbit[INPUT_DEVICE_ID_SND_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t ffbit[INPUT_DEVICE_ID_FF_MAX / BITS_PER_LONG + 1]; + kernel_ulong_t swbit[INPUT_DEVICE_ID_SW_MAX / BITS_PER_LONG + 1]; + + kernel_ulong_t driver_info; +}; + #endif /* LINUX_MOD_DEVICETABLE_H */ diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c index 84e21201f3c0..37f67c23e11b 100644 --- a/scripts/mod/file2alias.c +++ b/scripts/mod/file2alias.c @@ -374,10 +374,10 @@ static void do_input(char *alias, kernel_ulong_t *arr, unsigned int min, unsigned int max) { unsigned int i; - for (i = min; i < max; i++) { - if (arr[i/BITS_PER_LONG] & (1 << (i%BITS_PER_LONG))) - sprintf(alias+strlen(alias), "%X,*", i); - } + + for (i = min; i < max; i++) + if (arr[i / BITS_PER_LONG] & (1 << (i%BITS_PER_LONG))) + sprintf(alias + strlen(alias), "%X,*", i); } /* input:b0v0p0e0-eXkXrXaXmXlXsXfXwX where X is comma-separated %02X. */ @@ -386,39 +386,37 @@ static int do_input_entry(const char *filename, struct input_device_id *id, { sprintf(alias, "input:"); - ADD(alias, "b", id->flags&INPUT_DEVICE_ID_MATCH_BUS, id->id.bustype); - ADD(alias, "v", id->flags&INPUT_DEVICE_ID_MATCH_VENDOR, id->id.vendor); - ADD(alias, "p", id->flags&INPUT_DEVICE_ID_MATCH_PRODUCT, - id->id.product); - ADD(alias, "e", id->flags&INPUT_DEVICE_ID_MATCH_VERSION, - id->id.version); + ADD(alias, "b", id->flags & INPUT_DEVICE_ID_MATCH_BUS, id->bustype); + ADD(alias, "v", id->flags & INPUT_DEVICE_ID_MATCH_VENDOR, id->vendor); + ADD(alias, "p", id->flags & INPUT_DEVICE_ID_MATCH_PRODUCT, id->product); + ADD(alias, "e", id->flags & INPUT_DEVICE_ID_MATCH_VERSION, id->version); sprintf(alias + strlen(alias), "-e*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_EVBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_EVBIT) do_input(alias, id->evbit, 0, EV_MAX); sprintf(alias + strlen(alias), "k*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_KEYBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_KEYBIT) do_input(alias, id->keybit, KEY_MIN_INTERESTING, KEY_MAX); sprintf(alias + strlen(alias), "r*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_RELBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_RELBIT) do_input(alias, id->relbit, 0, REL_MAX); sprintf(alias + strlen(alias), "a*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_ABSBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_ABSBIT) do_input(alias, id->absbit, 0, ABS_MAX); sprintf(alias + strlen(alias), "m*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_MSCIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_MSCIT) do_input(alias, id->mscbit, 0, MSC_MAX); sprintf(alias + strlen(alias), "l*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_LEDBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_LEDBIT) do_input(alias, id->ledbit, 0, LED_MAX); sprintf(alias + strlen(alias), "s*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_SNDBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_SNDBIT) do_input(alias, id->sndbit, 0, SND_MAX); sprintf(alias + strlen(alias), "f*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_FFBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_FFBIT) do_input(alias, id->ffbit, 0, FF_MAX); sprintf(alias + strlen(alias), "w*"); - if (id->flags&INPUT_DEVICE_ID_MATCH_SWBIT) + if (id->flags & INPUT_DEVICE_ID_MATCH_SWBIT) do_input(alias, id->swbit, 0, SW_MAX); return 1; } -- cgit v1.2.3 From ebf43500ef148a380bd132743c3fc530111ac620 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 27 Apr 2006 08:46:01 +0200 Subject: [PATCH] Add find_get_pages_contig(): contiguous variant of find_get_pages() find_get_pages_contig() will break out if we hit a hole in the page cache. From Andrew Morton, small modifications and documentation by me. Signed-off-by: Jens Axboe --- include/linux/pagemap.h | 2 ++ mm/filemap.c | 32 ++++++++++++++++++++++++++++++++ 2 files changed, 34 insertions(+) (limited to 'include') diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index 9539efd4f7e6..7a1af574dedf 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -78,6 +78,8 @@ extern struct page * find_or_create_page(struct address_space *mapping, unsigned long index, gfp_t gfp_mask); unsigned find_get_pages(struct address_space *mapping, pgoff_t start, unsigned int nr_pages, struct page **pages); +unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t start, + unsigned int nr_pages, struct page **pages); unsigned find_get_pages_tag(struct address_space *mapping, pgoff_t *index, int tag, unsigned int nr_pages, struct page **pages); diff --git a/mm/filemap.c b/mm/filemap.c index 3ef20739e725..fd57442186cb 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -697,6 +697,38 @@ unsigned find_get_pages(struct address_space *mapping, pgoff_t start, return ret; } +/** + * find_get_pages_contig - gang contiguous pagecache lookup + * @mapping: The address_space to search + * @index: The starting page index + * @nr_pages: The maximum number of pages + * @pages: Where the resulting pages are placed + * + * find_get_pages_contig() works exactly like find_get_pages(), except + * that the returned number of pages are guaranteed to be contiguous. + * + * find_get_pages_contig() returns the number of pages which were found. + */ +unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t index, + unsigned int nr_pages, struct page **pages) +{ + unsigned int i; + unsigned int ret; + + read_lock_irq(&mapping->tree_lock); + ret = radix_tree_gang_lookup(&mapping->page_tree, + (void **)pages, index, nr_pages); + for (i = 0; i < ret; i++) { + if (pages[i]->mapping == NULL || pages[i]->index != index) + break; + + page_cache_get(pages[i]); + index++; + } + read_unlock_irq(&mapping->tree_lock); + return i; +} + /* * Like find_get_pages, except we only return pages which are tagged with * `tag'. We update *index to index the next page for the traversal. -- cgit v1.2.3 From bc818247203a7bfc40296a3f5b760de84fb8e0d1 Mon Sep 17 00:00:00 2001 From: Atsushi Nemoto Date: Mon, 17 Apr 2006 21:19:12 +0900 Subject: [MIPS] Fix bitops for MIPS32/MIPS64 CPUs. With recent rewrite for generic bitops, fls() for 32bit kernel with MIPS64_CPU is broken. Also, ffs(), fls() should be defined the same way as the libc and compiler built-in routines (returns int instead of unsigned long). Signed-off-by: Atsushi Nemoto Signed-off-by: Ralf Baechle --- include/asm-mips/bitops.h | 56 ++++++++++++++++++++--------------------------- 1 file changed, 24 insertions(+), 32 deletions(-) (limited to 'include') diff --git a/include/asm-mips/bitops.h b/include/asm-mips/bitops.h index a1728f8c0705..d2f444537e4b 100644 --- a/include/asm-mips/bitops.h +++ b/include/asm-mips/bitops.h @@ -467,64 +467,56 @@ static inline unsigned long __ffs(unsigned long word) } /* - * ffs - find first bit set. + * fls - find last bit set. * @word: The word to search * - * Returns 1..SZLONG - * Returns 0 if no bit exists + * This is defined the same way as ffs. + * Note fls(0) = 0, fls(1) = 1, fls(0x80000000) = 32. */ - -static inline unsigned long ffs(unsigned long word) +static inline int fls(int word) { - if (!word) - return 0; + __asm__ ("clz %0, %1" : "=r" (word) : "r" (word)); - return __ffs(word) + 1; + return 32 - word; } -/* - * ffz - find first zero in word. - * @word: The word to search - * - * Undefined if no zero exists, so code should check against ~0UL first. - */ -static inline unsigned long ffz(unsigned long word) +#if defined(CONFIG_64BIT) && defined(CONFIG_CPU_MIPS64) +static inline int fls64(__u64 word) { - return __ffs (~word); + __asm__ ("dclz %0, %1" : "=r" (word) : "r" (word)); + + return 64 - word; } +#else +#include +#endif /* - * fls - find last bit set. + * ffs - find first bit set. * @word: The word to search * - * Returns 1..SZLONG - * Returns 0 if no bit exists + * This is defined the same way as + * the libc and compiler builtin ffs routines, therefore + * differs in spirit from the above ffz (man ffs). */ -static inline unsigned long fls(unsigned long word) +static inline int ffs(int word) { -#ifdef CONFIG_CPU_MIPS32 - __asm__ ("clz %0, %1" : "=r" (word) : "r" (word)); - - return 32 - word; -#endif - -#ifdef CONFIG_CPU_MIPS64 - __asm__ ("dclz %0, %1" : "=r" (word) : "r" (word)); + if (!word) + return 0; - return 64 - word; -#endif + return fls(word & -word); } #else #include #include -#include #include +#include #endif /*defined(CONFIG_CPU_MIPS32) || defined(CONFIG_CPU_MIPS64) */ -#include +#include #include #ifdef __KERNEL__ -- cgit v1.2.3 From 7a8341969fe0df4a1fffa141435e742456270ffd Mon Sep 17 00:00:00 2001 From: Chris Dearman Date: Sat, 15 Apr 2006 00:31:16 +0100 Subject: [MIPS] 24K LV: Add core card id. Signed-off-by: Ralf Baechle --- arch/mips/mips-boards/generic/init.c | 1 + arch/mips/mips-boards/generic/pci.c | 1 + arch/mips/mips-boards/malta/malta_int.c | 3 +++ include/asm-mips/mips-boards/generic.h | 1 + 4 files changed, 6 insertions(+) (limited to 'include') diff --git a/arch/mips/mips-boards/generic/init.c b/arch/mips/mips-boards/generic/init.c index 17dfe6a8cab9..df4e94735604 100644 --- a/arch/mips/mips-boards/generic/init.c +++ b/arch/mips/mips-boards/generic/init.c @@ -337,6 +337,7 @@ void __init prom_init(void) case MIPS_REVISION_CORID_CORE_MSC: case MIPS_REVISION_CORID_CORE_FPGA2: case MIPS_REVISION_CORID_CORE_FPGA3: + case MIPS_REVISION_CORID_CORE_24K: case MIPS_REVISION_CORID_CORE_EMUL_MSC: _pcictrl_msc = (unsigned long)ioremap(MIPS_MSC01_PCI_REG_BASE, 0x2000); diff --git a/arch/mips/mips-boards/generic/pci.c b/arch/mips/mips-boards/generic/pci.c index 1f6f9df74ab2..9337f6c8873a 100644 --- a/arch/mips/mips-boards/generic/pci.c +++ b/arch/mips/mips-boards/generic/pci.c @@ -198,6 +198,7 @@ void __init mips_pcibios_init(void) case MIPS_REVISION_CORID_CORE_MSC: case MIPS_REVISION_CORID_CORE_FPGA2: case MIPS_REVISION_CORID_CORE_FPGA3: + case MIPS_REVISION_CORID_CORE_24K: case MIPS_REVISION_CORID_CORE_EMUL_MSC: /* Set up resource ranges from the controller's registers. */ MSC_READ(MSC01_PCI_SC2PMBASL, start); diff --git a/arch/mips/mips-boards/malta/malta_int.c b/arch/mips/mips-boards/malta/malta_int.c index 64db07d4dbe5..7cc0ba4f553a 100644 --- a/arch/mips/mips-boards/malta/malta_int.c +++ b/arch/mips/mips-boards/malta/malta_int.c @@ -57,6 +57,7 @@ static inline int mips_pcibios_iack(void) case MIPS_REVISION_CORID_CORE_MSC: case MIPS_REVISION_CORID_CORE_FPGA2: case MIPS_REVISION_CORID_CORE_FPGA3: + case MIPS_REVISION_CORID_CORE_24K: case MIPS_REVISION_CORID_CORE_EMUL_MSC: MSC_READ(MSC01_PCI_IACK, irq); irq &= 0xff; @@ -143,6 +144,7 @@ void corehi_irqdispatch(struct pt_regs *regs) case MIPS_REVISION_CORID_CORE_MSC: case MIPS_REVISION_CORID_CORE_FPGA2: case MIPS_REVISION_CORID_CORE_FPGA3: + case MIPS_REVISION_CORID_CORE_24K: case MIPS_REVISION_CORID_CORE_EMUL_MSC: ll_msc_irq(regs); break; @@ -309,6 +311,7 @@ void __init arch_init_irq(void) case MIPS_REVISION_CORID_CORE_MSC: case MIPS_REVISION_CORID_CORE_FPGA2: case MIPS_REVISION_CORID_CORE_FPGA3: + case MIPS_REVISION_CORID_CORE_24K: case MIPS_REVISION_CORID_CORE_EMUL_MSC: if (cpu_has_veic) init_msc_irqs (MSC01E_INT_BASE, msc_eicirqmap, msc_nr_eicirqs); diff --git a/include/asm-mips/mips-boards/generic.h b/include/asm-mips/mips-boards/generic.h index 25b6ffc26623..fa8b913cc3e0 100644 --- a/include/asm-mips/mips-boards/generic.h +++ b/include/asm-mips/mips-boards/generic.h @@ -67,6 +67,7 @@ #define MIPS_REVISION_CORID_CORE_FPGA2 7 #define MIPS_REVISION_CORID_CORE_FPGAR2 8 #define MIPS_REVISION_CORID_CORE_FPGA3 9 +#define MIPS_REVISION_CORID_CORE_24K 10 /**** Artificial corid defines ****/ /* -- cgit v1.2.3 From 6e5882cfa24e1456702e463f6920fc0ca3c3d2b8 Mon Sep 17 00:00:00 2001 From: Zachary Amsden Date: Thu, 27 Apr 2006 11:32:29 -0700 Subject: [PATCH] x86/PAE: Fix pte_clear for the >4GB RAM case Proposed fix for ptep_get_and_clear_full PAE bug. Pte_clear had the same bug, so use the same fix for both. Turns out pmd_clear had it as well, but pgds are not affected. The problem is rather intricate. Page table entries in PAE mode are 64-bits wide, but the only atomic 8-byte write operation available in 32-bit mode is cmpxchg8b, which is expensive (at least on P4), and thus avoided. But it can happen that the processor may prefetch entries into the TLB in the middle of an operation which clears a page table entry. So one must always clear the P-bit in the low word of the page table entry first when clearing it. Since the sequence *ptep = __pte(0) leaves the order of the write dependent on the compiler, it must be coded explicitly as a clear of the low word followed by a clear of the high word. Further, there must be a write memory barrier here to enforce proper ordering by the compiler (and, in the future, by the processor as well). On > 4GB memory machines, the implementation of pte_clear for PAE was clearly deficient, as it could leave virtual mappings of physical memory above 4GB aliased to memory below 4GB in the TLB. The implementation of ptep_get_and_clear_full has a similar bug, although not nearly as likely to occur, since the mappings being cleared are in the process of being destroyed, and should never be dereferenced again. But, as luck would have it, it is possible to trigger bugs even without ever dereferencing these bogus TLB mappings, even if the clear is followed fairly soon after with a TLB flush or invalidation. The problem is that memory above 4GB may now be aliased into the first 4GB of memory, and in fact, may hit a region of memory with non-memory semantics. These regions include AGP and PCI space. As such, these memory regions are not cached by the processor. This introduces the bug. The processor can speculate memory operations, including memory writes, as long as they are committed with the proper ordering. Speculating a memory write to a linear address that has a bogus TLB mapping is possible. Normally, the speculation is harmless. But for cached memory, it does leave the falsely speculated cacheline unmodified, but in a dirty state. This cache line will be eventually written back. If this cacheline happens to intersect a region of memory that is not protected by the cache coherency protocol, it can corrupt data in I/O memory, which is generally a very bad thing to do, and can cause total system failure or just plain undefined behavior. These bugs are extremely unlikely, but the severity is of such magnitude, and the fix so simple that I think fixing them immediately is justified. Also, they are nearly impossible to debug. Signed-off-by: Zachary Amsden Signed-off-by: Linus Torvalds --- include/asm-i386/pgtable-2level.h | 3 +++ include/asm-i386/pgtable-3level.h | 20 ++++++++++++++++++++ include/asm-i386/pgtable.h | 4 +--- 3 files changed, 24 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/asm-i386/pgtable-2level.h b/include/asm-i386/pgtable-2level.h index 27bde973abc7..2756d4b04c27 100644 --- a/include/asm-i386/pgtable-2level.h +++ b/include/asm-i386/pgtable-2level.h @@ -18,6 +18,9 @@ #define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval) #define set_pmd(pmdptr, pmdval) (*(pmdptr) = (pmdval)) +#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) +#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0) + #define ptep_get_and_clear(mm,addr,xp) __pte(xchg(&(xp)->pte_low, 0)) #define pte_same(a, b) ((a).pte_low == (b).pte_low) #define pte_page(x) pfn_to_page(pte_pfn(x)) diff --git a/include/asm-i386/pgtable-3level.h b/include/asm-i386/pgtable-3level.h index 36a5aa63cbbf..dccb1b3337ad 100644 --- a/include/asm-i386/pgtable-3level.h +++ b/include/asm-i386/pgtable-3level.h @@ -85,6 +85,26 @@ static inline void pud_clear (pud_t * pud) { } #define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \ pmd_index(address)) +/* + * For PTEs and PDEs, we must clear the P-bit first when clearing a page table + * entry, so clear the bottom half first and enforce ordering with a compiler + * barrier. + */ +static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) +{ + ptep->pte_low = 0; + smp_wmb(); + ptep->pte_high = 0; +} + +static inline void pmd_clear(pmd_t *pmd) +{ + u32 *tmp = (u32 *)pmd; + *tmp = 0; + smp_wmb(); + *(tmp + 1) = 0; +} + static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) { pte_t res; diff --git a/include/asm-i386/pgtable.h b/include/asm-i386/pgtable.h index ee056c41a9fb..672c3f76b9df 100644 --- a/include/asm-i386/pgtable.h +++ b/include/asm-i386/pgtable.h @@ -204,12 +204,10 @@ extern unsigned long long __PAGE_KERNEL, __PAGE_KERNEL_EXEC; extern unsigned long pg0[]; #define pte_present(x) ((x).pte_low & (_PAGE_PRESENT | _PAGE_PROTNONE)) -#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) /* To avoid harmful races, pmd_none(x) should check only the lower when PAE */ #define pmd_none(x) (!(unsigned long)pmd_val(x)) #define pmd_present(x) (pmd_val(x) & _PAGE_PRESENT) -#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0) #define pmd_bad(x) ((pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE) @@ -268,7 +266,7 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long pte_t pte; if (full) { pte = *ptep; - *ptep = __pte(0); + pte_clear(mm, addr, ptep); } else { pte = ptep_get_and_clear(mm, addr, ptep); } -- cgit v1.2.3 From b7d90a356a43f4609bd6290fc2e1ca4ef79d4458 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 25 Apr 2006 12:56:04 +0200 Subject: [ALSA] Fix Oops at rmmod with CONFIG_SND_VERBOSE_PROCFS=n Fixed Oops at rmmod with CONFIG_SND_VERBOSE_PROCFS=n. Add ifdef to struct fields for optimization and better compile checks. Signed-off-by: Takashi Iwai --- include/sound/pcm.h | 6 +++++- include/sound/pcm_oss.h | 2 ++ sound/core/Kconfig | 4 ++-- sound/core/oss/pcm_oss.c | 6 +++--- sound/core/pcm.c | 6 +++--- sound/core/pcm_memory.c | 8 +++++--- 6 files changed, 20 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/sound/pcm.h b/include/sound/pcm.h index df70e7592ab5..373425895faa 100644 --- a/include/sound/pcm.h +++ b/include/sound/pcm.h @@ -374,12 +374,14 @@ struct snd_pcm_substream { /* -- OSS things -- */ struct snd_pcm_oss_substream oss; #endif +#ifdef CONFIG_SND_VERBOSE_PROCFS struct snd_info_entry *proc_root; struct snd_info_entry *proc_info_entry; struct snd_info_entry *proc_hw_params_entry; struct snd_info_entry *proc_sw_params_entry; struct snd_info_entry *proc_status_entry; struct snd_info_entry *proc_prealloc_entry; +#endif /* misc flags */ unsigned int no_mmap_ctrl: 1; unsigned int hw_opened: 1; @@ -400,12 +402,14 @@ struct snd_pcm_str { struct snd_pcm_oss_stream oss; #endif struct snd_pcm_file *files; +#ifdef CONFIG_SND_VERBOSE_PROCFS struct snd_info_entry *proc_root; struct snd_info_entry *proc_info_entry; -#ifdef CONFIG_SND_DEBUG +#ifdef CONFIG_SND_PCM_XRUN_DEBUG unsigned int xrun_debug; /* 0 = disabled, 1 = verbose, 2 = stacktrace */ struct snd_info_entry *proc_xrun_debug_entry; #endif +#endif }; struct snd_pcm { diff --git a/include/sound/pcm_oss.h b/include/sound/pcm_oss.h index 39df2baca18a..c854647b6f3c 100644 --- a/include/sound/pcm_oss.h +++ b/include/sound/pcm_oss.h @@ -75,7 +75,9 @@ struct snd_pcm_oss_substream { struct snd_pcm_oss_stream { struct snd_pcm_oss_setup *setup_list; /* setup list */ struct mutex setup_mutex; +#ifdef CONFIG_SND_VERBOSE_PROCFS struct snd_info_entry *proc_entry; +#endif }; struct snd_pcm_oss { diff --git a/sound/core/Kconfig b/sound/core/Kconfig index f3a07fb2e690..4262a1c87731 100644 --- a/sound/core/Kconfig +++ b/sound/core/Kconfig @@ -142,7 +142,7 @@ config SND_SUPPORT_OLD_API config SND_VERBOSE_PROCFS bool "Verbose procfs contents" - depends on SND + depends on SND && PROC_FS default y help Say Y here to include code for verbose procfs contents (provides @@ -175,7 +175,7 @@ config SND_DEBUG_DETECT config SND_PCM_XRUN_DEBUG bool "Enable PCM ring buffer overrun/underrun debugging" default n - depends on SND_DEBUG + depends on SND_DEBUG && SND_VERBOSE_PROCFS help Say Y to enable the PCM ring buffer overrun/underrun debugging. It is usually not required, but if you have trouble with diff --git a/sound/core/oss/pcm_oss.c b/sound/core/oss/pcm_oss.c index a7567b891791..ac990bf0b48f 100644 --- a/sound/core/oss/pcm_oss.c +++ b/sound/core/oss/pcm_oss.c @@ -2214,7 +2214,7 @@ static int snd_pcm_oss_mmap(struct file *file, struct vm_area_struct *area) return 0; } -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_VERBOSE_PROCFS /* * /proc interface */ @@ -2368,10 +2368,10 @@ static void snd_pcm_oss_proc_done(struct snd_pcm *pcm) } } } -#else /* !CONFIG_PROC_FS */ +#else /* !CONFIG_SND_VERBOSE_PROCFS */ #define snd_pcm_oss_proc_init(pcm) #define snd_pcm_oss_proc_done(pcm) -#endif /* CONFIG_PROC_FS */ +#endif /* CONFIG_SND_VERBOSE_PROCFS */ /* * ENTRY functions diff --git a/sound/core/pcm.c b/sound/core/pcm.c index 48007a54c8b7..84b00038236d 100644 --- a/sound/core/pcm.c +++ b/sound/core/pcm.c @@ -142,7 +142,7 @@ static int snd_pcm_control_ioctl(struct snd_card *card, return -ENOIOCTLCMD; } -#if defined(CONFIG_PROC_FS) && defined(CONFIG_SND_VERBOSE_PROCFS) +#ifdef CONFIG_SND_VERBOSE_PROCFS #define STATE(v) [SNDRV_PCM_STATE_##v] = #v #define STREAM(v) [SNDRV_PCM_STREAM_##v] = #v @@ -599,12 +599,12 @@ static int snd_pcm_substream_proc_done(struct snd_pcm_substream *substream) } return 0; } -#else /* !CONFIG_PROC_FS */ +#else /* !CONFIG_SND_VERBOSE_PROCFS */ static inline int snd_pcm_stream_proc_init(struct snd_pcm_str *pstr) { return 0; } static inline int snd_pcm_stream_proc_done(struct snd_pcm_str *pstr) { return 0; } static inline int snd_pcm_substream_proc_init(struct snd_pcm_substream *substream) { return 0; } static inline int snd_pcm_substream_proc_done(struct snd_pcm_substream *substream) { return 0; } -#endif /* CONFIG_PROC_FS */ +#endif /* CONFIG_SND_VERBOSE_PROCFS */ /** * snd_pcm_new_stream - create a new PCM stream diff --git a/sound/core/pcm_memory.c b/sound/core/pcm_memory.c index a0119ae67dcd..428f8c169ee1 100644 --- a/sound/core/pcm_memory.c +++ b/sound/core/pcm_memory.c @@ -100,8 +100,10 @@ static void snd_pcm_lib_preallocate_dma_free(struct snd_pcm_substream *substream int snd_pcm_lib_preallocate_free(struct snd_pcm_substream *substream) { snd_pcm_lib_preallocate_dma_free(substream); +#ifdef CONFIG_SND_VERBOSE_PROCFS snd_info_unregister(substream->proc_prealloc_entry); substream->proc_prealloc_entry = NULL; +#endif return 0; } @@ -124,7 +126,7 @@ int snd_pcm_lib_preallocate_free_for_all(struct snd_pcm *pcm) return 0; } -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_VERBOSE_PROCFS /* * read callback for prealloc proc file * @@ -203,9 +205,9 @@ static inline void preallocate_info_init(struct snd_pcm_substream *substream) substream->proc_prealloc_entry = entry; } -#else /* !CONFIG_PROC_FS */ +#else /* !CONFIG_SND_VERBOSE_PROCFS */ #define preallocate_info_init(s) -#endif +#endif /* CONFIG_SND_VERBOSE_PROCFS */ /* * pre-allocate the buffer and create a proc file for the substream -- cgit v1.2.3 From 4d17ffda331ba6030bb8c233c73d6a87954d8ea7 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Tue, 25 Apr 2006 15:37:26 +0200 Subject: [PATCH] Kobject: fix build error This fixes a build error for various odd combinations of CONFIG_HOTPLUG and CONFIG_NET. Signed-off-by: Kay Sievers Cc: Nigel Cunningham Cc: Andrew Morton Signed-off-by: Greg Kroah-Hartman --- include/linux/kobject.h | 2 +- lib/kobject_uevent.c | 8 +++++++- 2 files changed, 8 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/kobject.h b/include/linux/kobject.h index dcd0623be892..e38bb357282a 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -259,7 +259,7 @@ struct subsys_attribute { extern int subsys_create_file(struct subsystem * , struct subsys_attribute *); extern void subsys_remove_file(struct subsystem * , struct subsys_attribute *); -#if defined(CONFIG_HOTPLUG) && defined(CONFIG_NET) +#if defined(CONFIG_HOTPLUG) void kobject_uevent(struct kobject *kobj, enum kobject_action action); int add_uevent_var(char **envp, int num_envp, int *cur_index, diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c index 982226daf939..7f20e7b857cb 100644 --- a/lib/kobject_uevent.c +++ b/lib/kobject_uevent.c @@ -25,11 +25,13 @@ #define BUFFER_SIZE 2048 /* buffer for the variables */ #define NUM_ENVP 32 /* number of env pointers */ -#if defined(CONFIG_HOTPLUG) && defined(CONFIG_NET) +#if defined(CONFIG_HOTPLUG) u64 uevent_seqnum; char uevent_helper[UEVENT_HELPER_PATH_LEN] = "/sbin/hotplug"; static DEFINE_SPINLOCK(sequence_lock); +#if defined(CONFIG_NET) static struct sock *uevent_sock; +#endif static char *action_to_string(enum kobject_action action) { @@ -155,6 +157,7 @@ void kobject_uevent(struct kobject *kobj, enum kobject_action action) spin_unlock(&sequence_lock); sprintf(seq_buff, "SEQNUM=%llu", (unsigned long long)seq); +#if defined(CONFIG_NET) /* send netlink message */ if (uevent_sock) { struct sk_buff *skb; @@ -179,6 +182,7 @@ void kobject_uevent(struct kobject *kobj, enum kobject_action action) netlink_broadcast(uevent_sock, skb, 0, 1, GFP_KERNEL); } } +#endif /* call uevent_helper, usually only enabled during early boot */ if (uevent_helper[0]) { @@ -249,6 +253,7 @@ int add_uevent_var(char **envp, int num_envp, int *cur_index, } EXPORT_SYMBOL_GPL(add_uevent_var); +#if defined(CONFIG_NET) static int __init kobject_uevent_init(void) { uevent_sock = netlink_kernel_create(NETLINK_KOBJECT_UEVENT, 1, NULL, @@ -264,5 +269,6 @@ static int __init kobject_uevent_init(void) } postcore_initcall(kobject_uevent_init); +#endif #endif /* CONFIG_HOTPLUG */ -- cgit v1.2.3 From bde11d794206ae8d72defd0e8a481181200f7dc4 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Tue, 18 Apr 2006 21:30:22 -0700 Subject: [PATCH] Fix OCFS2 warning when DEBUG_FS is not enabled Fix the following warning which happens when OCFS2_FS is enabled but DEBUG_FS isn't: fs/ocfs2/dlmglue.c: In function `ocfs2_dlm_init_debug': fs/ocfs2/dlmglue.c:2036: warning: passing arg 5 of `debugfs_create_file' discards qualifiers from pointer target type Signed-off-by: Jean Delvare Cc: Arjan van de Ven Cc: Joel Becker Acked-by: Mark Fasheh Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman --- include/linux/debugfs.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h index 176e2d371577..047567d34ca7 100644 --- a/include/linux/debugfs.h +++ b/include/linux/debugfs.h @@ -58,9 +58,8 @@ struct dentry *debugfs_create_blob(const char *name, mode_t mode, */ static inline struct dentry *debugfs_create_file(const char *name, mode_t mode, - struct dentry *parent, - void *data, - struct file_operations *fops) + struct dentry *parent, void *data, + const struct file_operations *fops) { return ERR_PTR(-ENODEV); } -- cgit v1.2.3 From 5b3ef14e3e9d745a512d65fcb4ef9be541226d80 Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Sat, 22 Apr 2006 12:14:44 +0200 Subject: [PATCH] Kobject: possible cleanups This patch contains the following possible cleanups: - #if 0 the following unused global function: - subsys_remove_file() - remove the following unused EXPORT_SYMBOL's: - kset_find_obj - subsystem_init - remove the following unused EXPORT_SYMBOL_GPL: - kobject_add_dir Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman --- include/linux/kobject.h | 1 - lib/kobject.c | 7 ++----- 2 files changed, 2 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/linux/kobject.h b/include/linux/kobject.h index e38bb357282a..c187c53cecd0 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -257,7 +257,6 @@ struct subsys_attribute { }; extern int subsys_create_file(struct subsystem * , struct subsys_attribute *); -extern void subsys_remove_file(struct subsystem * , struct subsys_attribute *); #if defined(CONFIG_HOTPLUG) void kobject_uevent(struct kobject *kobj, enum kobject_action action); diff --git a/lib/kobject.c b/lib/kobject.c index 01d957513940..b46350c27837 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -422,7 +422,6 @@ struct kobject *kobject_add_dir(struct kobject *parent, const char *name) return k; } -EXPORT_SYMBOL_GPL(kobject_add_dir); /** * kset_init - initialize a kset for use @@ -569,7 +568,7 @@ int subsys_create_file(struct subsystem * s, struct subsys_attribute * a) * @s: subsystem. * @a: attribute desciptor. */ - +#if 0 void subsys_remove_file(struct subsystem * s, struct subsys_attribute * a) { if (subsys_get(s)) { @@ -577,6 +576,7 @@ void subsys_remove_file(struct subsystem * s, struct subsys_attribute * a) subsys_put(s); } } +#endif /* 0 */ EXPORT_SYMBOL(kobject_init); EXPORT_SYMBOL(kobject_register); @@ -588,10 +588,7 @@ EXPORT_SYMBOL(kobject_del); EXPORT_SYMBOL(kset_register); EXPORT_SYMBOL(kset_unregister); -EXPORT_SYMBOL(kset_find_obj); -EXPORT_SYMBOL(subsystem_init); EXPORT_SYMBOL(subsystem_register); EXPORT_SYMBOL(subsystem_unregister); EXPORT_SYMBOL(subsys_create_file); -EXPORT_SYMBOL(subsys_remove_file); -- cgit v1.2.3 From f0fe253c4719faf76d40f581cdc0e8aef77273bb Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Sat, 22 Apr 2006 09:36:07 -0500 Subject: [IA64-SGI] - Fix discover of nearest cpu node to IO node Fix a bug that causes discovery of the nearest node/cpu to a TIO (IO node) to fail. Signed-off-by: Jack Steiner Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/sn2/sn_hwperf.c | 4 ++-- include/asm-ia64/sn/sn2/sn_hwperf.h | 6 +++++- 2 files changed, 7 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/arch/ia64/sn/kernel/sn2/sn_hwperf.c b/arch/ia64/sn/kernel/sn2/sn_hwperf.c index d917afa30b27..7ec65bc0ccf6 100644 --- a/arch/ia64/sn/kernel/sn2/sn_hwperf.c +++ b/arch/ia64/sn/kernel/sn2/sn_hwperf.c @@ -284,6 +284,8 @@ static int sn_hwperf_get_nearest_node_objdata(struct sn_hwperf_object_info *objb /* find nearest node with cpus and nearest memory */ for (router=NULL, j=0; j < op->ports; j++) { dest = sn_hwperf_findobj_id(objbuf, nobj, ptdata[j].conn_id); + if (dest && SN_HWPERF_IS_ROUTER(dest)) + router = dest; if (!dest || SN_HWPERF_FOREIGN(dest) || !SN_HWPERF_IS_NODE(dest) || SN_HWPERF_IS_IONODE(dest)) { continue; @@ -299,8 +301,6 @@ static int sn_hwperf_get_nearest_node_objdata(struct sn_hwperf_object_info *objb *near_mem_node = c; found_mem++; } - if (SN_HWPERF_IS_ROUTER(dest)) - router = dest; } if (router && (!found_cpu || !found_mem)) { diff --git a/include/asm-ia64/sn/sn2/sn_hwperf.h b/include/asm-ia64/sn/sn2/sn_hwperf.h index 291ef3d69da2..e61ebac38cdd 100644 --- a/include/asm-ia64/sn/sn2/sn_hwperf.h +++ b/include/asm-ia64/sn/sn2/sn_hwperf.h @@ -45,8 +45,12 @@ struct sn_hwperf_object_info { #define SN_HWPERF_IS_NODE(x) ((x) && strstr((x)->name, "SHub")) #define SN_HWPERF_IS_NODE_SHUB2(x) ((x) && strstr((x)->name, "SHub 2.")) #define SN_HWPERF_IS_IONODE(x) ((x) && strstr((x)->name, "TIO")) -#define SN_HWPERF_IS_ROUTER(x) ((x) && strstr((x)->name, "Router")) #define SN_HWPERF_IS_NL3ROUTER(x) ((x) && strstr((x)->name, "NL3Router")) +#define SN_HWPERF_IS_NL4ROUTER(x) ((x) && strstr((x)->name, "NL4Router")) +#define SN_HWPERF_IS_OLDROUTER(x) ((x) && strstr((x)->name, "Router")) +#define SN_HWPERF_IS_ROUTER(x) (SN_HWPERF_IS_NL3ROUTER(x) || \ + SN_HWPERF_IS_NL4ROUTER(x) || \ + SN_HWPERF_IS_OLDROUTER(x)) #define SN_HWPERF_FOREIGN(x) ((x) && !(x)->sn_hwp_this_part && !(x)->sn_hwp_is_shared) #define SN_HWPERF_SAME_OBJTYPE(x,y) ((SN_HWPERF_IS_NODE(x) && SN_HWPERF_IS_NODE(y)) ||\ (SN_HWPERF_IS_IONODE(x) && SN_HWPERF_IS_IONODE(y)) ||\ -- cgit v1.2.3 From 1df57c0c21c92a6d4fcfe5304c84151ed9beb7a2 Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Tue, 25 Apr 2006 10:47:48 -0500 Subject: [IA64] enable dumps to capture second page of kernel stack In SLES10 (2.6.16) crash dumping (in my experience, LKCD) is unable to capture the second page of the 2-page task/stack allocation. This is particularly troublesome for dump analysis, as the stack traceback cannot be done. (A similar convention is probably needed throughout the kernel to make kernel multi-page allocations detectable for dumping) Multi-page kernel allocations are represented by the single page structure associated with the first page of the allocation. The page structures associated with the other pages are unintialized. If the dumper is selecting only kernel pages it has no way to identify any but the first page of the allocation. The fix is to make the task/stack allocation a compound page. Signed-off-by: Cliff Wickman Signed-off-by: Tony Luck --- include/asm-ia64/thread_info.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/asm-ia64/thread_info.h b/include/asm-ia64/thread_info.h index 56394a2c7055..e5392c4d30c6 100644 --- a/include/asm-ia64/thread_info.h +++ b/include/asm-ia64/thread_info.h @@ -67,7 +67,7 @@ struct thread_info { #define end_of_stack(p) (unsigned long *)((void *)(p) + IA64_RBS_OFFSET) #define __HAVE_ARCH_TASK_STRUCT_ALLOCATOR -#define alloc_task_struct() ((task_t *)__get_free_pages(GFP_KERNEL, KERNEL_STACK_SIZE_ORDER)) +#define alloc_task_struct() ((task_t *)__get_free_pages(GFP_KERNEL | __GFP_COMP, KERNEL_STACK_SIZE_ORDER)) #define free_task_struct(tsk) free_pages((unsigned long) (tsk), KERNEL_STACK_SIZE_ORDER) #endif /* !__ASSEMBLY */ -- cgit v1.2.3 From f10a04c034c7285a1b15dfa4a83d3e56578e34e8 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Fri, 28 Apr 2006 15:02:51 +1000 Subject: [PATCH] powerpc: Fix pagetable bloat for hugepages At present, ARCH=powerpc kernels can waste considerable space in pagetables when making large hugepage mappings. Hugepage PTEs go in PMD pages, but each PMD page maps 256M and so contains only 16 hugepage PTEs (128 bytes of data), but takes up a 1024 byte allocation. With CONFIG_PPC_64K_PAGES enabled (64k base page size), the situation is worse. Now hugepage PTEs are at the PTE page level (also mapping 256M), so we store 16 hugepage PTEs in a 64k allocation. The PowerPC MMU already means that any 256M region is either all hugepage, or all normal pages. Thus, with some care, we can use a different allocation for the hugepage PTE tables and only allocate the 128 bytes necessary. Signed-off-by: Paul Mackerras --- arch/powerpc/mm/hugetlbpage.c | 295 ++++++++++++++++++++++++++++++++++++------ arch/powerpc/mm/init_64.c | 7 + include/asm-powerpc/page_64.h | 1 + include/asm-powerpc/pgalloc.h | 2 + 4 files changed, 269 insertions(+), 36 deletions(-) (limited to 'include') diff --git a/arch/powerpc/mm/hugetlbpage.c b/arch/powerpc/mm/hugetlbpage.c index 7370f9f33e29..266b8b2ceac9 100644 --- a/arch/powerpc/mm/hugetlbpage.c +++ b/arch/powerpc/mm/hugetlbpage.c @@ -30,13 +30,66 @@ #define NUM_LOW_AREAS (0x100000000UL >> SID_SHIFT) #define NUM_HIGH_AREAS (PGTABLE_RANGE >> HTLB_AREA_SHIFT) +#ifdef CONFIG_PPC_64K_PAGES +#define HUGEPTE_INDEX_SIZE (PMD_SHIFT-HPAGE_SHIFT) +#else +#define HUGEPTE_INDEX_SIZE (PUD_SHIFT-HPAGE_SHIFT) +#endif +#define PTRS_PER_HUGEPTE (1 << HUGEPTE_INDEX_SIZE) +#define HUGEPTE_TABLE_SIZE (sizeof(pte_t) << HUGEPTE_INDEX_SIZE) + +#define HUGEPD_SHIFT (HPAGE_SHIFT + HUGEPTE_INDEX_SIZE) +#define HUGEPD_SIZE (1UL << HUGEPD_SHIFT) +#define HUGEPD_MASK (~(HUGEPD_SIZE-1)) + +#define huge_pgtable_cache (pgtable_cache[HUGEPTE_CACHE_NUM]) + +/* Flag to mark huge PD pointers. This means pmd_bad() and pud_bad() + * will choke on pointers to hugepte tables, which is handy for + * catching screwups early. */ +#define HUGEPD_OK 0x1 + +typedef struct { unsigned long pd; } hugepd_t; + +#define hugepd_none(hpd) ((hpd).pd == 0) + +static inline pte_t *hugepd_page(hugepd_t hpd) +{ + BUG_ON(!(hpd.pd & HUGEPD_OK)); + return (pte_t *)(hpd.pd & ~HUGEPD_OK); +} + +static inline pte_t *hugepte_offset(hugepd_t *hpdp, unsigned long addr) +{ + unsigned long idx = ((addr >> HPAGE_SHIFT) & (PTRS_PER_HUGEPTE-1)); + pte_t *dir = hugepd_page(*hpdp); + + return dir + idx; +} + +static int __hugepte_alloc(struct mm_struct *mm, hugepd_t *hpdp, + unsigned long address) +{ + pte_t *new = kmem_cache_alloc(huge_pgtable_cache, + GFP_KERNEL|__GFP_REPEAT); + + if (! new) + return -ENOMEM; + + spin_lock(&mm->page_table_lock); + if (!hugepd_none(*hpdp)) + kmem_cache_free(huge_pgtable_cache, new); + else + hpdp->pd = (unsigned long)new | HUGEPD_OK; + spin_unlock(&mm->page_table_lock); + return 0; +} + /* Modelled after find_linux_pte() */ pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr) { pgd_t *pg; pud_t *pu; - pmd_t *pm; - pte_t *pt; BUG_ON(! in_hugepage_area(mm->context, addr)); @@ -46,26 +99,14 @@ pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr) if (!pgd_none(*pg)) { pu = pud_offset(pg, addr); if (!pud_none(*pu)) { - pm = pmd_offset(pu, addr); #ifdef CONFIG_PPC_64K_PAGES - /* Currently, we use the normal PTE offset within full - * size PTE pages, thus our huge PTEs are scattered in - * the PTE page and we do waste some. We may change - * that in the future, but the current mecanism keeps - * things much simpler - */ - if (!pmd_none(*pm)) { - /* Note: pte_offset_* are all equivalent on - * ppc64 as we don't have HIGHMEM - */ - pt = pte_offset_kernel(pm, addr); - return pt; - } -#else /* CONFIG_PPC_64K_PAGES */ - /* On 4k pages, we put huge PTEs in the PMD page */ - pt = (pte_t *)pm; - return pt; -#endif /* CONFIG_PPC_64K_PAGES */ + pmd_t *pm; + pm = pmd_offset(pu, addr); + if (!pmd_none(*pm)) + return hugepte_offset((hugepd_t *)pm, addr); +#else + return hugepte_offset((hugepd_t *)pu, addr); +#endif } } @@ -76,8 +117,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr) { pgd_t *pg; pud_t *pu; - pmd_t *pm; - pte_t *pt; + hugepd_t *hpdp = NULL; BUG_ON(! in_hugepage_area(mm->context, addr)); @@ -87,23 +127,182 @@ pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr) pu = pud_alloc(mm, pg, addr); if (pu) { +#ifdef CONFIG_PPC_64K_PAGES + pmd_t *pm; pm = pmd_alloc(mm, pu, addr); - if (pm) { + if (pm) + hpdp = (hugepd_t *)pm; +#else + hpdp = (hugepd_t *)pu; +#endif + } + + if (! hpdp) + return NULL; + + if (hugepd_none(*hpdp) && __hugepte_alloc(mm, hpdp, addr)) + return NULL; + + return hugepte_offset(hpdp, addr); +} + +static void free_hugepte_range(struct mmu_gather *tlb, hugepd_t *hpdp) +{ + pte_t *hugepte = hugepd_page(*hpdp); + + hpdp->pd = 0; + tlb->need_flush = 1; + pgtable_free_tlb(tlb, pgtable_free_cache(hugepte, HUGEPTE_CACHE_NUM, + HUGEPTE_TABLE_SIZE-1)); +} + #ifdef CONFIG_PPC_64K_PAGES - /* See comment in huge_pte_offset. Note that if we ever - * want to put the page size in the PMD, we would have - * to open code our own pte_alloc* function in order - * to populate and set the size atomically - */ - pt = pte_alloc_map(mm, pm, addr); -#else /* CONFIG_PPC_64K_PAGES */ - pt = (pte_t *)pm; -#endif /* CONFIG_PPC_64K_PAGES */ - return pt; - } +static void hugetlb_free_pmd_range(struct mmu_gather *tlb, pud_t *pud, + unsigned long addr, unsigned long end, + unsigned long floor, unsigned long ceiling) +{ + pmd_t *pmd; + unsigned long next; + unsigned long start; + + start = addr; + pmd = pmd_offset(pud, addr); + do { + next = pmd_addr_end(addr, end); + if (pmd_none(*pmd)) + continue; + free_hugepte_range(tlb, (hugepd_t *)pmd); + } while (pmd++, addr = next, addr != end); + + start &= PUD_MASK; + if (start < floor) + return; + if (ceiling) { + ceiling &= PUD_MASK; + if (!ceiling) + return; } + if (end - 1 > ceiling - 1) + return; - return NULL; + pmd = pmd_offset(pud, start); + pud_clear(pud); + pmd_free_tlb(tlb, pmd); +} +#endif + +static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, + unsigned long addr, unsigned long end, + unsigned long floor, unsigned long ceiling) +{ + pud_t *pud; + unsigned long next; + unsigned long start; + + start = addr; + pud = pud_offset(pgd, addr); + do { + next = pud_addr_end(addr, end); +#ifdef CONFIG_PPC_64K_PAGES + if (pud_none_or_clear_bad(pud)) + continue; + hugetlb_free_pmd_range(tlb, pud, addr, next, floor, ceiling); +#else + if (pud_none(*pud)) + continue; + free_hugepte_range(tlb, (hugepd_t *)pud); +#endif + } while (pud++, addr = next, addr != end); + + start &= PGDIR_MASK; + if (start < floor) + return; + if (ceiling) { + ceiling &= PGDIR_MASK; + if (!ceiling) + return; + } + if (end - 1 > ceiling - 1) + return; + + pud = pud_offset(pgd, start); + pgd_clear(pgd); + pud_free_tlb(tlb, pud); +} + +/* + * This function frees user-level page tables of a process. + * + * Must be called with pagetable lock held. + */ +void hugetlb_free_pgd_range(struct mmu_gather **tlb, + unsigned long addr, unsigned long end, + unsigned long floor, unsigned long ceiling) +{ + pgd_t *pgd; + unsigned long next; + unsigned long start; + + /* + * Comments below take from the normal free_pgd_range(). They + * apply here too. The tests against HUGEPD_MASK below are + * essential, because we *don't* test for this at the bottom + * level. Without them we'll attempt to free a hugepte table + * when we unmap just part of it, even if there are other + * active mappings using it. + * + * The next few lines have given us lots of grief... + * + * Why are we testing HUGEPD* at this top level? Because + * often there will be no work to do at all, and we'd prefer + * not to go all the way down to the bottom just to discover + * that. + * + * Why all these "- 1"s? Because 0 represents both the bottom + * of the address space and the top of it (using -1 for the + * top wouldn't help much: the masks would do the wrong thing). + * The rule is that addr 0 and floor 0 refer to the bottom of + * the address space, but end 0 and ceiling 0 refer to the top + * Comparisons need to use "end - 1" and "ceiling - 1" (though + * that end 0 case should be mythical). + * + * Wherever addr is brought up or ceiling brought down, we + * must be careful to reject "the opposite 0" before it + * confuses the subsequent tests. But what about where end is + * brought down by HUGEPD_SIZE below? no, end can't go down to + * 0 there. + * + * Whereas we round start (addr) and ceiling down, by different + * masks at different levels, in order to test whether a table + * now has no other vmas using it, so can be freed, we don't + * bother to round floor or end up - the tests don't need that. + */ + + addr &= HUGEPD_MASK; + if (addr < floor) { + addr += HUGEPD_SIZE; + if (!addr) + return; + } + if (ceiling) { + ceiling &= HUGEPD_MASK; + if (!ceiling) + return; + } + if (end - 1 > ceiling - 1) + end -= HUGEPD_SIZE; + if (addr > end - 1) + return; + + start = addr; + pgd = pgd_offset((*tlb)->mm, addr); + do { + BUG_ON(! in_hugepage_area((*tlb)->mm->context, addr)); + next = pgd_addr_end(addr, end); + if (pgd_none_or_clear_bad(pgd)) + continue; + hugetlb_free_pud_range(*tlb, pgd, addr, next, floor, ceiling); + } while (pgd++, addr = next, addr != end); } void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, @@ -841,3 +1040,27 @@ repeat: out: return err; } + +static void zero_ctor(void *addr, kmem_cache_t *cache, unsigned long flags) +{ + memset(addr, 0, kmem_cache_size(cache)); +} + +static int __init hugetlbpage_init(void) +{ + if (!cpu_has_feature(CPU_FTR_16M_PAGE)) + return -ENODEV; + + huge_pgtable_cache = kmem_cache_create("hugepte_cache", + HUGEPTE_TABLE_SIZE, + HUGEPTE_TABLE_SIZE, + SLAB_HWCACHE_ALIGN | + SLAB_MUST_HWCACHE_ALIGN, + zero_ctor, NULL); + if (! huge_pgtable_cache) + panic("hugetlbpage_init(): could not create hugepte cache\n"); + + return 0; +} + +module_init(hugetlbpage_init); diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c index babebd15bdc4..9e30f968c184 100644 --- a/arch/powerpc/mm/init_64.c +++ b/arch/powerpc/mm/init_64.c @@ -162,7 +162,14 @@ static const char *pgtable_cache_name[ARRAY_SIZE(pgtable_cache_size)] = { }; #endif /* CONFIG_PPC_64K_PAGES */ +#ifdef CONFIG_HUGETLB_PAGE +/* Hugepages need one extra cache, initialized in hugetlbpage.c. We + * can't put into the tables above, because HPAGE_SHIFT is not compile + * time constant. */ +kmem_cache_t *pgtable_cache[ARRAY_SIZE(pgtable_cache_size)+1]; +#else kmem_cache_t *pgtable_cache[ARRAY_SIZE(pgtable_cache_size)]; +#endif void pgtable_cache_init(void) { diff --git a/include/asm-powerpc/page_64.h b/include/asm-powerpc/page_64.h index 3fb061bab9ec..eab779c21995 100644 --- a/include/asm-powerpc/page_64.h +++ b/include/asm-powerpc/page_64.h @@ -101,6 +101,7 @@ extern unsigned int HPAGE_SHIFT; - (1U << GET_HTLB_AREA(addr))) & 0xffff) #define ARCH_HAS_HUGEPAGE_ONLY_RANGE +#define ARCH_HAS_HUGETLB_FREE_PGD_RANGE #define ARCH_HAS_PREPARE_HUGEPAGE_RANGE #define ARCH_HAS_SETCLEAR_HUGE_PTE diff --git a/include/asm-powerpc/pgalloc.h b/include/asm-powerpc/pgalloc.h index a00ee002cd11..9f0917c68659 100644 --- a/include/asm-powerpc/pgalloc.h +++ b/include/asm-powerpc/pgalloc.h @@ -17,11 +17,13 @@ extern kmem_cache_t *pgtable_cache[]; #define PTE_CACHE_NUM 0 #define PMD_CACHE_NUM 1 #define PGD_CACHE_NUM 2 +#define HUGEPTE_CACHE_NUM 3 #else #define PTE_CACHE_NUM 0 #define PMD_CACHE_NUM 1 #define PUD_CACHE_NUM 1 #define PGD_CACHE_NUM 0 +#define HUGEPTE_CACHE_NUM 2 #endif /* -- cgit v1.2.3 From 1269277a5e7c6d7ae1852e648a8bcdb78035e9fa Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Mon, 24 Apr 2006 23:22:17 +0100 Subject: [PATCH] powerpc: Use check_legacy_ioport() on ppc32 too. Some people report that we die on some Macs when we are expecting to catch machine checks after poking at some random I/O address. I'd seen it happen on my dual G4 with serial ports until we fixed those to use OF, but now other users are reporting it with i8042. This expands the use of check_legacy_ioport() to avoid that situation even on 32-bit kernels. Signed-off-by: David Woodhouse Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/setup-common.c | 8 ++++++++ arch/powerpc/kernel/setup_64.c | 8 -------- drivers/block/floppy.c | 2 +- drivers/input/serio/i8042-io.h | 4 ++-- include/asm-powerpc/io.h | 6 +++--- 5 files changed, 14 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index 1d93e73a7003..684ab1d49c65 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -516,3 +516,11 @@ void probe_machine(void) printk(KERN_INFO "Using %s machine description\n", ppc_md.name); } + +int check_legacy_ioport(unsigned long base_port) +{ + if (ppc_md.check_legacy_ioport == NULL) + return 0; + return ppc_md.check_legacy_ioport(base_port); +} +EXPORT_SYMBOL(check_legacy_ioport); diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index 13e91c4d70a8..4467c49903b6 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -594,14 +594,6 @@ void ppc64_terminate_msg(unsigned int src, const char *msg) printk("[terminate]%04x %s\n", src, msg); } -int check_legacy_ioport(unsigned long base_port) -{ - if (ppc_md.check_legacy_ioport == NULL) - return 0; - return ppc_md.check_legacy_ioport(base_port); -} -EXPORT_SYMBOL(check_legacy_ioport); - void cpu_die(void) { if (ppc_md.cpu_die) diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c index bedb689b051f..dff1e67b1dd4 100644 --- a/drivers/block/floppy.c +++ b/drivers/block/floppy.c @@ -4301,7 +4301,7 @@ static int __init floppy_init(void) } use_virtual_dma = can_use_virtual_dma & 1; -#if defined(CONFIG_PPC64) +#if defined(CONFIG_PPC_MERGE) if (check_legacy_ioport(FDC1)) { del_timer(&fd_timeout); err = -ENODEV; diff --git a/drivers/input/serio/i8042-io.h b/drivers/input/serio/i8042-io.h index 9a9221644250..cc21914fbc72 100644 --- a/drivers/input/serio/i8042-io.h +++ b/drivers/input/serio/i8042-io.h @@ -67,14 +67,14 @@ static inline int i8042_platform_init(void) * On some platforms touching the i8042 data register region can do really * bad things. Because of this the region is always reserved on such boxes. */ -#if !defined(__sh__) && !defined(__alpha__) && !defined(__mips__) && !defined(CONFIG_PPC64) +#if !defined(__sh__) && !defined(__alpha__) && !defined(__mips__) && !defined(CONFIG_PPC_MERGE) if (!request_region(I8042_DATA_REG, 16, "i8042")) return -EBUSY; #endif i8042_reset = 1; -#if defined(CONFIG_PPC64) +#if defined(CONFIG_PPC_MERGE) if (check_legacy_ioport(I8042_DATA_REG)) return -EBUSY; if (!request_region(I8042_DATA_REG, 16, "i8042")) diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h index 68efbea379c9..f1c2469b8844 100644 --- a/include/asm-powerpc/io.h +++ b/include/asm-powerpc/io.h @@ -9,6 +9,9 @@ * 2 of the License, or (at your option) any later version. */ +/* Check of existence of legacy devices */ +extern int check_legacy_ioport(unsigned long base_port); + #ifndef CONFIG_PPC64 #include #else @@ -437,9 +440,6 @@ out: #define dma_cache_wback(_start,_size) do { } while (0) #define dma_cache_wback_inv(_start,_size) do { } while (0) -/* Check of existence of legacy devices */ -extern int check_legacy_ioport(unsigned long base_port); - /* * Convert a physical pointer to a virtual kernel pointer for /dev/mem -- cgit v1.2.3 From 30aacebed0f0619f23ce84df7c59ad033ca08d77 Mon Sep 17 00:00:00 2001 From: Eugene Surovegin Date: Tue, 25 Apr 2006 01:22:44 -0700 Subject: [PATCH] ppc32: add 440GX erratum 440_43 workaround This patch adds workaround for PPC 440GX erratum 440_43. According to this erratum spurious MachineChecks (caused by L1 cache parity) can happen during DataTLB miss processing. We disable L1 cache parity checking for 440GX rev.C and rev.F Signed-off-by: Eugene Surovegin Signed-off-by: Paul Mackerras --- arch/ppc/platforms/4xx/ocotea.c | 2 +- arch/ppc/syslib/ibm440gx_common.c | 13 ++++++++++++- arch/ppc/syslib/ibm440gx_common.h | 4 ++++ include/asm-ppc/reg_booke.h | 1 + 4 files changed, 18 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/ppc/platforms/4xx/ocotea.c b/arch/ppc/platforms/4xx/ocotea.c index f841972f1fa9..554776d4b8ac 100644 --- a/arch/ppc/platforms/4xx/ocotea.c +++ b/arch/ppc/platforms/4xx/ocotea.c @@ -331,7 +331,7 @@ static void __init ocotea_init(void) void __init platform_init(unsigned long r3, unsigned long r4, unsigned long r5, unsigned long r6, unsigned long r7) { - ibm44x_platform_init(r3, r4, r5, r6, r7); + ibm440gx_platform_init(r3, r4, r5, r6, r7); ppc_md.setup_arch = ocotea_setup_arch; ppc_md.show_cpuinfo = ocotea_show_cpuinfo; diff --git a/arch/ppc/syslib/ibm440gx_common.c b/arch/ppc/syslib/ibm440gx_common.c index a7dd55f1c63e..f6cc16888527 100644 --- a/arch/ppc/syslib/ibm440gx_common.c +++ b/arch/ppc/syslib/ibm440gx_common.c @@ -2,7 +2,7 @@ * PPC440GX system library * * Eugene Surovegin or - * Copyright (c) 2003, 2004 Zultys Technologies + * Copyright (c) 2003 - 2006 Zultys Technologies * * 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 @@ -282,3 +282,14 @@ int ibm440gx_show_cpuinfo(struct seq_file *m){ return 0; } +void __init ibm440gx_platform_init(unsigned long r3, unsigned long r4, + unsigned long r5, unsigned long r6, + unsigned long r7) +{ + /* Erratum 440_43 workaround, disable L1 cache parity checking */ + if (!strcmp(cur_cpu_spec->cpu_name, "440GX Rev. C") || + !strcmp(cur_cpu_spec->cpu_name, "440GX Rev. F")) + mtspr(SPRN_CCR1, mfspr(SPRN_CCR1) | CCR1_DPC); + + ibm44x_platform_init(r3, r4, r5, r6, r7); +} diff --git a/arch/ppc/syslib/ibm440gx_common.h b/arch/ppc/syslib/ibm440gx_common.h index a2ab9fab8e34..a03ec6022e8f 100644 --- a/arch/ppc/syslib/ibm440gx_common.h +++ b/arch/ppc/syslib/ibm440gx_common.h @@ -29,6 +29,10 @@ void ibm440gx_get_clocks(struct ibm44x_clocks*, unsigned int sys_clk, unsigned int ser_clk) __init; +/* common 440GX platform init */ +void ibm440gx_platform_init(unsigned long r3, unsigned long r4, unsigned long r5, + unsigned long r6, unsigned long r7) __init; + /* Enable L2 cache */ void ibm440gx_l2c_enable(void) __init; diff --git a/include/asm-ppc/reg_booke.h b/include/asm-ppc/reg_booke.h index 00ad9c754c78..4944c0fb8bea 100644 --- a/include/asm-ppc/reg_booke.h +++ b/include/asm-ppc/reg_booke.h @@ -237,6 +237,7 @@ do { \ #endif /* Bit definitions for CCR1. */ +#define CCR1_DPC 0x00000100 /* Disable L1 I-Cache/D-Cache parity checking */ #define CCR1_TCS 0x00000080 /* Timer Clock Select */ /* Bit definitions for the MCSR. */ -- cgit v1.2.3 From 2833c28aa0d0326780acfa61149a2a02dcb2c9b4 Mon Sep 17 00:00:00 2001 From: Andreas Schwab Date: Thu, 27 Apr 2006 15:46:42 +0200 Subject: [PATCH] powerpc: Wire up *at syscalls Wire up *at syscalls. This patch has been tested on ppc64 (using glibc's testsuite, both 32bit and 64bit), and compile-tested for ppc32 (I have currently no ppc32 system available, but I expect no problems). Signed-off-by: Andreas Schwab Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/systbl.S | 13 +++++++++++++ arch/powerpc/platforms/cell/spu_callbacks.c | 13 +++++++++++++ fs/stat.c | 2 +- include/asm-powerpc/unistd.h | 20 +++++++++++++++++++- 4 files changed, 46 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/powerpc/kernel/systbl.S b/arch/powerpc/kernel/systbl.S index 0b98eea73c5e..cf56a1d499ff 100644 --- a/arch/powerpc/kernel/systbl.S +++ b/arch/powerpc/kernel/systbl.S @@ -325,6 +325,19 @@ SYSCALL(unshare) SYSCALL(splice) SYSCALL(tee) SYSCALL(vmsplice) +COMPAT_SYS(openat) +SYSCALL(mkdirat) +SYSCALL(mknodat) +SYSCALL(fchownat) +COMPAT_SYS(futimesat) +SYSX(sys_newfstatat, sys_fstatat64, sys_fstatat64) +SYSCALL(unlinkat) +SYSCALL(renameat) +SYSCALL(linkat) +SYSCALL(symlinkat) +SYSCALL(readlinkat) +SYSCALL(fchmodat) +SYSCALL(faccessat) /* * please add new calls to arch/powerpc/platforms/cell/spu_callbacks.c diff --git a/arch/powerpc/platforms/cell/spu_callbacks.c b/arch/powerpc/platforms/cell/spu_callbacks.c index b283380a2a18..95b36430aa0f 100644 --- a/arch/powerpc/platforms/cell/spu_callbacks.c +++ b/arch/powerpc/platforms/cell/spu_callbacks.c @@ -319,6 +319,19 @@ void *spu_syscall_table[] = { [__NR_splice] sys_splice, [__NR_tee] sys_tee, [__NR_vmsplice] sys_vmsplice, + [__NR_openat] sys_openat, + [__NR_mkdirat] sys_mkdirat, + [__NR_mknodat] sys_mknodat, + [__NR_fchownat] sys_fchownat, + [__NR_futimesat] sys_futimesat, + [__NR_newfstatat] sys_newfstatat, + [__NR_unlinkat] sys_unlinkat, + [__NR_renameat] sys_renameat, + [__NR_linkat] sys_linkat, + [__NR_symlinkat] sys_symlinkat, + [__NR_readlinkat] sys_readlinkat, + [__NR_fchmodat] sys_fchmodat, + [__NR_faccessat] sys_faccessat, }; long spu_sys_callback(struct spu_syscall_block *s) diff --git a/fs/stat.c b/fs/stat.c index 9948cc1685a4..0f282face322 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -261,7 +261,7 @@ asmlinkage long sys_newlstat(char __user *filename, struct stat __user *statbuf) return error; } -#ifndef __ARCH_WANT_STAT64 +#if !defined(__ARCH_WANT_STAT64) || defined(__ARCH_WANT_SYS_NEWFSTATAT) asmlinkage long sys_newfstatat(int dfd, char __user *filename, struct stat __user *statbuf, int flag) { diff --git a/include/asm-powerpc/unistd.h b/include/asm-powerpc/unistd.h index 34325e292596..908acb44cb8a 100644 --- a/include/asm-powerpc/unistd.h +++ b/include/asm-powerpc/unistd.h @@ -304,8 +304,25 @@ #define __NR_splice 283 #define __NR_tee 284 #define __NR_vmsplice 285 +#define __NR_openat 286 +#define __NR_mkdirat 287 +#define __NR_mknodat 288 +#define __NR_fchownat 289 +#define __NR_futimesat 290 +#ifdef __powerpc64__ +#define __NR_newfstatat 291 +#else +#define __NR_fstatat64 291 +#endif +#define __NR_unlinkat 292 +#define __NR_renameat 293 +#define __NR_linkat 294 +#define __NR_symlinkat 295 +#define __NR_readlinkat 296 +#define __NR_fchmodat 297 +#define __NR_faccessat 298 -#define __NR_syscalls 286 +#define __NR_syscalls 299 #ifdef __KERNEL__ #define __NR__exit __NR_exit @@ -458,6 +475,7 @@ type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4, type5 arg5, type6 arg6 #ifdef CONFIG_PPC64 #define __ARCH_WANT_COMPAT_SYS_TIME #define __ARCH_WANT_COMPAT_SYS_RT_SIGSUSPEND +#define __ARCH_WANT_SYS_NEWFSTATAT #endif /* -- cgit v1.2.3 From a73c87bfe98f4d54c919e920a6efb0a116115722 Mon Sep 17 00:00:00 2001 From: Vitaly Bordug Date: Tue, 25 Apr 2006 20:26:33 +0400 Subject: [PATCH] ppc32: odd fixes and improvements in ppc_sys This consists of offsets fix in ..._devices.c, and update of ppc_sys_fixup_mem_resource() function to prevent subsequent fixups Signed-off-by: Vitaly Bordug Signed-off-by: Paul Mackerras --- arch/ppc/syslib/mpc8xx_devices.c | 25 +++++++++++++++++++------ arch/ppc/syslib/ppc_sys.c | 4 +++- arch/ppc/syslib/pq2_sys.c | 8 ++++---- include/asm-ppc/ppc_sys.h | 2 ++ 4 files changed, 28 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/arch/ppc/syslib/mpc8xx_devices.c b/arch/ppc/syslib/mpc8xx_devices.c index bd41ed83beb3..6f536383866e 100644 --- a/arch/ppc/syslib/mpc8xx_devices.c +++ b/arch/ppc/syslib/mpc8xx_devices.c @@ -170,12 +170,18 @@ struct platform_device ppc_sys_platform_devices[] = { [MPC8xx_CPM_SMC1] = { .name = "fsl-cpm-smc", .id = 1, - .num_resources = 2, + .num_resources = 3, .resource = (struct resource[]) { { .name = "regs", - .start = 0xa82, - .end = 0xa91, + .start = 0xa80, + .end = 0xa8f, + .flags = IORESOURCE_MEM, + }, + { + .name = "pram", + .start = 0x3e80, + .end = 0x3ebf, .flags = IORESOURCE_MEM, }, { @@ -189,14 +195,21 @@ struct platform_device ppc_sys_platform_devices[] = { [MPC8xx_CPM_SMC2] = { .name = "fsl-cpm-smc", .id = 2, - .num_resources = 2, + .num_resources = 3, .resource = (struct resource[]) { { .name = "regs", - .start = 0xa92, - .end = 0xaa1, + .start = 0xa90, + .end = 0xa9f, .flags = IORESOURCE_MEM, }, + { + .name = "pram", + .start = 0x3f80, + .end = 0x3fbf, + .flags = IORESOURCE_MEM, + + }, { .name = "interrupt", .start = MPC8xx_INT_SMC2, diff --git a/arch/ppc/syslib/ppc_sys.c b/arch/ppc/syslib/ppc_sys.c index 7662c4e6e7d6..2d48018b71d9 100644 --- a/arch/ppc/syslib/ppc_sys.c +++ b/arch/ppc/syslib/ppc_sys.c @@ -109,9 +109,11 @@ ppc_sys_fixup_mem_resource(struct platform_device *pdev, phys_addr_t paddr) int i; for (i = 0; i < pdev->num_resources; i++) { struct resource *r = &pdev->resource[i]; - if ((r->flags & IORESOURCE_MEM) == IORESOURCE_MEM) { + if (((r->flags & IORESOURCE_MEM) == IORESOURCE_MEM) && + ((r->flags & PPC_SYS_IORESOURCE_FIXUPPED) != PPC_SYS_IORESOURCE_FIXUPPED)) { r->start += paddr; r->end += paddr; + r->flags |= PPC_SYS_IORESOURCE_FIXUPPED; } } } diff --git a/arch/ppc/syslib/pq2_sys.c b/arch/ppc/syslib/pq2_sys.c index 75e64f1c144d..433b0fa203e1 100644 --- a/arch/ppc/syslib/pq2_sys.c +++ b/arch/ppc/syslib/pq2_sys.c @@ -113,13 +113,13 @@ struct ppc_sys_spec ppc_sys_specs[] = { .ppc_sys_name = "8248", .mask = 0x0000ff00, .value = 0x00000c00, - .num_devices = 11, + .num_devices = 12, .device_list = (enum ppc_sys_devices[]) { MPC82xx_CPM_FCC1, MPC82xx_CPM_FCC2, MPC82xx_CPM_SCC1, - MPC82xx_CPM_SCC2, MPC82xx_CPM_SCC3, MPC82xx_CPM_SMC1, - MPC82xx_CPM_SMC2, MPC82xx_CPM_SPI, MPC82xx_CPM_I2C, - MPC82xx_CPM_USB, MPC82xx_SEC1, + MPC82xx_CPM_SCC2, MPC82xx_CPM_SCC3, MPC82xx_CPM_SCC4, + MPC82xx_CPM_SMC1, MPC82xx_CPM_SMC2, MPC82xx_CPM_SPI, + MPC82xx_CPM_I2C, MPC82xx_CPM_USB, MPC82xx_SEC1, }, }, { diff --git a/include/asm-ppc/ppc_sys.h b/include/asm-ppc/ppc_sys.h index 4b94f7059ebe..40f197af6508 100644 --- a/include/asm-ppc/ppc_sys.h +++ b/include/asm-ppc/ppc_sys.h @@ -39,6 +39,8 @@ #error "need definition of ppc_sys_devices" #endif +#define PPC_SYS_IORESOURCE_FIXUPPED 0x00000001 + struct ppc_sys_spec { /* PPC sys is matched via (ID & mask) == value, id could be * PVR, SVR, IMMR, * etc. */ -- cgit v1.2.3 From ec448a0a361ce3fa97a32a5c267a680163ffe8c5 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 18 Apr 2006 09:05:39 -0700 Subject: [SCSI] srp.h: avoid padding of structs Several structs in get padded to a multiple of 8 bytes on 64-bit architectures and end up with a size that does not match the definition in the SRP spec: SRP spec 64-bit sizeof (struct indirect_buf) 20 24 sizeof (struct srp_login_rsp) 52 56 sizeof (struct srp_rsp) 36 40 Fix this by adding __attribute__((packed)) to the offending structs. Problem pointed out by Arne Redlich . Signed-off-by: Roland Dreier Signed-off-by: James Bottomley --- include/scsi/srp.h | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/scsi/srp.h b/include/scsi/srp.h index 6c2681dc5b46..637f77eccf0c 100644 --- a/include/scsi/srp.h +++ b/include/scsi/srp.h @@ -95,14 +95,15 @@ struct srp_direct_buf { /* * We need the packed attribute because the SRP spec puts the list of - * descriptors at an offset of 20, which is not aligned to the size - * of struct srp_direct_buf. + * descriptors at an offset of 20, which is not aligned to the size of + * struct srp_direct_buf. The whole structure must be packed to avoid + * having the 20-byte structure padded to 24 bytes on 64-bit architectures. */ struct srp_indirect_buf { struct srp_direct_buf table_desc; __be32 len; - struct srp_direct_buf desc_list[0] __attribute__((packed)); -}; + struct srp_direct_buf desc_list[0]; +} __attribute__((packed)); enum { SRP_MULTICHAN_SINGLE = 0, @@ -122,6 +123,11 @@ struct srp_login_req { u8 target_port_id[16]; }; +/* + * The SRP spec defines the size of the LOGIN_RSP structure to be 52 + * bytes, so it needs to be packed to avoid having it padded to 56 + * bytes on 64-bit architectures. + */ struct srp_login_rsp { u8 opcode; u8 reserved1[3]; @@ -132,7 +138,7 @@ struct srp_login_rsp { __be16 buf_fmt; u8 rsp_flags; u8 reserved2[25]; -}; +} __attribute__((packed)); struct srp_login_rej { u8 opcode; @@ -207,6 +213,11 @@ enum { SRP_RSP_FLAG_DIUNDER = 1 << 5 }; +/* + * The SRP spec defines the size of the RSP structure to be 36 bytes, + * so it needs to be packed to avoid having it padded to 40 bytes on + * 64-bit architectures. + */ struct srp_rsp { u8 opcode; u8 sol_not; @@ -221,6 +232,6 @@ struct srp_rsp { __be32 sense_data_len; __be32 resp_data_len; u8 data[0]; -}; +} __attribute__((packed)); #endif /* SCSI_SRP_H */ -- cgit v1.2.3 From 13e87ec68641fd54f3fa04eef3419d034ed2115a Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 27 Apr 2006 18:39:18 -0700 Subject: [PATCH] request_irq(): remove warnings from irq probing - Add new SA_PROBEIRQ which suppresses the new sharing-mismatch warning. Some drivers like to use request_irq() to find an unused interrupt slot. - Use it in i82365.c - Kill unused SA_PROBE. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/pcmcia/i82365.c | 7 ++++--- include/asm-xtensa/signal.h | 2 +- include/linux/signal.h | 4 +++- kernel/irq/manage.c | 6 ++++-- 4 files changed, 12 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/pcmcia/i82365.c b/drivers/pcmcia/i82365.c index bd0308e89815..a2f05f485156 100644 --- a/drivers/pcmcia/i82365.c +++ b/drivers/pcmcia/i82365.c @@ -509,7 +509,8 @@ static irqreturn_t i365_count_irq(int irq, void *dev, struct pt_regs *regs) static u_int __init test_irq(u_short sock, int irq) { debug(2, " testing ISA irq %d\n", irq); - if (request_irq(irq, i365_count_irq, 0, "scan", i365_count_irq) != 0) + if (request_irq(irq, i365_count_irq, SA_PROBEIRQ, "scan", + i365_count_irq) != 0) return 1; irq_hits = 0; irq_sock = sock; msleep(10); @@ -561,7 +562,7 @@ static u_int __init isa_scan(u_short sock, u_int mask0) } else { /* Fallback: just find interrupts that aren't in use */ for (i = 0; i < 16; i++) - if ((mask0 & (1 << i)) && (_check_irq(i, 0) == 0)) + if ((mask0 & (1 << i)) && (_check_irq(i, SA_PROBEIRQ) == 0)) mask1 |= (1 << i); printk("default"); /* If scan failed, default to polled status */ @@ -725,7 +726,7 @@ static void __init add_pcic(int ns, int type) u_int cs_mask = mask & ((cs_irq) ? (1< 0; cs_irq--) if ((cs_mask & (1 << cs_irq)) && - (_check_irq(cs_irq, 0) == 0)) + (_check_irq(cs_irq, SA_PROBEIRQ) == 0)) break; if (cs_irq) { grab_irq = 1; diff --git a/include/asm-xtensa/signal.h b/include/asm-xtensa/signal.h index 5d6fc9cdf58d..a99c9aec64ec 100644 --- a/include/asm-xtensa/signal.h +++ b/include/asm-xtensa/signal.h @@ -118,9 +118,9 @@ typedef struct { * SA_INTERRUPT is also used by the irq handling routines. * SA_SHIRQ is for shared interrupt support on PCI and EISA. */ -#define SA_PROBE SA_ONESHOT #define SA_SAMPLE_RANDOM SA_RESTART #define SA_SHIRQ 0x04000000 +#define SA_PROBEIRQ 0x08000000 #endif #define SIG_BLOCK 0 /* for blocking signals */ diff --git a/include/linux/signal.h b/include/linux/signal.h index 162a8fd10b29..70739f51a09f 100644 --- a/include/linux/signal.h +++ b/include/linux/signal.h @@ -14,10 +14,12 @@ * * SA_INTERRUPT is also used by the irq handling routines. * SA_SHIRQ is for shared interrupt support on PCI and EISA. + * SA_PROBEIRQ is set by callers when they expect sharing mismatches to occur */ -#define SA_PROBE SA_ONESHOT #define SA_SAMPLE_RANDOM SA_RESTART #define SA_SHIRQ 0x04000000 +#define SA_PROBEIRQ 0x08000000 + /* * As above, these correspond to the IORESOURCE_IRQ_* defines in * linux/ioport.h to select the interrupt line behaviour. When diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index ac766ad573e8..1279e3499534 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -246,8 +246,10 @@ int setup_irq(unsigned int irq, struct irqaction * new) mismatch: spin_unlock_irqrestore(&desc->lock, flags); - printk(KERN_ERR "%s: irq handler mismatch\n", __FUNCTION__); - dump_stack(); + if (!(new->flags & SA_PROBEIRQ)) { + printk(KERN_ERR "%s: irq handler mismatch\n", __FUNCTION__); + dump_stack(); + } return -EBUSY; } -- cgit v1.2.3 From 3363fbdd6fb4992ffe6c17c0dd7388ffa22d99e6 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Thu, 27 Apr 2006 18:40:12 -0700 Subject: [PATCH] s390: futex atomic operations Add support for atomic futex operations. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-s390/futex.h | 123 +++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 119 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/asm-s390/futex.h b/include/asm-s390/futex.h index 6a332a9f099c..40c25e166a9b 100644 --- a/include/asm-s390/futex.h +++ b/include/asm-s390/futex.h @@ -1,6 +1,121 @@ -#ifndef _ASM_FUTEX_H -#define _ASM_FUTEX_H +#ifndef _ASM_S390_FUTEX_H +#define _ASM_S390_FUTEX_H -#include +#ifdef __KERNEL__ -#endif +#include +#include +#include + +#ifndef __s390x__ +#define __futex_atomic_fixup \ + ".section __ex_table,\"a\"\n" \ + " .align 4\n" \ + " .long 0b,2b,1b,2b\n" \ + ".previous" +#else /* __s390x__ */ +#define __futex_atomic_fixup \ + ".section __ex_table,\"a\"\n" \ + " .align 8\n" \ + " .quad 0b,2b,1b,2b\n" \ + ".previous" +#endif /* __s390x__ */ + +#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg) \ + asm volatile(" l %1,0(%6)\n" \ + "0: " insn \ + " cs %1,%2,0(%6)\n" \ + "1: jl 0b\n" \ + " lhi %0,0\n" \ + "2:\n" \ + __futex_atomic_fixup \ + : "=d" (ret), "=&d" (oldval), "=&d" (newval), \ + "=m" (*uaddr) \ + : "0" (-EFAULT), "d" (oparg), "a" (uaddr), \ + "m" (*uaddr) : "cc" ); + +static inline int futex_atomic_op_inuser (int encoded_op, int __user *uaddr) +{ + int op = (encoded_op >> 28) & 7; + int cmp = (encoded_op >> 24) & 15; + int oparg = (encoded_op << 8) >> 20; + int cmparg = (encoded_op << 20) >> 20; + int oldval = 0, newval, ret; + if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) + oparg = 1 << oparg; + + if (! access_ok (VERIFY_WRITE, uaddr, sizeof(int))) + return -EFAULT; + + inc_preempt_count(); + + switch (op) { + case FUTEX_OP_SET: + __futex_atomic_op("lr %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_ADD: + __futex_atomic_op("lr %2,%1\nar %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_OR: + __futex_atomic_op("lr %2,%1\nor %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_ANDN: + __futex_atomic_op("lr %2,%1\nnr %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + case FUTEX_OP_XOR: + __futex_atomic_op("lr %2,%1\nxr %2,%5\n", + ret, oldval, newval, uaddr, oparg); + break; + default: + ret = -ENOSYS; + } + + dec_preempt_count(); + + if (!ret) { + switch (cmp) { + case FUTEX_OP_CMP_EQ: ret = (oldval == cmparg); break; + case FUTEX_OP_CMP_NE: ret = (oldval != cmparg); break; + case FUTEX_OP_CMP_LT: ret = (oldval < cmparg); break; + case FUTEX_OP_CMP_GE: ret = (oldval >= cmparg); break; + case FUTEX_OP_CMP_LE: ret = (oldval <= cmparg); break; + case FUTEX_OP_CMP_GT: ret = (oldval > cmparg); break; + default: ret = -ENOSYS; + } + } + return ret; +} + +static inline int +futex_atomic_cmpxchg_inatomic(int __user *uaddr, int oldval, int newval) +{ + int ret; + + if (! access_ok (VERIFY_WRITE, uaddr, sizeof(int))) + return -EFAULT; + asm volatile(" cs %1,%4,0(%5)\n" + "0: lr %0,%1\n" + "1:\n" +#ifndef __s390x__ + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 0b,1b\n" + ".previous" +#else /* __s390x__ */ + ".section __ex_table,\"a\"\n" + " .align 8\n" + " .quad 0b,1b\n" + ".previous" +#endif /* __s390x__ */ + : "=d" (ret), "+d" (oldval), "=m" (*uaddr) + : "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr) + : "cc", "memory" ); + return oldval; +} + +#endif /* __KERNEL__ */ +#endif /* _ASM_S390_FUTEX_H */ -- cgit v1.2.3 From 58268b97f679108d32a882a7fc029585da801975 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 27 Apr 2006 18:40:24 -0700 Subject: [PATCH] s390: add read_mostly optimization Add a read_mostly section and define __read_mostly to prevent cache line pollution due to writes for mostly read variables. In addition fix the incorrect alignment of the cache_line_aligned data section. s390 has a cacheline size of 256 bytes. Signed-off-by: Christian Borntraeger Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/kernel/vmlinux.lds.S | 4 +++- include/asm-s390/cache.h | 2 ++ 2 files changed, 5 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index 9289face3027..9f34bb54c051 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -58,9 +58,11 @@ SECTIONS . = ALIGN(4096); .data.page_aligned : { *(.data.idt) } - . = ALIGN(32); + . = ALIGN(256); .data.cacheline_aligned : { *(.data.cacheline_aligned) } + . = ALIGN(256); + .data.read_mostly : { *(.data.read_mostly) } _edata = .; /* End of data section */ . = ALIGN(8192); /* init_task */ diff --git a/include/asm-s390/cache.h b/include/asm-s390/cache.h index e20cdd9074db..cdf431b061bb 100644 --- a/include/asm-s390/cache.h +++ b/include/asm-s390/cache.h @@ -16,4 +16,6 @@ #define ARCH_KMALLOC_MINALIGN 8 +#define __read_mostly __attribute__((__section__(".data.read_mostly"))) + #endif -- cgit v1.2.3 From 03054d51a70e8c273df5d9bc31fea6c843eaa1c3 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sat, 29 Apr 2006 09:51:06 +1000 Subject: [PATCH] powerpc: Add cputable entry for POWER6 Add a cputable entry for the POWER6 processor. The SIHV and SIPR bits in the mmcra have moved in POWER6, so disable support for that until oprofile is fixed. Also tell firmware that we know about POWER6. Signed-off-by: Anton Blanchard Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/cputable.c | 16 ++++++++++++++++ arch/powerpc/kernel/prom_init.c | 1 + include/asm-powerpc/cputable.h | 14 ++++++++++---- 3 files changed, 27 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c index 39e348a3ade2..3f7182db9ed5 100644 --- a/arch/powerpc/kernel/cputable.c +++ b/arch/powerpc/kernel/cputable.c @@ -57,6 +57,8 @@ extern void __setup_cpu_ppc970(unsigned long offset, struct cpu_spec* spec); PPC_FEATURE_SMT | PPC_FEATURE_ICACHE_SNOOP) #define COMMON_USER_POWER5_PLUS (COMMON_USER_PPC64 | PPC_FEATURE_POWER5_PLUS|\ PPC_FEATURE_SMT | PPC_FEATURE_ICACHE_SNOOP) +#define COMMON_USER_POWER6 (COMMON_USER_PPC64 | PPC_FEATURE_ARCH_2_05 |\ + PPC_FEATURE_SMT | PPC_FEATURE_ICACHE_SNOOP) #define COMMON_USER_BOOKE (PPC_FEATURE_32 | PPC_FEATURE_HAS_MMU | \ PPC_FEATURE_BOOKE) @@ -263,6 +265,20 @@ struct cpu_spec cpu_specs[] = { .oprofile_type = PPC_OPROFILE_POWER4, .platform = "power5+", }, + { /* Power6 */ + .pvr_mask = 0xffff0000, + .pvr_value = 0x003e0000, + .cpu_name = "POWER6", + .cpu_features = CPU_FTRS_POWER6, + .cpu_user_features = COMMON_USER_POWER6, + .icache_bsize = 128, + .dcache_bsize = 128, + .num_pmcs = 6, + .cpu_setup = __setup_cpu_power4, + .oprofile_cpu_type = "ppc64/power6", + .oprofile_type = PPC_OPROFILE_POWER4, + .platform = "power6", + }, { /* Cell Broadband Engine */ .pvr_mask = 0xffff0000, .pvr_value = 0x00700000, diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index 00ad7f981a0d..078fb5533541 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -686,6 +686,7 @@ static void __init early_cmdline_parse(void) */ static unsigned char ibm_architecture_vec[] = { W(0xfffe0000), W(0x003a0000), /* POWER5/POWER5+ */ + W(0xffff0000), W(0x003e0000), /* POWER6 */ W(0xfffffffe), W(0x0f000001), /* all 2.04-compliant and earlier */ 5 - 1, /* 5 option vectors */ diff --git a/include/asm-powerpc/cputable.h b/include/asm-powerpc/cputable.h index 4321483cce51..9fcf0162d859 100644 --- a/include/asm-powerpc/cputable.h +++ b/include/asm-powerpc/cputable.h @@ -22,6 +22,7 @@ #define PPC_FEATURE_BOOKE 0x00008000 #define PPC_FEATURE_SMT 0x00004000 #define PPC_FEATURE_ICACHE_SNOOP 0x00002000 +#define PPC_FEATURE_ARCH_2_05 0x00001000 #ifdef __KERNEL__ #ifndef __ASSEMBLY__ @@ -320,6 +321,11 @@ extern void do_cpu_ftr_fixups(unsigned long offset); CPU_FTR_MMCRA | CPU_FTR_SMT | \ CPU_FTR_COHERENT_ICACHE | CPU_FTR_LOCKLESS_TLBIE | \ CPU_FTR_MMCRA_SIHV | CPU_FTR_PURR) +#define CPU_FTRS_POWER6 (CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \ + CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \ + CPU_FTR_MMCRA | CPU_FTR_SMT | \ + CPU_FTR_COHERENT_ICACHE | CPU_FTR_LOCKLESS_TLBIE | \ + CPU_FTR_PURR | CPU_FTR_CI_LARGE_PAGE) #define CPU_FTRS_CELL (CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \ CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \ CPU_FTR_ALTIVEC_COMP | CPU_FTR_MMCRA | CPU_FTR_SMT | \ @@ -331,8 +337,8 @@ extern void do_cpu_ftr_fixups(unsigned long offset); #ifdef __powerpc64__ #define CPU_FTRS_POSSIBLE \ (CPU_FTRS_POWER3 | CPU_FTRS_RS64 | CPU_FTRS_POWER4 | \ - CPU_FTRS_PPC970 | CPU_FTRS_POWER5 | CPU_FTRS_CELL | \ - CPU_FTR_CI_LARGE_PAGE) + CPU_FTRS_PPC970 | CPU_FTRS_POWER5 | CPU_FTRS_POWER6 | \ + CPU_FTRS_CELL | CPU_FTR_CI_LARGE_PAGE) #else enum { CPU_FTRS_POSSIBLE = @@ -376,8 +382,8 @@ enum { #ifdef __powerpc64__ #define CPU_FTRS_ALWAYS \ (CPU_FTRS_POWER3 & CPU_FTRS_RS64 & CPU_FTRS_POWER4 & \ - CPU_FTRS_PPC970 & CPU_FTRS_POWER5 & CPU_FTRS_CELL & \ - CPU_FTRS_POSSIBLE) + CPU_FTRS_PPC970 & CPU_FTRS_POWER5 & CPU_FTRS_POWER6 & \ + CPU_FTRS_CELL & CPU_FTRS_POSSIBLE) #else enum { CPU_FTRS_ALWAYS = -- cgit v1.2.3 From 08791e5cf62b6952ca32106aebb79b6066005de4 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sat, 29 Apr 2006 01:13:21 -0400 Subject: Input: ressurect EVIOCGREP and EVIOCSREP While writing to an event device allows to set repeat rate for an individual input device there is no way to retrieve current settings so we need to ressurect EVIOCGREP. Also ressurect EVIOCSREP so we have a symmetrical interface. Signed-off-by: Dmitry Torokhov --- drivers/input/evdev.c | 21 +++++++++++++++++++++ include/linux/input.h | 2 ++ 2 files changed, 23 insertions(+) (limited to 'include') diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c index a34e3d91d9ed..ba325f16d077 100644 --- a/drivers/input/evdev.c +++ b/drivers/input/evdev.c @@ -403,6 +403,27 @@ static long evdev_ioctl_handler(struct file *file, unsigned int cmd, case EVIOCGID: if (copy_to_user(p, &dev->id, sizeof(struct input_id))) return -EFAULT; + return 0; + + case EVIOCGREP: + if (!test_bit(EV_REP, dev->evbit)) + return -ENOSYS; + if (put_user(dev->rep[REP_DELAY], ip)) + return -EFAULT; + if (put_user(dev->rep[REP_PERIOD], ip + 1)) + return -EFAULT; + return 0; + + case EVIOCSREP: + if (!test_bit(EV_REP, dev->evbit)) + return -ENOSYS; + if (get_user(u, ip)) + return -EFAULT; + if (get_user(v, ip + 1)) + return -EFAULT; + + input_event(dev, EV_REP, REP_DELAY, u); + input_event(dev, EV_REP, REP_PERIOD, v); return 0; diff --git a/include/linux/input.h b/include/linux/input.h index 8298b4bf5a07..50e338d2ffda 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -56,6 +56,8 @@ struct input_absinfo { #define EVIOCGVERSION _IOR('E', 0x01, int) /* get driver version */ #define EVIOCGID _IOR('E', 0x02, struct input_id) /* get device ID */ +#define EVIOCGREP _IOR('E', 0x03, int[2]) /* get repeat settings */ +#define EVIOCSREP _IOW('E', 0x03, int[2]) /* set repeat settings */ #define EVIOCGKEYCODE _IOR('E', 0x04, int[2]) /* get keycode */ #define EVIOCSKEYCODE _IOW('E', 0x04, int[2]) /* set keycode */ -- cgit v1.2.3 From 543f2a3382bd7abd7380903518c61f00c7c87577 Mon Sep 17 00:00:00 2001 From: Chuck Ebbert <76306.1226@compuserve.com> Date: Sat, 29 Apr 2006 14:07:49 -0400 Subject: [PATCH] i386: fix broken FP exception handling The FXSAVE information leak patch introduced a bug in FP exception handling: it clears FP exceptions only when there are already none outstanding. Mikael Pettersson reported that causes problems with the Erlang runtime and has tested this fix. Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com> Acked-by: Mikael Pettersson Signed-off-by: Linus Torvalds --- include/asm-i386/i387.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/asm-i386/i387.h b/include/asm-i386/i387.h index 7b1f01191e70..bc1d6edae1ed 100644 --- a/include/asm-i386/i387.h +++ b/include/asm-i386/i387.h @@ -58,13 +58,13 @@ static inline void __save_init_fpu( struct task_struct *tsk ) alternative_input( "fnsave %[fx] ;fwait;" GENERIC_NOP8 GENERIC_NOP4, "fxsave %[fx]\n" - "bt $7,%[fsw] ; jc 1f ; fnclex\n1:", + "bt $7,%[fsw] ; jnc 1f ; fnclex\n1:", X86_FEATURE_FXSR, [fx] "m" (tsk->thread.i387.fxsave), [fsw] "m" (tsk->thread.i387.fxsave.swd) : "memory"); /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception is pending. Clear the x87 state here by setting it to fixed - values. __per_cpu_offset[0] is a random variable that should be in L1 */ + values. safe_address is a random variable that should be in L1 */ alternative_input( GENERIC_NOP8 GENERIC_NOP2, "emms\n\t" /* clear stack tags */ -- cgit v1.2.3 From da753beaeb1446aa87bcca7e8a0026633a8914f0 Mon Sep 17 00:00:00 2001 From: Akinobu Mita Date: Fri, 28 Apr 2006 15:21:23 -0700 Subject: [NET]: use hlist_unhashed() Use hlist_unhashed() rather than accessing inside data structure. Signed-off-by: Akinobu Mita Signed-off-by: Andrew Morton Signed-off-by: David S. Miller --- include/linux/list.h | 2 +- include/net/inet_timewait_sock.h | 2 +- include/net/sock.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/list.h b/include/linux/list.h index 67258b47e9ca..76f05718342c 100644 --- a/include/linux/list.h +++ b/include/linux/list.h @@ -619,7 +619,7 @@ static inline void hlist_del_rcu(struct hlist_node *n) static inline void hlist_del_init(struct hlist_node *n) { - if (n->pprev) { + if (!hlist_unhashed(n)) { __hlist_del(n); INIT_HLIST_NODE(n); } diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h index 1da294c47522..e837f98fdb50 100644 --- a/include/net/inet_timewait_sock.h +++ b/include/net/inet_timewait_sock.h @@ -150,7 +150,7 @@ static inline void inet_twsk_add_bind_node(struct inet_timewait_sock *tw, static inline int inet_twsk_dead_hashed(const struct inet_timewait_sock *tw) { - return tw->tw_death_node.pprev != NULL; + return !hlist_unhashed(&tw->tw_death_node); } static inline void inet_twsk_dead_node_init(struct inet_timewait_sock *tw) diff --git a/include/net/sock.h b/include/net/sock.h index ff8b0dad7b0f..c9fad6fb629b 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -279,7 +279,7 @@ static inline int sk_unhashed(const struct sock *sk) static inline int sk_hashed(const struct sock *sk) { - return sk->sk_node.pprev != NULL; + return !sk_unhashed(sk); } static __inline__ void sk_node_init(struct hlist_node *node) -- cgit v1.2.3 From 68ac64cd3fd89fdaa091701f6ab98a9065e9b1b5 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 30 Apr 2006 11:13:50 +0100 Subject: [SERIAL] Clean up serial locking when obtaining a reference to a port The locking for the uart_port is over complicated, and can be simplified if we introduce a flag to indicate that a port is "dead" and will be removed. This also helps the validator because it removes a case of non-nested unlock ordering. Signed-off-by: Russell King Signed-off-by: Ingo Molnar Signed-off-by: Andrew Morton --- drivers/serial/serial_core.c | 114 +++++++++++++++++++++++-------------------- include/linux/serial_core.h | 1 + 2 files changed, 61 insertions(+), 54 deletions(-) (limited to 'include') diff --git a/drivers/serial/serial_core.c b/drivers/serial/serial_core.c index fcd7744c4253..aeb8153ccf24 100644 --- a/drivers/serial/serial_core.c +++ b/drivers/serial/serial_core.c @@ -1500,20 +1500,18 @@ uart_block_til_ready(struct file *filp, struct uart_state *state) static struct uart_state *uart_get(struct uart_driver *drv, int line) { struct uart_state *state; + int ret = 0; - mutex_lock(&port_mutex); state = drv->state + line; if (mutex_lock_interruptible(&state->mutex)) { - state = ERR_PTR(-ERESTARTSYS); - goto out; + ret = -ERESTARTSYS; + goto err; } state->count++; - if (!state->port) { - state->count--; - mutex_unlock(&state->mutex); - state = ERR_PTR(-ENXIO); - goto out; + if (!state->port || state->port->flags & UPF_DEAD) { + ret = -ENXIO; + goto err_unlock; } if (!state->info) { @@ -1531,15 +1529,17 @@ static struct uart_state *uart_get(struct uart_driver *drv, int line) tasklet_init(&state->info->tlet, uart_tasklet_action, (unsigned long)state); } else { - state->count--; - mutex_unlock(&state->mutex); - state = ERR_PTR(-ENOMEM); + ret = -ENOMEM; + goto err_unlock; } } - - out: - mutex_unlock(&port_mutex); return state; + + err_unlock: + state->count--; + mutex_unlock(&state->mutex); + err: + return ERR_PTR(ret); } /* @@ -2085,45 +2085,6 @@ uart_configure_port(struct uart_driver *drv, struct uart_state *state, } } -/* - * This reverses the effects of uart_configure_port, hanging up the - * port before removal. - */ -static void -uart_unconfigure_port(struct uart_driver *drv, struct uart_state *state) -{ - struct uart_port *port = state->port; - struct uart_info *info = state->info; - - if (info && info->tty) - tty_vhangup(info->tty); - - mutex_lock(&state->mutex); - - state->info = NULL; - - /* - * Free the port IO and memory resources, if any. - */ - if (port->type != PORT_UNKNOWN) - port->ops->release_port(port); - - /* - * Indicate that there isn't a port here anymore. - */ - port->type = PORT_UNKNOWN; - - /* - * Kill the tasklet, and free resources. - */ - if (info) { - tasklet_kill(&info->tlet); - kfree(info); - } - - mutex_unlock(&state->mutex); -} - static struct tty_operations uart_ops = { .open = uart_open, .close = uart_close, @@ -2270,6 +2231,7 @@ int uart_add_one_port(struct uart_driver *drv, struct uart_port *port) state = drv->state + port->line; mutex_lock(&port_mutex); + mutex_lock(&state->mutex); if (state->port) { ret = -EINVAL; goto out; @@ -2304,7 +2266,13 @@ int uart_add_one_port(struct uart_driver *drv, struct uart_port *port) port->cons && !(port->cons->flags & CON_ENABLED)) register_console(port->cons); + /* + * Ensure UPF_DEAD is not set. + */ + port->flags &= ~UPF_DEAD; + out: + mutex_unlock(&state->mutex); mutex_unlock(&port_mutex); return ret; @@ -2322,6 +2290,7 @@ int uart_add_one_port(struct uart_driver *drv, struct uart_port *port) int uart_remove_one_port(struct uart_driver *drv, struct uart_port *port) { struct uart_state *state = drv->state + port->line; + struct uart_info *info; BUG_ON(in_interrupt()); @@ -2331,12 +2300,49 @@ int uart_remove_one_port(struct uart_driver *drv, struct uart_port *port) mutex_lock(&port_mutex); + /* + * Mark the port "dead" - this prevents any opens from + * succeeding while we shut down the port. + */ + mutex_lock(&state->mutex); + port->flags |= UPF_DEAD; + mutex_unlock(&state->mutex); + /* * Remove the devices from devfs */ tty_unregister_device(drv->tty_driver, port->line); - uart_unconfigure_port(drv, state); + info = state->info; + if (info && info->tty) + tty_vhangup(info->tty); + + /* + * All users of this port should now be disconnected from + * this driver, and the port shut down. We should be the + * only thread fiddling with this port from now on. + */ + state->info = NULL; + + /* + * Free the port IO and memory resources, if any. + */ + if (port->type != PORT_UNKNOWN) + port->ops->release_port(port); + + /* + * Indicate that there isn't a port here anymore. + */ + port->type = PORT_UNKNOWN; + + /* + * Kill the tasklet, and free resources. + */ + if (info) { + tasklet_kill(&info->tlet); + kfree(info); + } + state->port = NULL; mutex_unlock(&port_mutex); diff --git a/include/linux/serial_core.h b/include/linux/serial_core.h index c32e60e79dea..bd14858121ea 100644 --- a/include/linux/serial_core.h +++ b/include/linux/serial_core.h @@ -254,6 +254,7 @@ struct uart_port { #define UPF_CONS_FLOW ((__force upf_t) (1 << 23)) #define UPF_SHARE_IRQ ((__force upf_t) (1 << 24)) #define UPF_BOOT_AUTOCONF ((__force upf_t) (1 << 28)) +#define UPF_DEAD ((__force upf_t) (1 << 30)) #define UPF_IOREMAP ((__force upf_t) (1 << 31)) #define UPF_CHANGE_MASK ((__force upf_t) (0x17fff)) -- cgit v1.2.3 From cd95842ca0ffb0e3df3b459832a60f9f4544ed9e Mon Sep 17 00:00:00 2001 From: Markus Gutschke Date: Sun, 30 Apr 2006 15:34:29 +0100 Subject: [ARM] 3486/1: Mark memory as clobbered by the ARM _syscallX() macros Patch from Markus Gutschke In order to prevent gcc from making incorrect optimizations, all asm() statements that define system calls should report memory as clobbered. Recent versions of the headers for i386 have been changed accordingly, but the ARM headers are still defective. This patch fixes the bug tracked at http://bugzilla.kernel.org/show_bug.cgi?id=6205 Signed-off-by: Markus Gutschke Signed-off-by: Russell King --- drivers/input/touchscreen/corgi_ts.c | 2 +- include/asm-arm/unistd.h | 21 ++++++++++++++------- 2 files changed, 15 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/drivers/input/touchscreen/corgi_ts.c b/drivers/input/touchscreen/corgi_ts.c index 1042987856f7..5013703db0e6 100644 --- a/drivers/input/touchscreen/corgi_ts.c +++ b/drivers/input/touchscreen/corgi_ts.c @@ -17,7 +17,7 @@ #include #include #include -#include +//#include #include #include diff --git a/include/asm-arm/unistd.h b/include/asm-arm/unistd.h index ee8dfea549bc..26f2f4828e03 100644 --- a/include/asm-arm/unistd.h +++ b/include/asm-arm/unistd.h @@ -410,7 +410,8 @@ type name(void) { \ __asm__ __volatile__ ( \ __syscall(name) \ : "=r" (__res_r0) \ - : __SYS_REG_LIST() ); \ + : __SYS_REG_LIST() \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } @@ -424,7 +425,8 @@ type name(type1 arg1) { \ __asm__ __volatile__ ( \ __syscall(name) \ : "=r" (__res_r0) \ - : __SYS_REG_LIST( "0" (__r0) ) ); \ + : __SYS_REG_LIST( "0" (__r0) ) \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } @@ -439,7 +441,8 @@ type name(type1 arg1,type2 arg2) { \ __asm__ __volatile__ ( \ __syscall(name) \ : "=r" (__res_r0) \ - : __SYS_REG_LIST( "0" (__r0), "r" (__r1) ) ); \ + : __SYS_REG_LIST( "0" (__r0), "r" (__r1) ) \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } @@ -456,7 +459,8 @@ type name(type1 arg1,type2 arg2,type3 arg3) { \ __asm__ __volatile__ ( \ __syscall(name) \ : "=r" (__res_r0) \ - : __SYS_REG_LIST( "0" (__r0), "r" (__r1), "r" (__r2) ) ); \ + : __SYS_REG_LIST( "0" (__r0), "r" (__r1), "r" (__r2) ) \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } @@ -474,7 +478,8 @@ type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4) { \ __asm__ __volatile__ ( \ __syscall(name) \ : "=r" (__res_r0) \ - : __SYS_REG_LIST( "0" (__r0), "r" (__r1), "r" (__r2), "r" (__r3) ) ); \ + : __SYS_REG_LIST( "0" (__r0), "r" (__r1), "r" (__r2), "r" (__r3) ) \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } @@ -494,7 +499,8 @@ type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4, type5 arg5) { \ __syscall(name) \ : "=r" (__res_r0) \ : __SYS_REG_LIST( "0" (__r0), "r" (__r1), "r" (__r2), \ - "r" (__r3), "r" (__r4) ) ); \ + "r" (__r3), "r" (__r4) ) \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } @@ -514,7 +520,8 @@ type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4, type5 arg5, type6 arg6 __syscall(name) \ : "=r" (__res_r0) \ : __SYS_REG_LIST( "0" (__r0), "r" (__r1), "r" (__r2), \ - "r" (__r3), "r" (__r4), "r" (__r5) ) ); \ + "r" (__r3), "r" (__r4), "r" (__r5) ) \ + : "memory" ); \ __res = __res_r0; \ __syscall_return(type,__res); \ } -- cgit v1.2.3 From 76bbb00288e569e7bd9ec18f45e4f814352260dd Mon Sep 17 00:00:00 2001 From: Deepak Saxena Date: Sun, 30 Apr 2006 15:34:29 +0100 Subject: [ARM] 3487/1: IXP4xx: Support non-PCI systems Patch from Deepak Saxena This patch allows for the addition of IXP4xx systems that do not make use of the PCI interface by moving the CONFIG_PCI symbol selection to be platform-specific instead of for all of IXP4xx. If at least one machine with PCI support is built, the PCI code will be compiled in, but when building !PCI, this will drastically shrink the kernel size. Signed-off-by: Deepak Saxena Signed-off-by: Russell King --- arch/arm/Kconfig | 2 -- arch/arm/mach-ixp4xx/Kconfig | 15 ++++++++++++++- arch/arm/mach-ixp4xx/Makefile | 3 ++- include/asm-arm/arch-ixp4xx/io.h | 7 +++++++ include/asm-arm/arch-ixp4xx/memory.h | 2 +- 5 files changed, 24 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 1dbf6ddb300d..08b7cc900cae 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -150,8 +150,6 @@ config ARCH_IOP3XX config ARCH_IXP4XX bool "IXP4xx-based" - select DMABOUNCE - select PCI help Support for Intel's IXP4XX (XScale) family of processors. diff --git a/arch/arm/mach-ixp4xx/Kconfig b/arch/arm/mach-ixp4xx/Kconfig index 5bf50a2a737d..2a39f9e481ad 100644 --- a/arch/arm/mach-ixp4xx/Kconfig +++ b/arch/arm/mach-ixp4xx/Kconfig @@ -11,6 +11,7 @@ comment "IXP4xx Platforms" config MACH_NSLU2 bool prompt "Linksys NSLU2" + select PCI help Say 'Y' here if you want your kernel to support Linksys's NSLU2 NAS device. For more information on this platform, @@ -18,6 +19,7 @@ config MACH_NSLU2 config ARCH_AVILA bool "Avila" + select PCI help Say 'Y' here if you want your kernel to support the Gateworks Avila Network Platform. For more information on this platform, @@ -25,6 +27,7 @@ config ARCH_AVILA config ARCH_ADI_COYOTE bool "Coyote" + select PCI help Say 'Y' here if you want your kernel to support the ADI Engineering Coyote Gateway Reference Platform. For more @@ -32,6 +35,7 @@ config ARCH_ADI_COYOTE config ARCH_IXDP425 bool "IXDP425" + select PCI help Say 'Y' here if you want your kernel to support Intel's IXDP425 Development Platform (Also known as Richfield). @@ -39,6 +43,7 @@ config ARCH_IXDP425 config MACH_IXDPG425 bool "IXDPG425" + select PCI help Say 'Y' here if you want your kernel to support Intel's IXDPG425 Development Platform (Also known as Montajade). @@ -46,6 +51,7 @@ config MACH_IXDPG425 config MACH_IXDP465 bool "IXDP465" + select PCI help Say 'Y' here if you want your kernel to support Intel's IXDP465 Development Platform (Also known as BMP). @@ -72,6 +78,7 @@ config ARCH_PRPMC1100 config MACH_NAS100D bool prompt "NAS100D" + select PCI help Say 'Y' here if you want your kernel to support Iomega's NAS 100d device. For more information on this platform, @@ -96,6 +103,7 @@ config CPU_IXP46X config MACH_GTWX5715 bool "Gemtek WX5715 (Linksys WRV54G)" depends on ARCH_IXP4XX + select PCI help This board is currently inside the Linksys WRV54G Gateways. @@ -110,11 +118,16 @@ config MACH_GTWX5715 "High Speed" UART is n/c (as far as I can tell) 20 Pin ARM/Xscale JTAG interface on J2 - comment "IXP4xx Options" +config DMABOUNCE + bool + default y + depends on PCI + config IXP4XX_INDIRECT_PCI bool "Use indirect PCI memory access" + depends on PCI help IXP4xx provides two methods of accessing PCI memory space: diff --git a/arch/arm/mach-ixp4xx/Makefile b/arch/arm/mach-ixp4xx/Makefile index 0471044fa179..5a4aaa0e0a09 100644 --- a/arch/arm/mach-ixp4xx/Makefile +++ b/arch/arm/mach-ixp4xx/Makefile @@ -2,8 +2,9 @@ # Makefile for the linux kernel. # -obj-y += common.o common-pci.o +obj-y += common.o +obj-$(CONFIG_PCI) += common-pci.o obj-$(CONFIG_ARCH_IXDP4XX) += ixdp425-pci.o ixdp425-setup.o obj-$(CONFIG_MACH_IXDPG425) += ixdpg425-pci.o coyote-setup.o obj-$(CONFIG_ARCH_ADI_COYOTE) += coyote-pci.o coyote-setup.o diff --git a/include/asm-arm/arch-ixp4xx/io.h b/include/asm-arm/arch-ixp4xx/io.h index 942b622455bc..b59520e56fc7 100644 --- a/include/asm-arm/arch-ixp4xx/io.h +++ b/include/asm-arm/arch-ixp4xx/io.h @@ -260,6 +260,12 @@ out: #endif +#ifndef CONFIG_PCI + +#define __io(v) v + +#else + /* * IXP4xx does not have a transparent cpu -> PCI I/O translation * window. Instead, it has a set of registers that must be tweaked @@ -578,6 +584,7 @@ __ixp4xx_iowrite32_rep(void __iomem *addr, const void *vaddr, u32 count) #define ioport_map(port, nr) ((void __iomem*)(port + PIO_OFFSET)) #define ioport_unmap(addr) +#endif // !CONFIG_PCI #endif // __ASM_ARM_ARCH_IO_H diff --git a/include/asm-arm/arch-ixp4xx/memory.h b/include/asm-arm/arch-ixp4xx/memory.h index ee211d28a3ef..af9667b57ab3 100644 --- a/include/asm-arm/arch-ixp4xx/memory.h +++ b/include/asm-arm/arch-ixp4xx/memory.h @@ -14,7 +14,7 @@ */ #define PHYS_OFFSET UL(0x00000000) -#ifndef __ASSEMBLY__ +#if !defined(__ASSEMBLY__) && defined(CONFIG_PCI) void ixp4xx_adjust_zones(int node, unsigned long *size, unsigned long *holes); -- cgit v1.2.3 From 1241140f5183db38393556832198a3b109bf9085 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Sun, 30 Apr 2006 21:40:13 -0700 Subject: [SPARC64]: Kill __flush_tlb_page() prototype. This function no longer exists. Signed-off-by: David S. Miller --- include/asm-sparc64/tlbflush.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/asm-sparc64/tlbflush.h b/include/asm-sparc64/tlbflush.h index 9ad5d9c51d42..e3a7c453b500 100644 --- a/include/asm-sparc64/tlbflush.h +++ b/include/asm-sparc64/tlbflush.h @@ -22,8 +22,6 @@ extern void flush_tlb_pending(void); /* Local cpu only. */ extern void __flush_tlb_all(void); -extern void __flush_tlb_page(unsigned long context, unsigned long page, unsigned long r); - extern void __flush_tlb_kernel_range(unsigned long start, unsigned long end); #ifndef CONFIG_SMP -- cgit v1.2.3 From 5411be59db80333039386f3b1ccfe5eb9023a916 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 29 Mar 2006 20:23:36 -0500 Subject: [PATCH] drop task argument of audit_syscall_{entry,exit} ... it's always current, and that's a good thing - allows simpler locking. Signed-off-by: Al Viro --- arch/i386/kernel/ptrace.c | 7 +++---- arch/i386/kernel/vm86.c | 2 +- arch/ia64/kernel/ptrace.c | 4 ++-- arch/mips/kernel/ptrace.c | 4 ++-- arch/powerpc/kernel/ptrace.c | 5 ++--- arch/s390/kernel/ptrace.c | 5 ++--- arch/sparc64/kernel/ptrace.c | 5 ++--- arch/um/kernel/ptrace.c | 6 ++---- arch/x86_64/kernel/ptrace.c | 6 +++--- include/linux/audit.h | 8 ++++---- kernel/auditsc.c | 8 ++++---- 11 files changed, 27 insertions(+), 33 deletions(-) (limited to 'include') diff --git a/arch/i386/kernel/ptrace.c b/arch/i386/kernel/ptrace.c index 506462ef36a0..fd7eaf7866e0 100644 --- a/arch/i386/kernel/ptrace.c +++ b/arch/i386/kernel/ptrace.c @@ -671,7 +671,7 @@ int do_syscall_trace(struct pt_regs *regs, int entryexit) if (unlikely(current->audit_context)) { if (entryexit) - audit_syscall_exit(current, AUDITSC_RESULT(regs->eax), + audit_syscall_exit(AUDITSC_RESULT(regs->eax), regs->eax); /* Debug traps, when using PTRACE_SINGLESTEP, must be sent only * on the syscall exit path. Normally, when TIF_SYSCALL_AUDIT is @@ -720,14 +720,13 @@ int do_syscall_trace(struct pt_regs *regs, int entryexit) ret = is_sysemu; out: if (unlikely(current->audit_context) && !entryexit) - audit_syscall_entry(current, AUDIT_ARCH_I386, regs->orig_eax, + audit_syscall_entry(AUDIT_ARCH_I386, regs->orig_eax, regs->ebx, regs->ecx, regs->edx, regs->esi); if (ret == 0) return 0; regs->orig_eax = -1; /* force skip of syscall restarting */ if (unlikely(current->audit_context)) - audit_syscall_exit(current, AUDITSC_RESULT(regs->eax), - regs->eax); + audit_syscall_exit(AUDITSC_RESULT(regs->eax), regs->eax); return 1; } diff --git a/arch/i386/kernel/vm86.c b/arch/i386/kernel/vm86.c index aee14fafd13d..00e0118e717c 100644 --- a/arch/i386/kernel/vm86.c +++ b/arch/i386/kernel/vm86.c @@ -312,7 +312,7 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk /*call audit_syscall_exit since we do not exit via the normal paths */ if (unlikely(current->audit_context)) - audit_syscall_exit(current, AUDITSC_RESULT(eax), eax); + audit_syscall_exit(AUDITSC_RESULT(eax), eax); __asm__ __volatile__( "movl %0,%%esp\n\t" diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 9887c8787e7a..e61e15e28d8b 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -1644,7 +1644,7 @@ syscall_trace_enter (long arg0, long arg1, long arg2, long arg3, arch = AUDIT_ARCH_IA64; } - audit_syscall_entry(current, arch, syscall, arg0, arg1, arg2, arg3); + audit_syscall_entry(arch, syscall, arg0, arg1, arg2, arg3); } } @@ -1662,7 +1662,7 @@ syscall_trace_leave (long arg0, long arg1, long arg2, long arg3, if (success != AUDITSC_SUCCESS) result = -result; - audit_syscall_exit(current, success, result); + audit_syscall_exit(success, result); } if (test_thread_flag(TIF_SYSCALL_TRACE) diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index f3106d0771b0..9b4733c12395 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -483,7 +483,7 @@ static inline int audit_arch(void) asmlinkage void do_syscall_trace(struct pt_regs *regs, int entryexit) { if (unlikely(current->audit_context) && entryexit) - audit_syscall_exit(current, AUDITSC_RESULT(regs->regs[2]), + audit_syscall_exit(AUDITSC_RESULT(regs->regs[2]), regs->regs[2]); if (!(current->ptrace & PT_PTRACED)) @@ -507,7 +507,7 @@ asmlinkage void do_syscall_trace(struct pt_regs *regs, int entryexit) } out: if (unlikely(current->audit_context) && !entryexit) - audit_syscall_entry(current, audit_arch(), regs->regs[2], + audit_syscall_entry(audit_arch(), regs->regs[2], regs->regs[4], regs->regs[5], regs->regs[6], regs->regs[7]); } diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c index bcb83574335b..4a677d1bd4ef 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -538,7 +538,7 @@ void do_syscall_trace_enter(struct pt_regs *regs) do_syscall_trace(); if (unlikely(current->audit_context)) - audit_syscall_entry(current, + audit_syscall_entry( #ifdef CONFIG_PPC32 AUDIT_ARCH_PPC, #else @@ -556,8 +556,7 @@ void do_syscall_trace_leave(struct pt_regs *regs) #endif if (unlikely(current->audit_context)) - audit_syscall_exit(current, - (regs->ccr&0x1000)?AUDITSC_FAILURE:AUDITSC_SUCCESS, + audit_syscall_exit((regs->ccr&0x1000)?AUDITSC_FAILURE:AUDITSC_SUCCESS, regs->result); if ((test_thread_flag(TIF_SYSCALL_TRACE) diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 37dfe33dab73..8f36504075ed 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -734,7 +734,7 @@ asmlinkage void syscall_trace(struct pt_regs *regs, int entryexit) { if (unlikely(current->audit_context) && entryexit) - audit_syscall_exit(current, AUDITSC_RESULT(regs->gprs[2]), regs->gprs[2]); + audit_syscall_exit(AUDITSC_RESULT(regs->gprs[2]), regs->gprs[2]); if (!test_thread_flag(TIF_SYSCALL_TRACE)) goto out; @@ -761,8 +761,7 @@ syscall_trace(struct pt_regs *regs, int entryexit) } out: if (unlikely(current->audit_context) && !entryexit) - audit_syscall_entry(current, - test_thread_flag(TIF_31BIT)?AUDIT_ARCH_S390:AUDIT_ARCH_S390X, + audit_syscall_entry(test_thread_flag(TIF_31BIT)?AUDIT_ARCH_S390:AUDIT_ARCH_S390X, regs->gprs[2], regs->orig_gpr2, regs->gprs[3], regs->gprs[4], regs->gprs[5]); } diff --git a/arch/sparc64/kernel/ptrace.c b/arch/sparc64/kernel/ptrace.c index 49e6dedd027d..d31975e6d6f6 100644 --- a/arch/sparc64/kernel/ptrace.c +++ b/arch/sparc64/kernel/ptrace.c @@ -653,7 +653,7 @@ asmlinkage void syscall_trace(struct pt_regs *regs, int syscall_exit_p) if (unlikely(tstate & (TSTATE_XCARRY | TSTATE_ICARRY))) result = AUDITSC_FAILURE; - audit_syscall_exit(current, result, regs->u_regs[UREG_I0]); + audit_syscall_exit(result, regs->u_regs[UREG_I0]); } if (!(current->ptrace & PT_PTRACED)) @@ -677,8 +677,7 @@ asmlinkage void syscall_trace(struct pt_regs *regs, int syscall_exit_p) out: if (unlikely(current->audit_context) && !syscall_exit_p) - audit_syscall_entry(current, - (test_thread_flag(TIF_32BIT) ? + audit_syscall_entry((test_thread_flag(TIF_32BIT) ? AUDIT_ARCH_SPARC : AUDIT_ARCH_SPARC64), regs->u_regs[UREG_G1], diff --git a/arch/um/kernel/ptrace.c b/arch/um/kernel/ptrace.c index 60d2eda995c1..9a77fb3c269d 100644 --- a/arch/um/kernel/ptrace.c +++ b/arch/um/kernel/ptrace.c @@ -275,15 +275,13 @@ void syscall_trace(union uml_pt_regs *regs, int entryexit) if (unlikely(current->audit_context)) { if (!entryexit) - audit_syscall_entry(current, - HOST_AUDIT_ARCH, + audit_syscall_entry(HOST_AUDIT_ARCH, UPT_SYSCALL_NR(regs), UPT_SYSCALL_ARG1(regs), UPT_SYSCALL_ARG2(regs), UPT_SYSCALL_ARG3(regs), UPT_SYSCALL_ARG4(regs)); - else audit_syscall_exit(current, - AUDITSC_RESULT(UPT_SYSCALL_RET(regs)), + else audit_syscall_exit(AUDITSC_RESULT(UPT_SYSCALL_RET(regs)), UPT_SYSCALL_RET(regs)); } diff --git a/arch/x86_64/kernel/ptrace.c b/arch/x86_64/kernel/ptrace.c index da8e7903d817..2d50024c9f30 100644 --- a/arch/x86_64/kernel/ptrace.c +++ b/arch/x86_64/kernel/ptrace.c @@ -600,12 +600,12 @@ asmlinkage void syscall_trace_enter(struct pt_regs *regs) if (unlikely(current->audit_context)) { if (test_thread_flag(TIF_IA32)) { - audit_syscall_entry(current, AUDIT_ARCH_I386, + audit_syscall_entry(AUDIT_ARCH_I386, regs->orig_rax, regs->rbx, regs->rcx, regs->rdx, regs->rsi); } else { - audit_syscall_entry(current, AUDIT_ARCH_X86_64, + audit_syscall_entry(AUDIT_ARCH_X86_64, regs->orig_rax, regs->rdi, regs->rsi, regs->rdx, regs->r10); @@ -616,7 +616,7 @@ asmlinkage void syscall_trace_enter(struct pt_regs *regs) asmlinkage void syscall_trace_leave(struct pt_regs *regs) { if (unlikely(current->audit_context)) - audit_syscall_exit(current, AUDITSC_RESULT(regs->rax), regs->rax); + audit_syscall_exit(AUDITSC_RESULT(regs->rax), regs->rax); if ((test_thread_flag(TIF_SYSCALL_TRACE) || test_thread_flag(TIF_SINGLESTEP)) diff --git a/include/linux/audit.h b/include/linux/audit.h index 1c47c59058c1..39fef6ebb854 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -287,10 +287,10 @@ struct netlink_skb_parms; /* Public API */ extern int audit_alloc(struct task_struct *task); extern void audit_free(struct task_struct *task); -extern void audit_syscall_entry(struct task_struct *task, int arch, +extern void audit_syscall_entry(int arch, int major, unsigned long a0, unsigned long a1, unsigned long a2, unsigned long a3); -extern void audit_syscall_exit(struct task_struct *task, int failed, long return_code); +extern void audit_syscall_exit(int failed, long return_code); extern void audit_getname(const char *name); extern void audit_putname(const char *name); extern void __audit_inode(const char *name, const struct inode *inode, unsigned flags); @@ -323,8 +323,8 @@ extern int audit_set_macxattr(const char *name); #else #define audit_alloc(t) ({ 0; }) #define audit_free(t) do { ; } while (0) -#define audit_syscall_entry(t,ta,a,b,c,d,e) do { ; } while (0) -#define audit_syscall_exit(t,f,r) do { ; } while (0) +#define audit_syscall_entry(ta,a,b,c,d,e) do { ; } while (0) +#define audit_syscall_exit(f,r) do { ; } while (0) #define audit_getname(n) do { ; } while (0) #define audit_putname(n) do { ; } while (0) #define __audit_inode(n,i,f) do { ; } while (0) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index ba0ec1ba6698..7ed82b088e4b 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -736,10 +736,11 @@ void audit_free(struct task_struct *tsk) * will only be written if another part of the kernel requests that it * be written). */ -void audit_syscall_entry(struct task_struct *tsk, int arch, int major, +void audit_syscall_entry(int arch, int major, unsigned long a1, unsigned long a2, unsigned long a3, unsigned long a4) { + struct task_struct *tsk = current; struct audit_context *context = tsk->audit_context; enum audit_state state; @@ -817,12 +818,11 @@ void audit_syscall_entry(struct task_struct *tsk, int arch, int major, * message), then write out the syscall information. In call cases, * free the names stored from getname(). */ -void audit_syscall_exit(struct task_struct *tsk, int valid, long return_code) +void audit_syscall_exit(int valid, long return_code) { + struct task_struct *tsk = current; struct audit_context *context; - /* tsk == current */ - get_task_struct(tsk); task_lock(tsk); context = audit_get_context(tsk, valid, return_code); -- cgit v1.2.3 From 376bd9cb357ec945ac893feaeb63af7370a6e70b Mon Sep 17 00:00:00 2001 From: Darrel Goeddel Date: Fri, 24 Feb 2006 15:44:05 -0600 Subject: [PATCH] support for context based audit filtering The following patch provides selinux interfaces that will allow the audit system to perform filtering based on the process context (user, role, type, sensitivity, and clearance). These interfaces will allow the selinux module to perform efficient matches based on lower level selinux constructs, rather than relying on context retrievals and string comparisons within the audit module. It also allows for dominance checks on the mls portion of the contexts that are impossible with only string comparisons. Signed-off-by: Darrel Goeddel Signed-off-by: Al Viro --- include/linux/audit.h | 5 + include/linux/selinux.h | 112 ++++++++++++++++++++ security/selinux/Makefile | 2 +- security/selinux/avc.c | 13 +-- security/selinux/exports.c | 28 +++++ security/selinux/ss/mls.c | 30 +++++- security/selinux/ss/mls.h | 4 +- security/selinux/ss/services.c | 235 ++++++++++++++++++++++++++++++++++++++++- 8 files changed, 419 insertions(+), 10 deletions(-) create mode 100644 include/linux/selinux.h create mode 100644 security/selinux/exports.c (limited to 'include') diff --git a/include/linux/audit.h b/include/linux/audit.h index 39fef6ebb854..740f950397b7 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -145,6 +145,11 @@ #define AUDIT_PERS 10 #define AUDIT_ARCH 11 #define AUDIT_MSGTYPE 12 +#define AUDIT_SE_USER 13 /* security label user */ +#define AUDIT_SE_ROLE 14 /* security label role */ +#define AUDIT_SE_TYPE 15 /* security label type */ +#define AUDIT_SE_SEN 16 /* security label sensitivity label */ +#define AUDIT_SE_CLR 17 /* security label clearance label */ /* These are ONLY useful when checking * at syscall exit time (AUDIT_AT_EXIT). */ diff --git a/include/linux/selinux.h b/include/linux/selinux.h new file mode 100644 index 000000000000..9d684b1728b0 --- /dev/null +++ b/include/linux/selinux.h @@ -0,0 +1,112 @@ +/* + * SELinux services exported to the rest of the kernel. + * + * Author: James Morris + * + * Copyright (C) 2005 Red Hat, Inc., James Morris + * Copyright (C) 2006 Trusted Computer Solutions, Inc. + * + * 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. + */ +#ifndef _LINUX_SELINUX_H +#define _LINUX_SELINUX_H + +struct selinux_audit_rule; +struct audit_context; + +#ifdef CONFIG_SECURITY_SELINUX + +/** + * selinux_audit_rule_init - alloc/init an selinux audit rule structure. + * @field: the field this rule refers to + * @op: the operater the rule uses + * @rulestr: the text "target" of the rule + * @rule: pointer to the new rule structure returned via this + * + * Returns 0 if successful, -errno if not. On success, the rule structure + * will be allocated internally. The caller must free this structure with + * selinux_audit_rule_free() after use. + */ +int selinux_audit_rule_init(u32 field, u32 op, char *rulestr, + struct selinux_audit_rule **rule); + +/** + * selinux_audit_rule_free - free an selinux audit rule structure. + * @rule: pointer to the audit rule to be freed + * + * This will free all memory associated with the given rule. + * If @rule is NULL, no operation is performed. + */ +void selinux_audit_rule_free(struct selinux_audit_rule *rule); + +/** + * selinux_audit_rule_match - determine if a context ID matches a rule. + * @ctxid: the context ID to check + * @field: the field this rule refers to + * @op: the operater the rule uses + * @rule: pointer to the audit rule to check against + * @actx: the audit context (can be NULL) associated with the check + * + * Returns 1 if the context id matches the rule, 0 if it does not, and + * -errno on failure. + */ +int selinux_audit_rule_match(u32 ctxid, u32 field, u32 op, + struct selinux_audit_rule *rule, + struct audit_context *actx); + +/** + * selinux_audit_set_callback - set the callback for policy reloads. + * @callback: the function to call when the policy is reloaded + * + * This sets the function callback function that will update the rules + * upon policy reloads. This callback should rebuild all existing rules + * using selinux_audit_rule_init(). + */ +void selinux_audit_set_callback(int (*callback)(void)); + +/** + * selinux_task_ctxid - determine a context ID for a process. + * @tsk: the task object + * @ctxid: ID value returned via this + * + * On return, ctxid will contain an ID for the context. This value + * should only be used opaquely. + */ +void selinux_task_ctxid(struct task_struct *tsk, u32 *ctxid); + +#else + +static inline int selinux_audit_rule_init(u32 field, u32 op, + char *rulestr, + struct selinux_audit_rule **rule) +{ + return -ENOTSUPP; +} + +static inline void selinux_audit_rule_free(struct selinux_audit_rule *rule) +{ + return; +} + +static inline int selinux_audit_rule_match(u32 ctxid, u32 field, u32 op, + struct selinux_audit_rule *rule, + struct audit_context *actx) +{ + return 0; +} + +static inline void selinux_audit_set_callback(int (*callback)(void)) +{ + return; +} + +static inline void selinux_task_ctxid(struct task_struct *tsk, u32 *ctxid) +{ + *ctxid = 0; +} + +#endif /* CONFIG_SECURITY_SELINUX */ + +#endif /* _LINUX_SELINUX_H */ diff --git a/security/selinux/Makefile b/security/selinux/Makefile index 688c0a267b62..faf2e02e4410 100644 --- a/security/selinux/Makefile +++ b/security/selinux/Makefile @@ -4,7 +4,7 @@ obj-$(CONFIG_SECURITY_SELINUX) := selinux.o ss/ -selinux-y := avc.o hooks.o selinuxfs.o netlink.o nlmsgtab.o netif.o +selinux-y := avc.o hooks.o selinuxfs.o netlink.o nlmsgtab.o netif.o exports.o selinux-$(CONFIG_SECURITY_NETWORK_XFRM) += xfrm.o diff --git a/security/selinux/avc.c b/security/selinux/avc.c index ac5d69bb3377..a300702da527 100644 --- a/security/selinux/avc.c +++ b/security/selinux/avc.c @@ -800,7 +800,7 @@ out: int avc_ss_reset(u32 seqno) { struct avc_callback_node *c; - int i, rc = 0; + int i, rc = 0, tmprc; unsigned long flag; struct avc_node *node; @@ -813,15 +813,16 @@ int avc_ss_reset(u32 seqno) for (c = avc_callbacks; c; c = c->next) { if (c->events & AVC_CALLBACK_RESET) { - rc = c->callback(AVC_CALLBACK_RESET, - 0, 0, 0, 0, NULL); - if (rc) - goto out; + tmprc = c->callback(AVC_CALLBACK_RESET, + 0, 0, 0, 0, NULL); + /* save the first error encountered for the return + value and continue processing the callbacks */ + if (!rc) + rc = tmprc; } } avc_latest_notif_update(seqno, 0); -out: return rc; } diff --git a/security/selinux/exports.c b/security/selinux/exports.c new file mode 100644 index 000000000000..333c4c7824d8 --- /dev/null +++ b/security/selinux/exports.c @@ -0,0 +1,28 @@ +/* + * SELinux services exported to the rest of the kernel. + * + * Author: James Morris + * + * Copyright (C) 2005 Red Hat, Inc., James Morris + * Copyright (C) 2006 Trusted Computer Solutions, Inc. + * + * 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. + */ +#include +#include +#include +#include + +#include "security.h" +#include "objsec.h" + +void selinux_task_ctxid(struct task_struct *tsk, u32 *ctxid) +{ + struct task_security_struct *tsec = tsk->security; + if (selinux_enabled) + *ctxid = tsec->sid; + else + *ctxid = 0; +} diff --git a/security/selinux/ss/mls.c b/security/selinux/ss/mls.c index 84047f69f9c1..7bc5b6440f70 100644 --- a/security/selinux/ss/mls.c +++ b/security/selinux/ss/mls.c @@ -8,7 +8,7 @@ * * Support for enhanced MLS infrastructure. * - * Copyright (C) 2004-2005 Trusted Computer Solutions, Inc. + * Copyright (C) 2004-2006 Trusted Computer Solutions, Inc. */ #include @@ -384,6 +384,34 @@ out: return rc; } +/* + * Set the MLS fields in the security context structure + * `context' based on the string representation in + * the string `str'. This function will allocate temporary memory with the + * given constraints of gfp_mask. + */ +int mls_from_string(char *str, struct context *context, gfp_t gfp_mask) +{ + char *tmpstr, *freestr; + int rc; + + if (!selinux_mls_enabled) + return -EINVAL; + + /* we need freestr because mls_context_to_sid will change + the value of tmpstr */ + tmpstr = freestr = kstrdup(str, gfp_mask); + if (!tmpstr) { + rc = -ENOMEM; + } else { + rc = mls_context_to_sid(':', &tmpstr, context, + NULL, SECSID_NULL); + kfree(freestr); + } + + return rc; +} + /* * Copies the effective MLS range from `src' into `dst'. */ diff --git a/security/selinux/ss/mls.h b/security/selinux/ss/mls.h index 03de697c8058..fbb42f07dd7c 100644 --- a/security/selinux/ss/mls.h +++ b/security/selinux/ss/mls.h @@ -8,7 +8,7 @@ * * Support for enhanced MLS infrastructure. * - * Copyright (C) 2004-2005 Trusted Computer Solutions, Inc. + * Copyright (C) 2004-2006 Trusted Computer Solutions, Inc. */ #ifndef _SS_MLS_H_ @@ -27,6 +27,8 @@ int mls_context_to_sid(char oldc, struct sidtab *s, u32 def_sid); +int mls_from_string(char *str, struct context *context, gfp_t gfp_mask); + int mls_convert_context(struct policydb *oldp, struct policydb *newp, struct context *context); diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c index 61492485de84..7177e98df7f3 100644 --- a/security/selinux/ss/services.c +++ b/security/selinux/ss/services.c @@ -7,12 +7,13 @@ * Updated: Trusted Computer Solutions, Inc. * * Support for enhanced MLS infrastructure. + * Support for context based audit filters. * * Updated: Frank Mayer and Karl MacMillan * * Added conditional policy language extensions * - * Copyright (C) 2004-2005 Trusted Computer Solutions, Inc. + * Copyright (C) 2004-2006 Trusted Computer Solutions, Inc. * Copyright (C) 2003 - 2004 Tresys Technology, LLC * Copyright (C) 2003 Red Hat, Inc., James Morris * This program is free software; you can redistribute it and/or modify @@ -1811,3 +1812,235 @@ out: POLICY_RDUNLOCK; return rc; } + +struct selinux_audit_rule { + u32 au_seqno; + struct context au_ctxt; +}; + +void selinux_audit_rule_free(struct selinux_audit_rule *rule) +{ + if (rule) { + context_destroy(&rule->au_ctxt); + kfree(rule); + } +} + +int selinux_audit_rule_init(u32 field, u32 op, char *rulestr, + struct selinux_audit_rule **rule) +{ + struct selinux_audit_rule *tmprule; + struct role_datum *roledatum; + struct type_datum *typedatum; + struct user_datum *userdatum; + int rc = 0; + + *rule = NULL; + + if (!ss_initialized) + return -ENOTSUPP; + + switch (field) { + case AUDIT_SE_USER: + case AUDIT_SE_ROLE: + case AUDIT_SE_TYPE: + /* only 'equals' and 'not equals' fit user, role, and type */ + if (op != AUDIT_EQUAL && op != AUDIT_NOT_EQUAL) + return -EINVAL; + break; + case AUDIT_SE_SEN: + case AUDIT_SE_CLR: + /* we do not allow a range, indicated by the presense of '-' */ + if (strchr(rulestr, '-')) + return -EINVAL; + break; + default: + /* only the above fields are valid */ + return -EINVAL; + } + + tmprule = kzalloc(sizeof(struct selinux_audit_rule), GFP_KERNEL); + if (!tmprule) + return -ENOMEM; + + context_init(&tmprule->au_ctxt); + + POLICY_RDLOCK; + + tmprule->au_seqno = latest_granting; + + switch (field) { + case AUDIT_SE_USER: + userdatum = hashtab_search(policydb.p_users.table, rulestr); + if (!userdatum) + rc = -EINVAL; + else + tmprule->au_ctxt.user = userdatum->value; + break; + case AUDIT_SE_ROLE: + roledatum = hashtab_search(policydb.p_roles.table, rulestr); + if (!roledatum) + rc = -EINVAL; + else + tmprule->au_ctxt.role = roledatum->value; + break; + case AUDIT_SE_TYPE: + typedatum = hashtab_search(policydb.p_types.table, rulestr); + if (!typedatum) + rc = -EINVAL; + else + tmprule->au_ctxt.type = typedatum->value; + break; + case AUDIT_SE_SEN: + case AUDIT_SE_CLR: + rc = mls_from_string(rulestr, &tmprule->au_ctxt, GFP_ATOMIC); + break; + } + + POLICY_RDUNLOCK; + + if (rc) { + selinux_audit_rule_free(tmprule); + tmprule = NULL; + } + + *rule = tmprule; + + return rc; +} + +int selinux_audit_rule_match(u32 ctxid, u32 field, u32 op, + struct selinux_audit_rule *rule, + struct audit_context *actx) +{ + struct context *ctxt; + struct mls_level *level; + int match = 0; + + if (!rule) { + audit_log(actx, GFP_ATOMIC, AUDIT_SELINUX_ERR, + "selinux_audit_rule_match: missing rule\n"); + return -ENOENT; + } + + POLICY_RDLOCK; + + if (rule->au_seqno < latest_granting) { + audit_log(actx, GFP_ATOMIC, AUDIT_SELINUX_ERR, + "selinux_audit_rule_match: stale rule\n"); + match = -ESTALE; + goto out; + } + + ctxt = sidtab_search(&sidtab, ctxid); + if (!ctxt) { + audit_log(actx, GFP_ATOMIC, AUDIT_SELINUX_ERR, + "selinux_audit_rule_match: unrecognized SID %d\n", + ctxid); + match = -ENOENT; + goto out; + } + + /* a field/op pair that is not caught here will simply fall through + without a match */ + switch (field) { + case AUDIT_SE_USER: + switch (op) { + case AUDIT_EQUAL: + match = (ctxt->user == rule->au_ctxt.user); + break; + case AUDIT_NOT_EQUAL: + match = (ctxt->user != rule->au_ctxt.user); + break; + } + break; + case AUDIT_SE_ROLE: + switch (op) { + case AUDIT_EQUAL: + match = (ctxt->role == rule->au_ctxt.role); + break; + case AUDIT_NOT_EQUAL: + match = (ctxt->role != rule->au_ctxt.role); + break; + } + break; + case AUDIT_SE_TYPE: + switch (op) { + case AUDIT_EQUAL: + match = (ctxt->type == rule->au_ctxt.type); + break; + case AUDIT_NOT_EQUAL: + match = (ctxt->type != rule->au_ctxt.type); + break; + } + break; + case AUDIT_SE_SEN: + case AUDIT_SE_CLR: + level = (op == AUDIT_SE_SEN ? + &ctxt->range.level[0] : &ctxt->range.level[1]); + switch (op) { + case AUDIT_EQUAL: + match = mls_level_eq(&rule->au_ctxt.range.level[0], + level); + break; + case AUDIT_NOT_EQUAL: + match = !mls_level_eq(&rule->au_ctxt.range.level[0], + level); + break; + case AUDIT_LESS_THAN: + match = (mls_level_dom(&rule->au_ctxt.range.level[0], + level) && + !mls_level_eq(&rule->au_ctxt.range.level[0], + level)); + break; + case AUDIT_LESS_THAN_OR_EQUAL: + match = mls_level_dom(&rule->au_ctxt.range.level[0], + level); + break; + case AUDIT_GREATER_THAN: + match = (mls_level_dom(level, + &rule->au_ctxt.range.level[0]) && + !mls_level_eq(level, + &rule->au_ctxt.range.level[0])); + break; + case AUDIT_GREATER_THAN_OR_EQUAL: + match = mls_level_dom(level, + &rule->au_ctxt.range.level[0]); + break; + } + } + +out: + POLICY_RDUNLOCK; + return match; +} + +static int (*aurule_callback)(void) = NULL; + +static int aurule_avc_callback(u32 event, u32 ssid, u32 tsid, + u16 class, u32 perms, u32 *retained) +{ + int err = 0; + + if (event == AVC_CALLBACK_RESET && aurule_callback) + err = aurule_callback(); + return err; +} + +static int __init aurule_init(void) +{ + int err; + + err = avc_add_callback(aurule_avc_callback, AVC_CALLBACK_RESET, + SECSID_NULL, SECSID_NULL, SECCLASS_NULL, 0); + if (err) + panic("avc_add_callback() failed, error %d\n", err); + + return err; +} +__initcall(aurule_init); + +void selinux_audit_set_callback(int (*callback)(void)) +{ + aurule_callback = callback; +} -- cgit v1.2.3 From 1b50eed9cac0e8e5e4d3a522d8aa267f7f8f8acb Mon Sep 17 00:00:00 2001 From: Steve Grubb Date: Mon, 3 Apr 2006 14:06:13 -0400 Subject: [PATCH] audit inode patch Previously, we were gathering the context instead of the sid. Now in this patch, we gather just the sid and convert to context only if an audit event is being output. This patch brings the performance hit from 146% down to 23% Signed-off-by: Al Viro --- include/linux/selinux.h | 34 +++++++++++++++++++++++++++++ kernel/auditsc.c | 53 ++++++++++++++-------------------------------- security/selinux/exports.c | 24 +++++++++++++++++++++ 3 files changed, 74 insertions(+), 37 deletions(-) (limited to 'include') diff --git a/include/linux/selinux.h b/include/linux/selinux.h index 9d684b1728b0..84a6c7404687 100644 --- a/include/linux/selinux.h +++ b/include/linux/selinux.h @@ -15,6 +15,7 @@ struct selinux_audit_rule; struct audit_context; +struct inode; #ifdef CONFIG_SECURITY_SELINUX @@ -76,6 +77,27 @@ void selinux_audit_set_callback(int (*callback)(void)); */ void selinux_task_ctxid(struct task_struct *tsk, u32 *ctxid); +/** + * selinux_ctxid_to_string - map a security context ID to a string + * @ctxid: security context ID to be converted. + * @ctx: address of context string to be returned + * @ctxlen: length of returned context string. + * + * Returns 0 if successful, -errno if not. On success, the context + * string will be allocated internally, and the caller must call + * kfree() on it after use. + */ +int selinux_ctxid_to_string(u32 ctxid, char **ctx, u32 *ctxlen); + +/** + * selinux_get_inode_sid - get the inode's security context ID + * @inode: inode structure to get the sid from. + * @sid: pointer to security context ID to be filled in. + * + * Returns nothing + */ +void selinux_get_inode_sid(const struct inode *inode, u32 *sid); + #else static inline int selinux_audit_rule_init(u32 field, u32 op, @@ -107,6 +129,18 @@ static inline void selinux_task_ctxid(struct task_struct *tsk, u32 *ctxid) *ctxid = 0; } +static inline int selinux_ctxid_to_string(u32 ctxid, char **ctx, u32 *ctxlen) +{ + *ctx = NULL; + *ctxlen = 0; + return 0; +} + +static inline void selinux_get_inode_sid(const struct inode *inode, u32 *sid) +{ + *sid = 0; +} + #endif /* CONFIG_SECURITY_SELINUX */ #endif /* _LINUX_SELINUX_H */ diff --git a/kernel/auditsc.c b/kernel/auditsc.c index d3d97d28b69a..2e123a8a0d60 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -90,7 +90,7 @@ struct audit_names { uid_t uid; gid_t gid; dev_t rdev; - char *ctx; + u32 osid; }; struct audit_aux_data { @@ -410,9 +410,6 @@ static inline void audit_free_names(struct audit_context *context) #endif for (i = 0; i < context->name_count; i++) { - char *p = context->names[i].ctx; - context->names[i].ctx = NULL; - kfree(p); if (context->names[i].name) __putname(context->names[i].name); } @@ -674,6 +671,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts } } for (i = 0; i < context->name_count; i++) { + int call_panic = 0; unsigned long ino = context->names[i].ino; unsigned long pino = context->names[i].pino; @@ -703,12 +701,22 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts context->names[i].gid, MAJOR(context->names[i].rdev), MINOR(context->names[i].rdev)); - if (context->names[i].ctx) { - audit_log_format(ab, " obj=%s", - context->names[i].ctx); + if (context->names[i].osid != 0) { + char *ctx = NULL; + u32 len; + if (selinux_ctxid_to_string( + context->names[i].osid, &ctx, &len)) { + audit_log_format(ab, " obj=%u", + context->names[i].osid); + call_panic = 1; + } else + audit_log_format(ab, " obj=%s", ctx); + kfree(ctx); } audit_log_end(ab); + if (call_panic) + audit_panic("error converting sid to string"); } } @@ -946,37 +954,8 @@ void audit_putname(const char *name) void audit_inode_context(int idx, const struct inode *inode) { struct audit_context *context = current->audit_context; - const char *suffix = security_inode_xattr_getsuffix(); - char *ctx = NULL; - int len = 0; - - if (!suffix) - goto ret; - - len = security_inode_getsecurity(inode, suffix, NULL, 0, 0); - if (len == -EOPNOTSUPP) - goto ret; - if (len < 0) - goto error_path; - - ctx = kmalloc(len, GFP_KERNEL); - if (!ctx) - goto error_path; - - len = security_inode_getsecurity(inode, suffix, ctx, len, 0); - if (len < 0) - goto error_path; - - kfree(context->names[idx].ctx); - context->names[idx].ctx = ctx; - goto ret; -error_path: - if (ctx) - kfree(ctx); - audit_panic("error in audit_inode_context"); -ret: - return; + selinux_get_inode_sid(inode, &context->names[idx].osid); } diff --git a/security/selinux/exports.c b/security/selinux/exports.c index 333c4c7824d8..07ddce7bf374 100644 --- a/security/selinux/exports.c +++ b/security/selinux/exports.c @@ -14,6 +14,7 @@ #include #include #include +#include #include "security.h" #include "objsec.h" @@ -26,3 +27,26 @@ void selinux_task_ctxid(struct task_struct *tsk, u32 *ctxid) else *ctxid = 0; } + +int selinux_ctxid_to_string(u32 ctxid, char **ctx, u32 *ctxlen) +{ + if (selinux_enabled) + return security_sid_to_context(ctxid, ctx, ctxlen); + else { + *ctx = NULL; + *ctxlen = 0; + } + + return 0; +} + +void selinux_get_inode_sid(const struct inode *inode, u32 *sid) +{ + if (selinux_enabled) { + struct inode_security_struct *isec = inode->i_security; + *sid = isec->sid; + return; + } + *sid = 0; +} + -- cgit v1.2.3 From 9c7aa6aa74fa8a5cda36e54cbbe4fffe0214497d Mon Sep 17 00:00:00 2001 From: Steve Grubb Date: Fri, 31 Mar 2006 15:22:49 -0500 Subject: [PATCH] change lspp ipc auditing Hi, The patch below converts IPC auditing to collect sid's and convert to context string only if it needs to output an audit record. This patch depends on the inode audit change patch already being applied. Signed-off-by: Steve Grubb Signed-off-by: Al Viro --- include/linux/security.h | 16 ----------- include/linux/selinux.h | 15 ++++++++++ kernel/auditsc.c | 68 ++++++++++++++-------------------------------- security/dummy.c | 6 ---- security/selinux/exports.c | 11 ++++++++ security/selinux/hooks.c | 8 ------ 6 files changed, 47 insertions(+), 77 deletions(-) (limited to 'include') diff --git a/include/linux/security.h b/include/linux/security.h index aaa0a5cdbf75..1bab48f6aeac 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -869,11 +869,6 @@ struct swap_info_struct; * @ipcp contains the kernel IPC permission structure * @flag contains the desired (requested) permission set * Return 0 if permission is granted. - * @ipc_getsecurity: - * Copy the security label associated with the ipc object into - * @buffer. @buffer may be NULL to request the size of the buffer - * required. @size indicates the size of @buffer in bytes. Return - * number of bytes used/required on success. * * Security hooks for individual messages held in System V IPC message queues * @msg_msg_alloc_security: @@ -1223,7 +1218,6 @@ struct security_operations { void (*task_to_inode)(struct task_struct *p, struct inode *inode); int (*ipc_permission) (struct kern_ipc_perm * ipcp, short flag); - int (*ipc_getsecurity)(struct kern_ipc_perm *ipcp, void *buffer, size_t size); int (*msg_msg_alloc_security) (struct msg_msg * msg); void (*msg_msg_free_security) (struct msg_msg * msg); @@ -1887,11 +1881,6 @@ static inline int security_ipc_permission (struct kern_ipc_perm *ipcp, return security_ops->ipc_permission (ipcp, flag); } -static inline int security_ipc_getsecurity(struct kern_ipc_perm *ipcp, void *buffer, size_t size) -{ - return security_ops->ipc_getsecurity(ipcp, buffer, size); -} - static inline int security_msg_msg_alloc (struct msg_msg * msg) { return security_ops->msg_msg_alloc_security (msg); @@ -2532,11 +2521,6 @@ static inline int security_ipc_permission (struct kern_ipc_perm *ipcp, return 0; } -static inline int security_ipc_getsecurity(struct kern_ipc_perm *ipcp, void *buffer, size_t size) -{ - return -EOPNOTSUPP; -} - static inline int security_msg_msg_alloc (struct msg_msg * msg) { return 0; diff --git a/include/linux/selinux.h b/include/linux/selinux.h index 84a6c7404687..413d66773b91 100644 --- a/include/linux/selinux.h +++ b/include/linux/selinux.h @@ -16,6 +16,7 @@ struct selinux_audit_rule; struct audit_context; struct inode; +struct kern_ipc_perm; #ifdef CONFIG_SECURITY_SELINUX @@ -98,6 +99,15 @@ int selinux_ctxid_to_string(u32 ctxid, char **ctx, u32 *ctxlen); */ void selinux_get_inode_sid(const struct inode *inode, u32 *sid); +/** + * selinux_get_ipc_sid - get the ipc security context ID + * @ipcp: ipc structure to get the sid from. + * @sid: pointer to security context ID to be filled in. + * + * Returns nothing + */ +void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid); + #else static inline int selinux_audit_rule_init(u32 field, u32 op, @@ -141,6 +151,11 @@ static inline void selinux_get_inode_sid(const struct inode *inode, u32 *sid) *sid = 0; } +static inline void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid) +{ + *sid = 0; +} + #endif /* CONFIG_SECURITY_SELINUX */ #endif /* _LINUX_SELINUX_H */ diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 2e123a8a0d60..b4f7223811fe 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -107,7 +107,7 @@ struct audit_aux_data_ipcctl { uid_t uid; gid_t gid; mode_t mode; - char *ctx; + u32 osid; }; struct audit_aux_data_socketcall { @@ -432,11 +432,6 @@ static inline void audit_free_aux(struct audit_context *context) dput(axi->dentry); mntput(axi->mnt); } - if ( aux->type == AUDIT_IPC ) { - struct audit_aux_data_ipcctl *axi = (void *)aux; - if (axi->ctx) - kfree(axi->ctx); - } context->aux = aux->next; kfree(aux); @@ -584,7 +579,7 @@ static void audit_log_task_info(struct audit_buffer *ab, struct task_struct *tsk static void audit_log_exit(struct audit_context *context, struct task_struct *tsk) { - int i; + int i, call_panic = 0; struct audit_buffer *ab; struct audit_aux_data *aux; const char *tty; @@ -635,8 +630,20 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts case AUDIT_IPC: { struct audit_aux_data_ipcctl *axi = (void *)aux; audit_log_format(ab, - " qbytes=%lx iuid=%u igid=%u mode=%x obj=%s", - axi->qbytes, axi->uid, axi->gid, axi->mode, axi->ctx); + " qbytes=%lx iuid=%u igid=%u mode=%x", + axi->qbytes, axi->uid, axi->gid, axi->mode); + if (axi->osid != 0) { + char *ctx = NULL; + u32 len; + if (selinux_ctxid_to_string( + axi->osid, &ctx, &len)) { + audit_log_format(ab, " obj=%u", + axi->osid); + call_panic = 1; + } else + audit_log_format(ab, " obj=%s", ctx); + kfree(ctx); + } break; } case AUDIT_SOCKETCALL: { @@ -671,7 +678,6 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts } } for (i = 0; i < context->name_count; i++) { - int call_panic = 0; unsigned long ino = context->names[i].ino; unsigned long pino = context->names[i].pino; @@ -708,16 +714,16 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts context->names[i].osid, &ctx, &len)) { audit_log_format(ab, " obj=%u", context->names[i].osid); - call_panic = 1; + call_panic = 2; } else audit_log_format(ab, " obj=%s", ctx); kfree(ctx); } audit_log_end(ab); - if (call_panic) - audit_panic("error converting sid to string"); } + if (call_panic) + audit_panic("error converting sid to string"); } /** @@ -951,7 +957,7 @@ void audit_putname(const char *name) #endif } -void audit_inode_context(int idx, const struct inode *inode) +static void audit_inode_context(int idx, const struct inode *inode) { struct audit_context *context = current->audit_context; @@ -1141,38 +1147,6 @@ uid_t audit_get_loginuid(struct audit_context *ctx) return ctx ? ctx->loginuid : -1; } -static char *audit_ipc_context(struct kern_ipc_perm *ipcp) -{ - struct audit_context *context = current->audit_context; - char *ctx = NULL; - int len = 0; - - if (likely(!context)) - return NULL; - - len = security_ipc_getsecurity(ipcp, NULL, 0); - if (len == -EOPNOTSUPP) - goto ret; - if (len < 0) - goto error_path; - - ctx = kmalloc(len, GFP_ATOMIC); - if (!ctx) - goto error_path; - - len = security_ipc_getsecurity(ipcp, ctx, len); - if (len < 0) - goto error_path; - - return ctx; - -error_path: - kfree(ctx); - audit_panic("error in audit_ipc_context"); -ret: - return NULL; -} - /** * audit_ipc_perms - record audit data for ipc * @qbytes: msgq bytes @@ -1198,7 +1172,7 @@ int audit_ipc_perms(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, str ax->uid = uid; ax->gid = gid; ax->mode = mode; - ax->ctx = audit_ipc_context(ipcp); + selinux_get_ipc_sid(ipcp, &ax->osid); ax->d.type = AUDIT_IPC; ax->d.next = context->aux; diff --git a/security/dummy.c b/security/dummy.c index fd99429278e9..8ccccccc12ac 100644 --- a/security/dummy.c +++ b/security/dummy.c @@ -563,11 +563,6 @@ static int dummy_ipc_permission (struct kern_ipc_perm *ipcp, short flag) return 0; } -static int dummy_ipc_getsecurity(struct kern_ipc_perm *ipcp, void *buffer, size_t size) -{ - return -EOPNOTSUPP; -} - static int dummy_msg_msg_alloc_security (struct msg_msg *msg) { return 0; @@ -976,7 +971,6 @@ void security_fixup_ops (struct security_operations *ops) set_to_dummy_if_null(ops, task_reparent_to_init); set_to_dummy_if_null(ops, task_to_inode); set_to_dummy_if_null(ops, ipc_permission); - set_to_dummy_if_null(ops, ipc_getsecurity); set_to_dummy_if_null(ops, msg_msg_alloc_security); set_to_dummy_if_null(ops, msg_msg_free_security); set_to_dummy_if_null(ops, msg_queue_alloc_security); diff --git a/security/selinux/exports.c b/security/selinux/exports.c index 07ddce7bf374..7357cf247f60 100644 --- a/security/selinux/exports.c +++ b/security/selinux/exports.c @@ -15,6 +15,7 @@ #include #include #include +#include #include "security.h" #include "objsec.h" @@ -50,3 +51,13 @@ void selinux_get_inode_sid(const struct inode *inode, u32 *sid) *sid = 0; } +void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid) +{ + if (selinux_enabled) { + struct ipc_security_struct *isec = ipcp->security; + *sid = isec->sid; + return; + } + *sid = 0; +} + diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index b61b9554bc27..3cf368a16448 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -4052,13 +4052,6 @@ static int selinux_ipc_permission(struct kern_ipc_perm *ipcp, short flag) return ipc_has_perm(ipcp, av); } -static int selinux_ipc_getsecurity(struct kern_ipc_perm *ipcp, void *buffer, size_t size) -{ - struct ipc_security_struct *isec = ipcp->security; - - return selinux_getsecurity(isec->sid, buffer, size); -} - /* module stacking operations */ static int selinux_register_security (const char *name, struct security_operations *ops) { @@ -4321,7 +4314,6 @@ static struct security_operations selinux_ops = { .task_to_inode = selinux_task_to_inode, .ipc_permission = selinux_ipc_permission, - .ipc_getsecurity = selinux_ipc_getsecurity, .msg_msg_alloc_security = selinux_msg_msg_alloc_security, .msg_msg_free_security = selinux_msg_msg_free_security, -- cgit v1.2.3 From e7c3497013a7e5496ce3d5fd3c73b5cf5af7a56e Mon Sep 17 00:00:00 2001 From: Steve Grubb Date: Mon, 3 Apr 2006 09:08:13 -0400 Subject: [PATCH] Reworked patch for labels on user space messages The below patch should be applied after the inode and ipc sid patches. This patch is a reworking of Tim's patch that has been updated to match the inode and ipc patches since its similar. [updated: > Stephen Smalley also wanted to change a variable from isec to tsec in the > user sid patch. ] Signed-off-by: Steve Grubb Signed-off-by: Al Viro --- include/linux/netlink.h | 1 + include/linux/selinux.h | 16 ++++++++++++++++ kernel/audit.c | 22 +++++++++++++++++++--- net/netlink/af_netlink.c | 2 ++ security/selinux/exports.c | 11 +++++++++++ 5 files changed, 49 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/netlink.h b/include/linux/netlink.h index f8f3d1c927f8..87b8a5703ebc 100644 --- a/include/linux/netlink.h +++ b/include/linux/netlink.h @@ -143,6 +143,7 @@ struct netlink_skb_parms __u32 dst_group; kernel_cap_t eff_cap; __u32 loginuid; /* Login (audit) uid */ + __u32 sid; /* SELinux security id */ }; #define NETLINK_CB(skb) (*(struct netlink_skb_parms*)&((skb)->cb)) diff --git a/include/linux/selinux.h b/include/linux/selinux.h index 413d66773b91..4047bcde4484 100644 --- a/include/linux/selinux.h +++ b/include/linux/selinux.h @@ -5,6 +5,7 @@ * * Copyright (C) 2005 Red Hat, Inc., James Morris * Copyright (C) 2006 Trusted Computer Solutions, Inc. + * Copyright (C) 2006 IBM Corporation, Timothy R. Chavez * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2, @@ -108,6 +109,16 @@ void selinux_get_inode_sid(const struct inode *inode, u32 *sid); */ void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid); +/** + * selinux_get_task_sid - return the SID of task + * @tsk: the task whose SID will be returned + * @sid: pointer to security context ID to be filled in. + * + * Returns nothing + */ +void selinux_get_task_sid(struct task_struct *tsk, u32 *sid); + + #else static inline int selinux_audit_rule_init(u32 field, u32 op, @@ -156,6 +167,11 @@ static inline void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *si *sid = 0; } +static inline void selinux_get_task_sid(struct task_struct *tsk, u32 *sid) +{ + *sid = 0; +} + #endif /* CONFIG_SECURITY_SELINUX */ #endif /* _LINUX_SELINUX_H */ diff --git a/kernel/audit.c b/kernel/audit.c index 9060be750c48..7ec9ccae1299 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -390,7 +390,7 @@ static int audit_netlink_ok(kernel_cap_t eff_cap, u16 msg_type) static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) { - u32 uid, pid, seq; + u32 uid, pid, seq, sid; void *data; struct audit_status *status_get, status_set; int err; @@ -416,6 +416,7 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) pid = NETLINK_CREDS(skb)->pid; uid = NETLINK_CREDS(skb)->uid; loginuid = NETLINK_CB(skb).loginuid; + sid = NETLINK_CB(skb).sid; seq = nlh->nlmsg_seq; data = NLMSG_DATA(nlh); @@ -468,8 +469,23 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) ab = audit_log_start(NULL, GFP_KERNEL, msg_type); if (ab) { audit_log_format(ab, - "user pid=%d uid=%u auid=%u msg='%.1024s'", - pid, uid, loginuid, (char *)data); + "user pid=%d uid=%u auid=%u", + pid, uid, loginuid); + if (sid) { + char *ctx = NULL; + u32 len; + if (selinux_ctxid_to_string( + sid, &ctx, &len)) { + audit_log_format(ab, + " subj=%u", sid); + /* Maybe call audit_panic? */ + } else + audit_log_format(ab, + " subj=%s", ctx); + kfree(ctx); + } + audit_log_format(ab, " msg='%.1024s'", + (char *)data); audit_set_pid(ab, pid); audit_log_end(ab); } diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 2a233ffcf618..09fbc4bc7088 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -56,6 +56,7 @@ #include #include #include +#include #include #include @@ -1157,6 +1158,7 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock, NETLINK_CB(skb).dst_pid = dst_pid; NETLINK_CB(skb).dst_group = dst_group; NETLINK_CB(skb).loginuid = audit_get_loginuid(current->audit_context); + selinux_get_task_sid(current, &(NETLINK_CB(skb).sid)); memcpy(NETLINK_CREDS(skb), &siocb->scm->creds, sizeof(struct ucred)); /* What can I do? Netlink is asynchronous, so that diff --git a/security/selinux/exports.c b/security/selinux/exports.c index 7357cf247f60..ae4c73eb3085 100644 --- a/security/selinux/exports.c +++ b/security/selinux/exports.c @@ -5,6 +5,7 @@ * * Copyright (C) 2005 Red Hat, Inc., James Morris * Copyright (C) 2006 Trusted Computer Solutions, Inc. + * Copyright (C) 2006 IBM Corporation, Timothy R. Chavez * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2, @@ -61,3 +62,13 @@ void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid) *sid = 0; } +void selinux_get_task_sid(struct task_struct *tsk, u32 *sid) +{ + if (selinux_enabled) { + struct task_security_struct *tsec = tsk->security; + *sid = tsec->sid; + return; + } + *sid = 0; +} + -- cgit v1.2.3 From ce29b682e228c70cdc91a1b2935c5adb2087bab8 Mon Sep 17 00:00:00 2001 From: Steve Grubb Date: Sat, 1 Apr 2006 18:29:34 -0500 Subject: [PATCH] More user space subject labels Hi, The patch below builds upon the patch sent earlier and adds subject label to all audit events generated via the netlink interface. It also cleans up a few other minor things. Signed-off-by: Steve Grubb Signed-off-by: Al Viro --- include/linux/audit.h | 2 +- kernel/audit.c | 132 ++++++++++++++++++++++++++++++++++++++------------ kernel/auditfilter.c | 44 ++++++++++++++--- kernel/auditsc.c | 4 +- 4 files changed, 142 insertions(+), 40 deletions(-) (limited to 'include') diff --git a/include/linux/audit.h b/include/linux/audit.h index 740f950397b7..d5c40823e166 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -371,7 +371,7 @@ extern void audit_log_d_path(struct audit_buffer *ab, extern int audit_filter_user(struct netlink_skb_parms *cb, int type); extern int audit_filter_type(int type); extern int audit_receive_filter(int type, int pid, int uid, int seq, - void *data, size_t datasz, uid_t loginuid); + void *data, size_t datasz, uid_t loginuid, u32 sid); #else #define audit_log(c,g,t,f,...) do { ; } while (0) #define audit_log_start(c,g,t) ({ NULL; }) diff --git a/kernel/audit.c b/kernel/audit.c index 7ec9ccae1299..df57b493e1cb 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -230,49 +230,103 @@ void audit_log_lost(const char *message) } } -static int audit_set_rate_limit(int limit, uid_t loginuid) +static int audit_set_rate_limit(int limit, uid_t loginuid, u32 sid) { - int old = audit_rate_limit; - audit_rate_limit = limit; - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + int old = audit_rate_limit; + + if (sid) { + char *ctx = NULL; + u32 len; + int rc; + if ((rc = selinux_ctxid_to_string(sid, &ctx, &len))) + return rc; + else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "audit_rate_limit=%d old=%d by auid=%u subj=%s", + limit, old, loginuid, ctx); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, "audit_rate_limit=%d old=%d by auid=%u", - audit_rate_limit, old, loginuid); + limit, old, loginuid); + audit_rate_limit = limit; return old; } -static int audit_set_backlog_limit(int limit, uid_t loginuid) +static int audit_set_backlog_limit(int limit, uid_t loginuid, u32 sid) { - int old = audit_backlog_limit; - audit_backlog_limit = limit; - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + int old = audit_backlog_limit; + + if (sid) { + char *ctx = NULL; + u32 len; + int rc; + if ((rc = selinux_ctxid_to_string(sid, &ctx, &len))) + return rc; + else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "audit_backlog_limit=%d old=%d by auid=%u subj=%s", + limit, old, loginuid, ctx); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, "audit_backlog_limit=%d old=%d by auid=%u", - audit_backlog_limit, old, loginuid); + limit, old, loginuid); + audit_backlog_limit = limit; return old; } -static int audit_set_enabled(int state, uid_t loginuid) +static int audit_set_enabled(int state, uid_t loginuid, u32 sid) { - int old = audit_enabled; + int old = audit_enabled; + if (state != 0 && state != 1) return -EINVAL; - audit_enabled = state; - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + + if (sid) { + char *ctx = NULL; + u32 len; + int rc; + if ((rc = selinux_ctxid_to_string(sid, &ctx, &len))) + return rc; + else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "audit_enabled=%d old=%d by auid=%u subj=%s", + state, old, loginuid, ctx); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, "audit_enabled=%d old=%d by auid=%u", - audit_enabled, old, loginuid); + state, old, loginuid); + audit_enabled = state; return old; } -static int audit_set_failure(int state, uid_t loginuid) +static int audit_set_failure(int state, uid_t loginuid, u32 sid) { - int old = audit_failure; + int old = audit_failure; + if (state != AUDIT_FAIL_SILENT && state != AUDIT_FAIL_PRINTK && state != AUDIT_FAIL_PANIC) return -EINVAL; - audit_failure = state; - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + + if (sid) { + char *ctx = NULL; + u32 len; + int rc; + if ((rc = selinux_ctxid_to_string(sid, &ctx, &len))) + return rc; + else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "audit_failure=%d old=%d by auid=%u subj=%s", + state, old, loginuid, ctx); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, "audit_failure=%d old=%d by auid=%u", - audit_failure, old, loginuid); + state, old, loginuid); + audit_failure = state; return old; } @@ -437,25 +491,43 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) return -EINVAL; status_get = (struct audit_status *)data; if (status_get->mask & AUDIT_STATUS_ENABLED) { - err = audit_set_enabled(status_get->enabled, loginuid); + err = audit_set_enabled(status_get->enabled, + loginuid, sid); if (err < 0) return err; } if (status_get->mask & AUDIT_STATUS_FAILURE) { - err = audit_set_failure(status_get->failure, loginuid); + err = audit_set_failure(status_get->failure, + loginuid, sid); if (err < 0) return err; } if (status_get->mask & AUDIT_STATUS_PID) { int old = audit_pid; + if (sid) { + char *ctx = NULL; + u32 len; + int rc; + if ((rc = selinux_ctxid_to_string( + sid, &ctx, &len))) + return rc; + else + audit_log(NULL, GFP_KERNEL, + AUDIT_CONFIG_CHANGE, + "audit_pid=%d old=%d by auid=%u subj=%s", + status_get->pid, old, + loginuid, ctx); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "audit_pid=%d old=%d by auid=%u", + status_get->pid, old, loginuid); audit_pid = status_get->pid; - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, - "audit_pid=%d old=%d by auid=%u", - audit_pid, old, loginuid); } if (status_get->mask & AUDIT_STATUS_RATE_LIMIT) - audit_set_rate_limit(status_get->rate_limit, loginuid); + audit_set_rate_limit(status_get->rate_limit, + loginuid, sid); if (status_get->mask & AUDIT_STATUS_BACKLOG_LIMIT) audit_set_backlog_limit(status_get->backlog_limit, - loginuid); + loginuid, sid); break; case AUDIT_USER: case AUDIT_FIRST_USER_MSG...AUDIT_LAST_USER_MSG: @@ -477,7 +549,7 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) if (selinux_ctxid_to_string( sid, &ctx, &len)) { audit_log_format(ab, - " subj=%u", sid); + " ssid=%u", sid); /* Maybe call audit_panic? */ } else audit_log_format(ab, @@ -499,7 +571,7 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) case AUDIT_LIST: err = audit_receive_filter(nlh->nlmsg_type, NETLINK_CB(skb).pid, uid, seq, data, nlmsg_len(nlh), - loginuid); + loginuid, sid); break; case AUDIT_ADD_RULE: case AUDIT_DEL_RULE: @@ -509,7 +581,7 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) case AUDIT_LIST_RULES: err = audit_receive_filter(nlh->nlmsg_type, NETLINK_CB(skb).pid, uid, seq, data, nlmsg_len(nlh), - loginuid); + loginuid, sid); break; case AUDIT_SIGNAL_INFO: sig_data.uid = audit_sig_uid; diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index 85a7862143a1..7c134906d689 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -586,9 +586,10 @@ static int audit_list_rules(void *_dest) * @data: payload data * @datasz: size of payload data * @loginuid: loginuid of sender + * @sid: SE Linux Security ID of sender */ int audit_receive_filter(int type, int pid, int uid, int seq, void *data, - size_t datasz, uid_t loginuid) + size_t datasz, uid_t loginuid, u32 sid) { struct task_struct *tsk; int *dest; @@ -631,9 +632,23 @@ int audit_receive_filter(int type, int pid, int uid, int seq, void *data, err = audit_add_rule(entry, &audit_filter_list[entry->rule.listnr]); - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, - "auid=%u add rule to list=%d res=%d\n", - loginuid, entry->rule.listnr, !err); + if (sid) { + char *ctx = NULL; + u32 len; + if (selinux_ctxid_to_string(sid, &ctx, &len)) { + /* Maybe call audit_panic? */ + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "auid=%u ssid=%u add rule to list=%d res=%d", + loginuid, sid, entry->rule.listnr, !err); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "auid=%u subj=%s add rule to list=%d res=%d", + loginuid, ctx, entry->rule.listnr, !err); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "auid=%u add rule to list=%d res=%d", + loginuid, entry->rule.listnr, !err); if (err) audit_free_rule(entry); @@ -649,9 +664,24 @@ int audit_receive_filter(int type, int pid, int uid, int seq, void *data, err = audit_del_rule(entry, &audit_filter_list[entry->rule.listnr]); - audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, - "auid=%u remove rule from list=%d res=%d\n", - loginuid, entry->rule.listnr, !err); + + if (sid) { + char *ctx = NULL; + u32 len; + if (selinux_ctxid_to_string(sid, &ctx, &len)) { + /* Maybe call audit_panic? */ + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "auid=%u ssid=%u remove rule from list=%d res=%d", + loginuid, sid, entry->rule.listnr, !err); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "auid=%u subj=%s remove rule from list=%d res=%d", + loginuid, ctx, entry->rule.listnr, !err); + kfree(ctx); + } else + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "auid=%u remove rule from list=%d res=%d", + loginuid, entry->rule.listnr, !err); audit_free_rule(entry); break; diff --git a/kernel/auditsc.c b/kernel/auditsc.c index b4f7223811fe..d94e0404113c 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -637,7 +637,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts u32 len; if (selinux_ctxid_to_string( axi->osid, &ctx, &len)) { - audit_log_format(ab, " obj=%u", + audit_log_format(ab, " osid=%u", axi->osid); call_panic = 1; } else @@ -712,7 +712,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts u32 len; if (selinux_ctxid_to_string( context->names[i].osid, &ctx, &len)) { - audit_log_format(ab, " obj=%u", + audit_log_format(ab, " osid=%u", context->names[i].osid); call_panic = 2; } else -- cgit v1.2.3 From 073115d6b29c7910feaa08241c6484637f5ca958 Mon Sep 17 00:00:00 2001 From: Steve Grubb Date: Sun, 2 Apr 2006 17:07:33 -0400 Subject: [PATCH] Rework of IPC auditing 1) The audit_ipc_perms() function has been split into two different functions: - audit_ipc_obj() - audit_ipc_set_perm() There's a key shift here... The audit_ipc_obj() collects the uid, gid, mode, and SElinux context label of the current ipc object. This audit_ipc_obj() hook is now found in several places. Most notably, it is hooked in ipcperms(), which is called in various places around the ipc code permforming a MAC check. Additionally there are several places where *checkid() is used to validate that an operation is being performed on a valid object while not necessarily having a nearby ipcperms() call. In these locations, audit_ipc_obj() is called to ensure that the information is captured by the audit system. The audit_set_new_perm() function is called any time the permissions on the ipc object changes. In this case, the NEW permissions are recorded (and note that an audit_ipc_obj() call exists just a few lines before each instance). 2) Support for an AUDIT_IPC_SET_PERM audit message type. This allows for separate auxiliary audit records for normal operations on an IPC object and permissions changes. Note that the same struct audit_aux_data_ipcctl is used and populated, however there are separate audit_log_format statements based on the type of the message. Finally, the AUDIT_IPC block of code in audit_free_aux() was extended to handle aux messages of this new type. No more mem leaks I hope ;-) Signed-off-by: Al Viro --- include/linux/audit.h | 7 +++++-- ipc/msg.c | 11 ++++++++++- ipc/sem.c | 11 ++++++++++- ipc/shm.c | 19 +++++++++++++++--- ipc/util.c | 7 ++++++- kernel/auditsc.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++--- 6 files changed, 98 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/include/linux/audit.h b/include/linux/audit.h index d5c40823e166..b74c148f14e3 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -83,6 +83,7 @@ #define AUDIT_CONFIG_CHANGE 1305 /* Audit system configuration change */ #define AUDIT_SOCKADDR 1306 /* sockaddr copied as syscall arg */ #define AUDIT_CWD 1307 /* Current working directory */ +#define AUDIT_IPC_SET_PERM 1311 /* IPC new permissions record type */ #define AUDIT_AVC 1400 /* SE Linux avc denial or grant */ #define AUDIT_SELINUX_ERR 1401 /* Internal SE Linux Errors */ @@ -319,7 +320,8 @@ extern void auditsc_get_stamp(struct audit_context *ctx, struct timespec *t, unsigned int *serial); extern int audit_set_loginuid(struct task_struct *task, uid_t loginuid); extern uid_t audit_get_loginuid(struct audit_context *ctx); -extern int audit_ipc_perms(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, struct kern_ipc_perm *ipcp); +extern int audit_ipc_obj(struct kern_ipc_perm *ipcp); +extern int audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, struct kern_ipc_perm *ipcp); extern int audit_socketcall(int nargs, unsigned long *args); extern int audit_sockaddr(int len, void *addr); extern int audit_avc_path(struct dentry *dentry, struct vfsmount *mnt); @@ -338,7 +340,8 @@ extern int audit_set_macxattr(const char *name); #define audit_inode_child(d,i,p) do { ; } while (0) #define auditsc_get_stamp(c,t,s) do { BUG(); } while (0) #define audit_get_loginuid(c) ({ -1; }) -#define audit_ipc_perms(q,u,g,m,i) ({ 0; }) +#define audit_ipc_obj(i) ({ 0; }) +#define audit_ipc_set_perm(q,u,g,m,i) ({ 0; }) #define audit_socketcall(n,a) ({ 0; }) #define audit_sockaddr(len, addr) ({ 0; }) #define audit_avc_path(dentry, mnt) ({ 0; }) diff --git a/ipc/msg.c b/ipc/msg.c index 48a7f17a7236..7d1340ccb16b 100644 --- a/ipc/msg.c +++ b/ipc/msg.c @@ -13,6 +13,9 @@ * mostly rewritten, threaded and wake-one semantics added * MSGMAX limit removed, sysctl's added * (c) 1999 Manfred Spraul + * + * support for audit of ipc object properties and permission changes + * Dustin Kirkland */ #include @@ -447,6 +450,11 @@ asmlinkage long sys_msgctl (int msqid, int cmd, struct msqid_ds __user *buf) if (msg_checkid(msq,msqid)) goto out_unlock_up; ipcp = &msq->q_perm; + + err = audit_ipc_obj(ipcp); + if (err) + goto out_unlock_up; + err = -EPERM; if (current->euid != ipcp->cuid && current->euid != ipcp->uid && !capable(CAP_SYS_ADMIN)) @@ -460,7 +468,8 @@ asmlinkage long sys_msgctl (int msqid, int cmd, struct msqid_ds __user *buf) switch (cmd) { case IPC_SET: { - if ((err = audit_ipc_perms(setbuf.qbytes, setbuf.uid, setbuf.gid, setbuf.mode, ipcp))) + err = audit_ipc_set_perm(setbuf.qbytes, setbuf.uid, setbuf.gid, setbuf.mode, ipcp); + if (err) goto out_unlock_up; err = -EPERM; diff --git a/ipc/sem.c b/ipc/sem.c index 642659cd596b..7919f8ece6ba 100644 --- a/ipc/sem.c +++ b/ipc/sem.c @@ -61,6 +61,9 @@ * (c) 2001 Red Hat Inc * Lockless wakeup * (c) 2003 Manfred Spraul + * + * support for audit of ipc object properties and permission changes + * Dustin Kirkland */ #include @@ -820,6 +823,11 @@ static int semctl_down(int semid, int semnum, int cmd, int version, union semun goto out_unlock; } ipcp = &sma->sem_perm; + + err = audit_ipc_obj(ipcp); + if (err) + goto out_unlock; + if (current->euid != ipcp->cuid && current->euid != ipcp->uid && !capable(CAP_SYS_ADMIN)) { err=-EPERM; @@ -836,7 +844,8 @@ static int semctl_down(int semid, int semnum, int cmd, int version, union semun err = 0; break; case IPC_SET: - if ((err = audit_ipc_perms(0, setbuf.uid, setbuf.gid, setbuf.mode, ipcp))) + err = audit_ipc_set_perm(0, setbuf.uid, setbuf.gid, setbuf.mode, ipcp); + if (err) goto out_unlock; ipcp->uid = setbuf.uid; ipcp->gid = setbuf.gid; diff --git a/ipc/shm.c b/ipc/shm.c index 1c2faf62bc73..809896851902 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -13,6 +13,8 @@ * Shared /dev/zero support, Kanoj Sarcar * Move the mm functionality over to mm/shmem.c, Christoph Rohland * + * support for audit of ipc object properties and permission changes + * Dustin Kirkland */ #include @@ -542,6 +544,10 @@ asmlinkage long sys_shmctl (int shmid, int cmd, struct shmid_ds __user *buf) if(err) goto out_unlock; + err = audit_ipc_obj(&(shp->shm_perm)); + if (err) + goto out_unlock; + if (!capable(CAP_IPC_LOCK)) { err = -EPERM; if (current->euid != shp->shm_perm.uid && @@ -594,6 +600,10 @@ asmlinkage long sys_shmctl (int shmid, int cmd, struct shmid_ds __user *buf) if(err) goto out_unlock_up; + err = audit_ipc_obj(&(shp->shm_perm)); + if (err) + goto out_unlock_up; + if (current->euid != shp->shm_perm.uid && current->euid != shp->shm_perm.cuid && !capable(CAP_SYS_ADMIN)) { @@ -627,12 +637,15 @@ asmlinkage long sys_shmctl (int shmid, int cmd, struct shmid_ds __user *buf) err=-EINVAL; if(shp==NULL) goto out_up; - if ((err = audit_ipc_perms(0, setbuf.uid, setbuf.gid, - setbuf.mode, &(shp->shm_perm)))) - goto out_unlock_up; err = shm_checkid(shp,shmid); if(err) goto out_unlock_up; + err = audit_ipc_obj(&(shp->shm_perm)); + if (err) + goto out_unlock_up; + err = audit_ipc_set_perm(0, setbuf.uid, setbuf.gid, setbuf.mode, &(shp->shm_perm)); + if (err) + goto out_unlock_up; err=-EPERM; if (current->euid != shp->shm_perm.uid && current->euid != shp->shm_perm.cuid && diff --git a/ipc/util.c b/ipc/util.c index b3dcfad3b4f7..8193299f45f6 100644 --- a/ipc/util.c +++ b/ipc/util.c @@ -10,6 +10,8 @@ * Manfred Spraul * Oct 2002 - One lock per IPC id. RCU ipc_free for lock-free grow_ary(). * Mingming Cao + * Mar 2006 - support for audit of ipc object properties + * Dustin Kirkland */ #include @@ -27,6 +29,7 @@ #include #include #include +#include #include @@ -464,8 +467,10 @@ void ipc_rcu_putref(void *ptr) int ipcperms (struct kern_ipc_perm *ipcp, short flag) { /* flag will most probably be 0 or S_...UGO from */ - int requested_mode, granted_mode; + int requested_mode, granted_mode, err; + if (unlikely((err = audit_ipc_obj(ipcp)))) + return err; requested_mode = (flag >> 6) | (flag >> 3) | flag; granted_mode = ipcp->mode; if (current->euid == ipcp->cuid || current->euid == ipcp->uid) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index d94e0404113c..a300736ee037 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -646,6 +646,25 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts } break; } + case AUDIT_IPC_SET_PERM: { + struct audit_aux_data_ipcctl *axi = (void *)aux; + audit_log_format(ab, + " new qbytes=%lx new iuid=%u new igid=%u new mode=%x", + axi->qbytes, axi->uid, axi->gid, axi->mode); + if (axi->osid != 0) { + char *ctx = NULL; + u32 len; + if (selinux_ctxid_to_string( + axi->osid, &ctx, &len)) { + audit_log_format(ab, " osid=%u", + axi->osid); + call_panic = 1; + } else + audit_log_format(ab, " obj=%s", ctx); + kfree(ctx); + } + break; } + case AUDIT_SOCKETCALL: { int i; struct audit_aux_data_socketcall *axs = (void *)aux; @@ -1148,7 +1167,36 @@ uid_t audit_get_loginuid(struct audit_context *ctx) } /** - * audit_ipc_perms - record audit data for ipc + * audit_ipc_obj - record audit data for ipc object + * @ipcp: ipc permissions + * + * Returns 0 for success or NULL context or < 0 on error. + */ +int audit_ipc_obj(struct kern_ipc_perm *ipcp) +{ + struct audit_aux_data_ipcctl *ax; + struct audit_context *context = current->audit_context; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + + ax->uid = ipcp->uid; + ax->gid = ipcp->gid; + ax->mode = ipcp->mode; + selinux_get_ipc_sid(ipcp, &ax->osid); + + ax->d.type = AUDIT_IPC; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +/** + * audit_ipc_set_perm - record audit data for new ipc permissions * @qbytes: msgq bytes * @uid: msgq user id * @gid: msgq group id @@ -1156,7 +1204,7 @@ uid_t audit_get_loginuid(struct audit_context *ctx) * * Returns 0 for success or NULL context or < 0 on error. */ -int audit_ipc_perms(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, struct kern_ipc_perm *ipcp) +int audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, struct kern_ipc_perm *ipcp) { struct audit_aux_data_ipcctl *ax; struct audit_context *context = current->audit_context; @@ -1174,7 +1222,7 @@ int audit_ipc_perms(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, str ax->mode = mode; selinux_get_ipc_sid(ipcp, &ax->osid); - ax->d.type = AUDIT_IPC; + ax->d.type = AUDIT_IPC_SET_PERM; ax->d.next = context->aux; context->aux = (void *)ax; return 0; -- cgit v1.2.3 From 0568b409c74f7a125d92a09a3f386785700ef688 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 1 May 2006 19:50:48 +0200 Subject: [PATCH] splice: fix bugs in pipe_to_file() Found by Oleg Nesterov , fixed by me. - Only allow full pages to go to the page cache. - Check page != buf->page instead of using PIPE_BUF_FLAG_STOLEN. - Remember to clear 'stolen' if add_to_page_cache() fails. And as a cleanup on that: - Make the bottom fall-through logic a little less convoluted. Also make the steal path hold an extra reference to the page, so we don't have to differentiate between stolen and non-stolen at the end. Signed-off-by: Jens Axboe --- fs/pipe.c | 3 --- fs/splice.c | 37 +++++++++++++++++++------------------ include/linux/pipe_fs_i.h | 3 +-- 3 files changed, 20 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/fs/pipe.c b/fs/pipe.c index 5a369273c51b..888f265011bf 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -99,8 +99,6 @@ static void anon_pipe_buf_release(struct pipe_inode_info *pipe, { struct page *page = buf->page; - buf->flags &= ~PIPE_BUF_FLAG_STOLEN; - /* * If nobody else uses this page, and we don't already have a * temporary page, let's keep track of it as a one-deep @@ -130,7 +128,6 @@ static int anon_pipe_buf_steal(struct pipe_inode_info *pipe, struct page *page = buf->page; if (page_count(page) == 1) { - buf->flags |= PIPE_BUF_FLAG_STOLEN; lock_page(page); return 0; } diff --git a/fs/splice.c b/fs/splice.c index 9df28d30efa0..1633778f3652 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -78,7 +78,7 @@ static int page_cache_pipe_buf_steal(struct pipe_inode_info *info, return 1; } - buf->flags |= PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU; + buf->flags |= PIPE_BUF_FLAG_LRU; return 0; } @@ -87,7 +87,7 @@ static void page_cache_pipe_buf_release(struct pipe_inode_info *info, { page_cache_release(buf->page); buf->page = NULL; - buf->flags &= ~(PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU); + buf->flags &= ~PIPE_BUF_FLAG_LRU; } static void *page_cache_pipe_buf_map(struct file *file, @@ -587,9 +587,10 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, this_len = PAGE_CACHE_SIZE - offset; /* - * Reuse buf page, if SPLICE_F_MOVE is set. + * Reuse buf page, if SPLICE_F_MOVE is set and we are doing a full + * page. */ - if (sd->flags & SPLICE_F_MOVE) { + if ((sd->flags & SPLICE_F_MOVE) && this_len == PAGE_CACHE_SIZE) { /* * If steal succeeds, buf->page is now pruned from the vm * side (LRU and page cache) and we can reuse it. The page @@ -604,6 +605,8 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, goto find_page; } + page_cache_get(page); + if (!(buf->flags & PIPE_BUF_FLAG_LRU)) lru_cache_add(page); } else { @@ -662,7 +665,7 @@ find_page: } else if (ret) goto out; - if (!(buf->flags & PIPE_BUF_FLAG_STOLEN)) { + if (buf->page != page) { char *dst = kmap_atomic(page, KM_USER0); memcpy(dst + offset, src + buf->offset, this_len); @@ -671,22 +674,20 @@ find_page: } ret = mapping->a_ops->commit_write(file, page, offset, offset+this_len); - if (ret == AOP_TRUNCATED_PAGE) { + if (!ret) { + /* + * Return the number of bytes written and mark page as + * accessed, we are now done! + */ + ret = this_len; + mark_page_accessed(page); + balance_dirty_pages_ratelimited(mapping); + } else if (ret == AOP_TRUNCATED_PAGE) { page_cache_release(page); goto find_page; - } else if (ret) - goto out; - - /* - * Return the number of bytes written. - */ - ret = this_len; - mark_page_accessed(page); - balance_dirty_pages_ratelimited(mapping); + } out: - if (!(buf->flags & PIPE_BUF_FLAG_STOLEN)) - page_cache_release(page); - + page_cache_release(page); unlock_page(page); out_nomem: buf->ops->unmap(info, buf); diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 0008d4bd4059..3130977fc6ab 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -5,8 +5,7 @@ #define PIPE_BUFFERS (16) -#define PIPE_BUF_FLAG_STOLEN 0x01 -#define PIPE_BUF_FLAG_LRU 0x02 +#define PIPE_BUF_FLAG_LRU 0x01 struct pipe_buffer { struct page *page; -- cgit v1.2.3 From f84d751994441292593523c7069ed147176f6cab Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 1 May 2006 19:59:03 +0200 Subject: [PATCH] pipe: introduce ->pin() buffer operation The ->map() function is really expensive on highmem machines right now, since it has to use the slower kmap() instead of kmap_atomic(). Splice rarely needs to access the virtual address of a page, so it's a waste of time doing it. Introduce ->pin() to take over the responsibility of making sure the page data is valid. ->map() is then reduced to just kmap(). That way we can also share a most of the pipe buffer ops between pipe.c and splice.c Signed-off-by: Jens Axboe --- fs/pipe.c | 39 +++++++++++--------- fs/splice.c | 91 ++++++++++++++++------------------------------- include/linux/pipe_fs_i.h | 21 ++++++++++- 3 files changed, 73 insertions(+), 78 deletions(-) (limited to 'include') diff --git a/fs/pipe.c b/fs/pipe.c index 888f265011bf..d9644fd9cc0d 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -110,14 +110,14 @@ static void anon_pipe_buf_release(struct pipe_inode_info *pipe, page_cache_release(page); } -static void * anon_pipe_buf_map(struct file *file, struct pipe_inode_info *pipe, - struct pipe_buffer *buf) +void *generic_pipe_buf_map(struct pipe_inode_info *pipe, + struct pipe_buffer *buf) { return kmap(buf->page); } -static void anon_pipe_buf_unmap(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) +void generic_pipe_buf_unmap(struct pipe_inode_info *pipe, + struct pipe_buffer *buf) { kunmap(buf->page); } @@ -135,19 +135,24 @@ static int anon_pipe_buf_steal(struct pipe_inode_info *pipe, return 1; } -static void anon_pipe_buf_get(struct pipe_inode_info *info, - struct pipe_buffer *buf) +void generic_pipe_buf_get(struct pipe_inode_info *info, struct pipe_buffer *buf) { page_cache_get(buf->page); } +int generic_pipe_buf_pin(struct pipe_inode_info *info, struct pipe_buffer *buf) +{ + return 0; +} + static struct pipe_buf_operations anon_pipe_buf_ops = { .can_merge = 1, - .map = anon_pipe_buf_map, - .unmap = anon_pipe_buf_unmap, + .map = generic_pipe_buf_map, + .unmap = generic_pipe_buf_unmap, + .pin = generic_pipe_buf_pin, .release = anon_pipe_buf_release, .steal = anon_pipe_buf_steal, - .get = anon_pipe_buf_get, + .get = generic_pipe_buf_get, }; static ssize_t @@ -183,12 +188,14 @@ pipe_readv(struct file *filp, const struct iovec *_iov, if (chars > total_len) chars = total_len; - addr = ops->map(filp, pipe, buf); - if (IS_ERR(addr)) { + error = ops->pin(pipe, buf); + if (error) { if (!ret) - ret = PTR_ERR(addr); + error = ret; break; } + + addr = ops->map(pipe, buf); error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars); ops->unmap(pipe, buf); if (unlikely(error)) { @@ -300,11 +307,11 @@ pipe_writev(struct file *filp, const struct iovec *_iov, void *addr; int error; - addr = ops->map(filp, pipe, buf); - if (IS_ERR(addr)) { - error = PTR_ERR(addr); + error = ops->pin(pipe, buf); + if (error) goto out; - } + + addr = ops->map(pipe, buf); error = pipe_iov_copy_from_user(offset + addr, iov, chars); ops->unmap(pipe, buf); diff --git a/fs/splice.c b/fs/splice.c index 1633778f3652..d7538d83c367 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -90,9 +90,8 @@ static void page_cache_pipe_buf_release(struct pipe_inode_info *info, buf->flags &= ~PIPE_BUF_FLAG_LRU; } -static void *page_cache_pipe_buf_map(struct file *file, - struct pipe_inode_info *info, - struct pipe_buffer *buf) +static int page_cache_pipe_buf_pin(struct pipe_inode_info *info, + struct pipe_buffer *buf) { struct page *page = buf->page; int err; @@ -118,49 +117,25 @@ static void *page_cache_pipe_buf_map(struct file *file, } /* - * Page is ok afterall, fall through to mapping. + * Page is ok afterall, we are done. */ unlock_page(page); } - return kmap(page); + return 0; error: unlock_page(page); - return ERR_PTR(err); -} - -static void page_cache_pipe_buf_unmap(struct pipe_inode_info *info, - struct pipe_buffer *buf) -{ - kunmap(buf->page); -} - -static void *user_page_pipe_buf_map(struct file *file, - struct pipe_inode_info *pipe, - struct pipe_buffer *buf) -{ - return kmap(buf->page); -} - -static void user_page_pipe_buf_unmap(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) -{ - kunmap(buf->page); -} - -static void page_cache_pipe_buf_get(struct pipe_inode_info *info, - struct pipe_buffer *buf) -{ - page_cache_get(buf->page); + return err; } static struct pipe_buf_operations page_cache_pipe_buf_ops = { .can_merge = 0, - .map = page_cache_pipe_buf_map, - .unmap = page_cache_pipe_buf_unmap, + .map = generic_pipe_buf_map, + .unmap = generic_pipe_buf_unmap, + .pin = page_cache_pipe_buf_pin, .release = page_cache_pipe_buf_release, .steal = page_cache_pipe_buf_steal, - .get = page_cache_pipe_buf_get, + .get = generic_pipe_buf_get, }; static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, @@ -171,11 +146,12 @@ static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, static struct pipe_buf_operations user_page_pipe_buf_ops = { .can_merge = 0, - .map = user_page_pipe_buf_map, - .unmap = user_page_pipe_buf_unmap, + .map = generic_pipe_buf_map, + .unmap = generic_pipe_buf_unmap, + .pin = generic_pipe_buf_pin, .release = page_cache_pipe_buf_release, .steal = user_page_pipe_buf_steal, - .get = page_cache_pipe_buf_get, + .get = generic_pipe_buf_get, }; /* @@ -517,26 +493,16 @@ static int pipe_to_sendpage(struct pipe_inode_info *info, { struct file *file = sd->file; loff_t pos = sd->pos; - ssize_t ret; - void *ptr; - int more; + int ret, more; - /* - * Sub-optimal, but we are limited by the pipe ->map. We don't - * need a kmap'ed buffer here, we just want to make sure we - * have the page pinned if the pipe page originates from the - * page cache. - */ - ptr = buf->ops->map(file, info, buf); - if (IS_ERR(ptr)) - return PTR_ERR(ptr); - - more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len; + ret = buf->ops->pin(info, buf); + if (!ret) { + more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len; - ret = file->f_op->sendpage(file, buf->page, buf->offset, sd->len, - &pos, more); + ret = file->f_op->sendpage(file, buf->page, buf->offset, + sd->len, &pos, more); + } - buf->ops->unmap(info, buf); return ret; } @@ -569,15 +535,14 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, unsigned int offset, this_len; struct page *page; pgoff_t index; - char *src; int ret; /* * make sure the data in this buffer is uptodate */ - src = buf->ops->map(file, info, buf); - if (IS_ERR(src)) - return PTR_ERR(src); + ret = buf->ops->pin(info, buf); + if (unlikely(ret)) + return ret; index = sd->pos >> PAGE_CACHE_SHIFT; offset = sd->pos & ~PAGE_CACHE_MASK; @@ -666,11 +631,16 @@ find_page: goto out; if (buf->page != page) { - char *dst = kmap_atomic(page, KM_USER0); + /* + * Careful, ->map() uses KM_USER0! + */ + char *src = buf->ops->map(info, buf); + char *dst = kmap_atomic(page, KM_USER1); memcpy(dst + offset, src + buf->offset, this_len); flush_dcache_page(page); - kunmap_atomic(dst, KM_USER0); + kunmap_atomic(dst, KM_USER1); + buf->ops->unmap(info, buf); } ret = mapping->a_ops->commit_write(file, page, offset, offset+this_len); @@ -690,7 +660,6 @@ out: page_cache_release(page); unlock_page(page); out_nomem: - buf->ops->unmap(info, buf); return ret; } diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 3130977fc6ab..b8aae1fc5185 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -14,10 +14,23 @@ struct pipe_buffer { unsigned int flags; }; +/* + * Note on the nesting of these functions: + * + * ->pin() + * ->steal() + * ... + * ->map() + * ... + * ->unmap() + * + * That is, ->map() must be called on a pinned buffer, same goes for ->steal(). + */ struct pipe_buf_operations { int can_merge; - void * (*map)(struct file *, struct pipe_inode_info *, struct pipe_buffer *); + void * (*map)(struct pipe_inode_info *, struct pipe_buffer *); void (*unmap)(struct pipe_inode_info *, struct pipe_buffer *); + int (*pin)(struct pipe_inode_info *, struct pipe_buffer *); void (*release)(struct pipe_inode_info *, struct pipe_buffer *); int (*steal)(struct pipe_inode_info *, struct pipe_buffer *); void (*get)(struct pipe_inode_info *, struct pipe_buffer *); @@ -50,6 +63,12 @@ struct pipe_inode_info * alloc_pipe_info(struct inode * inode); void free_pipe_info(struct inode * inode); void __free_pipe_info(struct pipe_inode_info *); +/* Generic pipe buffer ops functions */ +void *generic_pipe_buf_map(struct pipe_inode_info *, struct pipe_buffer *); +void generic_pipe_buf_unmap(struct pipe_inode_info *, struct pipe_buffer *); +void generic_pipe_buf_get(struct pipe_inode_info *, struct pipe_buffer *); +int generic_pipe_buf_pin(struct pipe_inode_info *, struct pipe_buffer *); + /* * splice is tied to pipes as a transport (at least for now), so we'll just * add the splice flags here. -- cgit v1.2.3 From f6762b7ad8edd6abc802542ce845d3bc8adcb92f Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 1 May 2006 20:02:05 +0200 Subject: [PATCH] pipe: enable atomic copying of pipe data to/from user space The pipe ->map() method uses kmap() to virtually map the pages, which is both slow and has known scalability issues on SMP. This patch enables atomic copying of pipe pages, by pre-faulting data and using kmap_atomic() instead. lmbench bw_pipe and lat_pipe measurements agree this is a Good Thing. Here are results from that on a UP machine with highmem (1.5GiB of RAM), running first a UP kernel, SMP kernel, and SMP kernel patched. Vanilla-UP: Pipe bandwidth: 1622.28 MB/sec Pipe bandwidth: 1610.59 MB/sec Pipe bandwidth: 1608.30 MB/sec Pipe latency: 7.3275 microseconds Pipe latency: 7.2995 microseconds Pipe latency: 7.3097 microseconds Vanilla-SMP: Pipe bandwidth: 1382.19 MB/sec Pipe bandwidth: 1317.27 MB/sec Pipe bandwidth: 1355.61 MB/sec Pipe latency: 9.6402 microseconds Pipe latency: 9.6696 microseconds Pipe latency: 9.6153 microseconds Patched-SMP: Pipe bandwidth: 1578.70 MB/sec Pipe bandwidth: 1579.95 MB/sec Pipe bandwidth: 1578.63 MB/sec Pipe latency: 9.1654 microseconds Pipe latency: 9.2266 microseconds Pipe latency: 9.1527 microseconds Signed-off-by: Jens Axboe --- fs/pipe.c | 141 ++++++++++++++++++++++++++++++++++++++-------- fs/splice.c | 4 +- include/linux/pipe_fs_i.h | 11 ++-- 3 files changed, 126 insertions(+), 30 deletions(-) (limited to 'include') diff --git a/fs/pipe.c b/fs/pipe.c index d9644fd9cc0d..3941a7f78b5d 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -55,7 +55,8 @@ void pipe_wait(struct pipe_inode_info *pipe) } static int -pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len) +pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len, + int atomic) { unsigned long copy; @@ -64,8 +65,13 @@ pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len) iov++; copy = min_t(unsigned long, len, iov->iov_len); - if (copy_from_user(to, iov->iov_base, copy)) - return -EFAULT; + if (atomic) { + if (__copy_from_user_inatomic(to, iov->iov_base, copy)) + return -EFAULT; + } else { + if (copy_from_user(to, iov->iov_base, copy)) + return -EFAULT; + } to += copy; len -= copy; iov->iov_base += copy; @@ -75,7 +81,8 @@ pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len) } static int -pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len) +pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len, + int atomic) { unsigned long copy; @@ -84,8 +91,13 @@ pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len) iov++; copy = min_t(unsigned long, len, iov->iov_len); - if (copy_to_user(iov->iov_base, from, copy)) - return -EFAULT; + if (atomic) { + if (__copy_to_user_inatomic(iov->iov_base, from, copy)) + return -EFAULT; + } else { + if (copy_to_user(iov->iov_base, from, copy)) + return -EFAULT; + } from += copy; len -= copy; iov->iov_base += copy; @@ -94,6 +106,47 @@ pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len) return 0; } +/* + * Attempt to pre-fault in the user memory, so we can use atomic copies. + * Returns the number of bytes not faulted in. + */ +static int iov_fault_in_pages_write(struct iovec *iov, unsigned long len) +{ + while (!iov->iov_len) + iov++; + + while (len > 0) { + unsigned long this_len; + + this_len = min_t(unsigned long, len, iov->iov_len); + if (fault_in_pages_writeable(iov->iov_base, this_len)) + break; + + len -= this_len; + iov++; + } + + return len; +} + +/* + * Pre-fault in the user memory, so we can use atomic copies. + */ +static void iov_fault_in_pages_read(struct iovec *iov, unsigned long len) +{ + while (!iov->iov_len) + iov++; + + while (len > 0) { + unsigned long this_len; + + this_len = min_t(unsigned long, len, iov->iov_len); + fault_in_pages_readable(iov->iov_base, this_len); + len -= this_len; + iov++; + } +} + static void anon_pipe_buf_release(struct pipe_inode_info *pipe, struct pipe_buffer *buf) { @@ -111,15 +164,24 @@ static void anon_pipe_buf_release(struct pipe_inode_info *pipe, } void *generic_pipe_buf_map(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) + struct pipe_buffer *buf, int atomic) { + if (atomic) { + buf->flags |= PIPE_BUF_FLAG_ATOMIC; + return kmap_atomic(buf->page, KM_USER0); + } + return kmap(buf->page); } void generic_pipe_buf_unmap(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) + struct pipe_buffer *buf, void *map_data) { - kunmap(buf->page); + if (buf->flags & PIPE_BUF_FLAG_ATOMIC) { + buf->flags &= ~PIPE_BUF_FLAG_ATOMIC; + kunmap_atomic(map_data, KM_USER0); + } else + kunmap(buf->page); } static int anon_pipe_buf_steal(struct pipe_inode_info *pipe, @@ -183,7 +245,7 @@ pipe_readv(struct file *filp, const struct iovec *_iov, struct pipe_buf_operations *ops = buf->ops; void *addr; size_t chars = buf->len; - int error; + int error, atomic; if (chars > total_len) chars = total_len; @@ -195,12 +257,21 @@ pipe_readv(struct file *filp, const struct iovec *_iov, break; } - addr = ops->map(pipe, buf); - error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars); - ops->unmap(pipe, buf); + atomic = !iov_fault_in_pages_write(iov, chars); +redo: + addr = ops->map(pipe, buf, atomic); + error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars, atomic); + ops->unmap(pipe, buf, addr); if (unlikely(error)) { + /* + * Just retry with the slow path if we failed. + */ + if (atomic) { + atomic = 0; + goto redo; + } if (!ret) - ret = -EFAULT; + ret = error; break; } ret += chars; @@ -304,21 +375,28 @@ pipe_writev(struct file *filp, const struct iovec *_iov, int offset = buf->offset + buf->len; if (ops->can_merge && offset + chars <= PAGE_SIZE) { + int error, atomic = 1; void *addr; - int error; error = ops->pin(pipe, buf); if (error) goto out; - addr = ops->map(pipe, buf); + iov_fault_in_pages_read(iov, chars); +redo1: + addr = ops->map(pipe, buf, atomic); error = pipe_iov_copy_from_user(offset + addr, iov, - chars); - ops->unmap(pipe, buf); + chars, atomic); + ops->unmap(pipe, buf, addr); ret = error; do_wakeup = 1; - if (error) + if (error) { + if (atomic) { + atomic = 0; + goto redo1; + } goto out; + } buf->len += chars; total_len -= chars; ret = chars; @@ -341,7 +419,8 @@ pipe_writev(struct file *filp, const struct iovec *_iov, int newbuf = (pipe->curbuf + bufs) & (PIPE_BUFFERS-1); struct pipe_buffer *buf = pipe->bufs + newbuf; struct page *page = pipe->tmp_page; - int error; + char *src; + int error, atomic = 1; if (!page) { page = alloc_page(GFP_HIGHUSER); @@ -361,11 +440,27 @@ pipe_writev(struct file *filp, const struct iovec *_iov, if (chars > total_len) chars = total_len; - error = pipe_iov_copy_from_user(kmap(page), iov, chars); - kunmap(page); + iov_fault_in_pages_read(iov, chars); +redo2: + if (atomic) + src = kmap_atomic(page, KM_USER0); + else + src = kmap(page); + + error = pipe_iov_copy_from_user(src, iov, chars, + atomic); + if (atomic) + kunmap_atomic(src, KM_USER0); + else + kunmap(page); + if (unlikely(error)) { + if (atomic) { + atomic = 0; + goto redo2; + } if (!ret) - ret = -EFAULT; + ret = error; break; } ret += chars; diff --git a/fs/splice.c b/fs/splice.c index 0a6916423e7d..d4664a297bab 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -640,13 +640,13 @@ find_page: /* * Careful, ->map() uses KM_USER0! */ - char *src = buf->ops->map(info, buf); + char *src = buf->ops->map(info, buf, 1); char *dst = kmap_atomic(page, KM_USER1); memcpy(dst + offset, src + buf->offset, this_len); flush_dcache_page(page); kunmap_atomic(dst, KM_USER1); - buf->ops->unmap(info, buf); + buf->ops->unmap(info, buf, src); } ret = mapping->a_ops->commit_write(file, page, offset, offset+this_len); diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index b8aae1fc5185..4c054491e38e 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -5,7 +5,8 @@ #define PIPE_BUFFERS (16) -#define PIPE_BUF_FLAG_LRU 0x01 +#define PIPE_BUF_FLAG_LRU 0x01 /* page is on the LRU */ +#define PIPE_BUF_FLAG_ATOMIC 0x02 /* was atomically mapped */ struct pipe_buffer { struct page *page; @@ -28,8 +29,8 @@ struct pipe_buffer { */ struct pipe_buf_operations { int can_merge; - void * (*map)(struct pipe_inode_info *, struct pipe_buffer *); - void (*unmap)(struct pipe_inode_info *, struct pipe_buffer *); + void * (*map)(struct pipe_inode_info *, struct pipe_buffer *, int); + void (*unmap)(struct pipe_inode_info *, struct pipe_buffer *, void *); int (*pin)(struct pipe_inode_info *, struct pipe_buffer *); void (*release)(struct pipe_inode_info *, struct pipe_buffer *); int (*steal)(struct pipe_inode_info *, struct pipe_buffer *); @@ -64,8 +65,8 @@ void free_pipe_info(struct inode * inode); void __free_pipe_info(struct pipe_inode_info *); /* Generic pipe buffer ops functions */ -void *generic_pipe_buf_map(struct pipe_inode_info *, struct pipe_buffer *); -void generic_pipe_buf_unmap(struct pipe_inode_info *, struct pipe_buffer *); +void *generic_pipe_buf_map(struct pipe_inode_info *, struct pipe_buffer *, int); +void generic_pipe_buf_unmap(struct pipe_inode_info *, struct pipe_buffer *, void *); void generic_pipe_buf_get(struct pipe_inode_info *, struct pipe_buffer *); int generic_pipe_buf_pin(struct pipe_inode_info *, struct pipe_buffer *); -- cgit v1.2.3 From 7afa6fd037e51e95d322990cb127bb2b1217251a Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 1 May 2006 20:02:33 +0200 Subject: [PATCH] vmsplice: allow user to pass in gift pages If SPLICE_F_GIFT is set, the user is basically giving this pages away to the kernel. That means we can steal them for eg page cache uses instead of copying it. The data must be properly page aligned and also a multiple of the page size in length. Signed-off-by: Jens Axboe --- fs/splice.c | 28 +++++++++++++++++++++++++--- include/linux/pipe_fs_i.h | 2 ++ 2 files changed, 27 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/splice.c b/fs/splice.c index d4664a297bab..b150493b6fc3 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -141,7 +141,10 @@ static struct pipe_buf_operations page_cache_pipe_buf_ops = { static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, struct pipe_buffer *buf) { - return 1; + if (!(buf->flags & PIPE_BUF_FLAG_GIFT)) + return 1; + + return 0; } static struct pipe_buf_operations user_page_pipe_buf_ops = { @@ -186,6 +189,9 @@ static ssize_t splice_to_pipe(struct pipe_inode_info *pipe, buf->offset = spd->partial[page_nr].offset; buf->len = spd->partial[page_nr].len; buf->ops = spd->ops; + if (spd->flags & SPLICE_F_GIFT) + buf->flags |= PIPE_BUF_FLAG_GIFT; + pipe->nrbufs++; page_nr++; ret += buf->len; @@ -1073,7 +1079,7 @@ static long do_splice(struct file *in, loff_t __user *off_in, */ static int get_iovec_page_array(const struct iovec __user *iov, unsigned int nr_vecs, struct page **pages, - struct partial_page *partial) + struct partial_page *partial, int aligned) { int buffers = 0, error = 0; @@ -1113,6 +1119,15 @@ static int get_iovec_page_array(const struct iovec __user *iov, * in the user pages. */ off = (unsigned long) base & ~PAGE_MASK; + + /* + * If asked for alignment, the offset must be zero and the + * length a multiple of the PAGE_SIZE. + */ + error = -EINVAL; + if (aligned && (off || len & ~PAGE_MASK)) + break; + npages = (off + len + PAGE_SIZE - 1) >> PAGE_SHIFT; if (npages > PIPE_BUFFERS - buffers) npages = PIPE_BUFFERS - buffers; @@ -1206,7 +1221,8 @@ static long do_vmsplice(struct file *file, const struct iovec __user *iov, else if (unlikely(!nr_segs)) return 0; - spd.nr_pages = get_iovec_page_array(iov, nr_segs, pages, partial); + spd.nr_pages = get_iovec_page_array(iov, nr_segs, pages, partial, + flags & SPLICE_F_GIFT); if (spd.nr_pages <= 0) return spd.nr_pages; @@ -1314,6 +1330,12 @@ static int link_pipe(struct pipe_inode_info *ipipe, obuf = opipe->bufs + nbuf; *obuf = *ibuf; + /* + * Don't inherit the gift flag, we need to + * prevent multiple steals of this page. + */ + obuf->flags &= ~PIPE_BUF_FLAG_GIFT; + if (obuf->len > len) obuf->len = len; diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 4c054491e38e..df4d3fa7d3dc 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -7,6 +7,7 @@ #define PIPE_BUF_FLAG_LRU 0x01 /* page is on the LRU */ #define PIPE_BUF_FLAG_ATOMIC 0x02 /* was atomically mapped */ +#define PIPE_BUF_FLAG_GIFT 0x04 /* page is a gift */ struct pipe_buffer { struct page *page; @@ -79,6 +80,7 @@ int generic_pipe_buf_pin(struct pipe_inode_info *, struct pipe_buffer *); /* we may still block on the fd we splice */ /* from/to, of course */ #define SPLICE_F_MORE (0x04) /* expect more data */ +#define SPLICE_F_GIFT (0x08) /* pages passed in are a gift */ /* * Passed to the actors -- cgit v1.2.3 From 953039c8df7beb2694814e20e2707a77d335a2e3 Mon Sep 17 00:00:00 2001 From: Jeremy Kerr Date: Mon, 1 May 2006 12:16:12 -0700 Subject: [PATCH] powerpc: Allow devices to register with numa topology Change of_node_to_nid() to traverse the device tree, looking for a numa id. Cell uses this to assign ids to SPUs, which are children of the CPU node. Existing users of of_node_to_nid() are altered to use of_node_to_nid_single(), which doesn't do the traversal. Export an attach_sysdev_to_node() function, allowing system devices (eg. SPUs) to link themselves into the numa topology in sysfs. Signed-off-by: Arnd Bergmann Cc: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/kernel/sysfs.c | 18 ++++++++++++++++++ arch/powerpc/mm/numa.c | 32 +++++++++++++++++++++++++++----- include/asm-powerpc/topology.h | 24 ++++++++++++++++++++++++ 3 files changed, 69 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/arch/powerpc/kernel/sysfs.c b/arch/powerpc/kernel/sysfs.c index ed737cacf92d..5bc2585c8036 100644 --- a/arch/powerpc/kernel/sysfs.c +++ b/arch/powerpc/kernel/sysfs.c @@ -322,13 +322,31 @@ static void register_nodes(void) } } } + +int sysfs_add_device_to_node(struct sys_device *dev, int nid) +{ + struct node *node = &node_devices[nid]; + return sysfs_create_link(&node->sysdev.kobj, &dev->kobj, + kobject_name(&dev->kobj)); +} + +void sysfs_remove_device_from_node(struct sys_device *dev, int nid) +{ + struct node *node = &node_devices[nid]; + sysfs_remove_link(&node->sysdev.kobj, kobject_name(&dev->kobj)); +} + #else static void register_nodes(void) { return; } + #endif +EXPORT_SYMBOL_GPL(sysfs_add_device_to_node); +EXPORT_SYMBOL_GPL(sysfs_remove_device_from_node); + /* Only valid if CPU is present. */ static ssize_t show_physical_id(struct sys_device *dev, char *buf) { diff --git a/arch/powerpc/mm/numa.c b/arch/powerpc/mm/numa.c index 0a335f34974c..092355f37399 100644 --- a/arch/powerpc/mm/numa.c +++ b/arch/powerpc/mm/numa.c @@ -194,7 +194,7 @@ static int *of_get_associativity(struct device_node *dev) /* Returns nid in the range [0..MAX_NUMNODES-1], or -1 if no useful numa * info is found. */ -static int of_node_to_nid(struct device_node *device) +static int of_node_to_nid_single(struct device_node *device) { int nid = -1; unsigned int *tmp; @@ -216,6 +216,28 @@ out: return nid; } +/* Walk the device tree upwards, looking for an associativity id */ +int of_node_to_nid(struct device_node *device) +{ + struct device_node *tmp; + int nid = -1; + + of_node_get(device); + while (device) { + nid = of_node_to_nid_single(device); + if (nid != -1) + break; + + tmp = device; + device = of_get_parent(tmp); + of_node_put(tmp); + } + of_node_put(device); + + return nid; +} +EXPORT_SYMBOL_GPL(of_node_to_nid); + /* * In theory, the "ibm,associativity" property may contain multiple * associativity lists because a resource may be multiply connected @@ -300,7 +322,7 @@ static int __cpuinit numa_setup_cpu(unsigned long lcpu) goto out; } - nid = of_node_to_nid(cpu); + nid = of_node_to_nid_single(cpu); if (nid < 0 || !node_online(nid)) nid = any_online_node(NODE_MASK_ALL); @@ -393,7 +415,7 @@ static int __init parse_numa_properties(void) cpu = find_cpu_node(i); BUG_ON(!cpu); - nid = of_node_to_nid(cpu); + nid = of_node_to_nid_single(cpu); of_node_put(cpu); /* @@ -437,7 +459,7 @@ new_range: * have associativity properties. If none, then * everything goes to default_nid. */ - nid = of_node_to_nid(memory); + nid = of_node_to_nid_single(memory); if (nid < 0) nid = default_nid; node_set_online(nid); @@ -776,7 +798,7 @@ int hot_add_scn_to_nid(unsigned long scn_addr) ha_new_range: start = read_n_cells(n_mem_addr_cells, &memcell_buf); size = read_n_cells(n_mem_size_cells, &memcell_buf); - nid = of_node_to_nid(memory); + nid = of_node_to_nid_single(memory); /* Domains not present at boot default to 0 */ if (nid < 0 || !node_online(nid)) diff --git a/include/asm-powerpc/topology.h b/include/asm-powerpc/topology.h index 1e19cd00af25..87362a05542b 100644 --- a/include/asm-powerpc/topology.h +++ b/include/asm-powerpc/topology.h @@ -4,6 +4,9 @@ #include +struct sys_device; +struct device_node; + #ifdef CONFIG_NUMA #include @@ -27,6 +30,8 @@ static inline int node_to_first_cpu(int node) return first_cpu(tmp); } +int of_node_to_nid(struct device_node *device); + #define pcibus_to_node(node) (-1) #define pcibus_to_cpumask(bus) (cpu_online_map) @@ -57,10 +62,29 @@ static inline int node_to_first_cpu(int node) extern void __init dump_numa_cpu_topology(void); +extern int sysfs_add_device_to_node(struct sys_device *dev, int nid); +extern void sysfs_remove_device_from_node(struct sys_device *dev, int nid); + #else +static inline int of_node_to_nid(struct device_node *device) +{ + return 0; +} + static inline void dump_numa_cpu_topology(void) {} +static inline int sysfs_add_device_to_node(struct sys_device *dev, int nid) +{ + return 0; +} + +static inline void sysfs_remove_device_from_node(struct sys_device *dev, + int nid) +{ +} + + #include #endif /* CONFIG_NUMA */ -- cgit v1.2.3 From 8261aa600943046a5305564a1cd7432009971799 Mon Sep 17 00:00:00 2001 From: Jeremy Kerr Date: Mon, 1 May 2006 12:16:13 -0700 Subject: [PATCH] powerpc: cell: Add numa id to struct spu Add an nid member to the spu structure, and store the numa id of the spu there on creation. Signed-off-by: Arnd Bergmann Cc: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/platforms/cell/spu_base.c | 43 ++++++++++++++++------------------ include/asm-powerpc/spu.h | 1 + 2 files changed, 21 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/arch/powerpc/platforms/cell/spu_base.c b/arch/powerpc/platforms/cell/spu_base.c index b788e165c557..ad141fe8d52d 100644 --- a/arch/powerpc/platforms/cell/spu_base.c +++ b/arch/powerpc/platforms/cell/spu_base.c @@ -529,8 +529,8 @@ static int __init find_spu_node_id(struct device_node *spe) return id ? *id : 0; } -static int __init cell_spuprop_present(struct device_node *spe, - const char *prop) +static int __init cell_spuprop_present(struct spu *spu, struct device_node *spe, + const char *prop) { static DEFINE_MUTEX(add_spumem_mutex); @@ -541,7 +541,6 @@ static int __init cell_spuprop_present(struct device_node *spe, int proplen; unsigned long start_pfn, nr_pages; - int node_id; struct pglist_data *pgdata; struct zone *zone; int ret; @@ -552,14 +551,7 @@ static int __init cell_spuprop_present(struct device_node *spe, start_pfn = p->address >> PAGE_SHIFT; nr_pages = ((unsigned long)p->len + PAGE_SIZE - 1) >> PAGE_SHIFT; - /* - * XXX need to get the correct NUMA node in here. This may - * be different from the spe::node_id property, e.g. when - * the host firmware is not NUMA aware. - */ - node_id = 0; - - pgdata = NODE_DATA(node_id); + pgdata = NODE_DATA(spu->nid); zone = pgdata->node_zones; /* XXX rethink locking here */ @@ -570,8 +562,8 @@ static int __init cell_spuprop_present(struct device_node *spe, return ret; } -static void __iomem * __init map_spe_prop(struct device_node *n, - const char *name) +static void __iomem * __init map_spe_prop(struct spu *spu, + struct device_node *n, const char *name) { struct address_prop { unsigned long address; @@ -589,7 +581,7 @@ static void __iomem * __init map_spe_prop(struct device_node *n, prop = p; - err = cell_spuprop_present(n, name); + err = cell_spuprop_present(spu, n, name); if (err && (err != -EEXIST)) goto out; @@ -607,44 +599,45 @@ static void spu_unmap(struct spu *spu) iounmap((u8 __iomem *)spu->local_store); } -static int __init spu_map_device(struct spu *spu, struct device_node *spe) +static int __init spu_map_device(struct spu *spu, struct device_node *node) { char *prop; int ret; ret = -ENODEV; - prop = get_property(spe, "isrc", NULL); + prop = get_property(node, "isrc", NULL); if (!prop) goto out; spu->isrc = *(unsigned int *)prop; - spu->name = get_property(spe, "name", NULL); + spu->name = get_property(node, "name", NULL); if (!spu->name) goto out; - prop = get_property(spe, "local-store", NULL); + prop = get_property(node, "local-store", NULL); if (!prop) goto out; spu->local_store_phys = *(unsigned long *)prop; /* we use local store as ram, not io memory */ - spu->local_store = (void __force *)map_spe_prop(spe, "local-store"); + spu->local_store = (void __force *) + map_spe_prop(spu, node, "local-store"); if (!spu->local_store) goto out; - prop = get_property(spe, "problem", NULL); + prop = get_property(node, "problem", NULL); if (!prop) goto out_unmap; spu->problem_phys = *(unsigned long *)prop; - spu->problem= map_spe_prop(spe, "problem"); + spu->problem= map_spe_prop(spu, node, "problem"); if (!spu->problem) goto out_unmap; - spu->priv1= map_spe_prop(spe, "priv1"); + spu->priv1= map_spe_prop(spu, node, "priv1"); /* priv1 is not available on a hypervisor */ - spu->priv2= map_spe_prop(spe, "priv2"); + spu->priv2= map_spe_prop(spu, node, "priv2"); if (!spu->priv2) goto out_unmap; ret = 0; @@ -672,6 +665,10 @@ static int __init create_spu(struct device_node *spe) goto out_free; spu->node = find_spu_node_id(spe); + spu->nid = of_node_to_nid(spe); + if (spu->nid == -1) + spu->nid = 0; + spu->stop_code = 0; spu->slb_replace = 0; spu->mm = NULL; diff --git a/include/asm-powerpc/spu.h b/include/asm-powerpc/spu.h index f431d8b0b651..7cfcff3ef027 100644 --- a/include/asm-powerpc/spu.h +++ b/include/asm-powerpc/spu.h @@ -117,6 +117,7 @@ struct spu { struct list_head list; struct list_head sched_list; int number; + int nid; u32 isrc; u32 node; u64 flags; -- cgit v1.2.3 From 46c5ea3c9ae7fbc6e52a13c92e59d4fc7f4ca80a Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 2 May 2006 05:12:22 +0200 Subject: [NETFILTER] x_tables: fix compat related crash on non-x86 When iptables userspace adds an ipt_standard_target, it calculates the size of the entire entry as: sizeof(struct ipt_entry) + XT_ALIGN(sizeof(struct ipt_standard_target)) ipt_standard_target looks like this: struct xt_standard_target { struct xt_entry_target target; int verdict; }; xt_entry_target contains a pointer, so when compiled for 64 bit the structure gets an extra 4 byte of padding at the end. On 32 bit architectures where iptables aligns to 8 byte it will also have 4 byte padding at the end because it is only 36 bytes large. The compat_ipt_standard_fn in the kernel adjusts the offsets by sizeof(struct ipt_standard_target) - sizeof(struct compat_ipt_standard_target), which will always result in 4, even if the structure from userspace was already padded to a multiple of 8. On x86 this works out by accident because userspace only aligns to 4, on all other architectures this is broken and causes incorrect adjustments to the size and following offsets. Thanks to Linus for lots of debugging help and testing. Signed-off-by: Patrick McHardy Signed-off-by: Linus Torvalds --- include/linux/netfilter/x_tables.h | 8 ++++++++ net/ipv4/netfilter/ip_tables.c | 33 ++++++++++++++------------------- 2 files changed, 22 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h index 38701454e197..48cc32d83f77 100644 --- a/include/linux/netfilter/x_tables.h +++ b/include/linux/netfilter/x_tables.h @@ -337,6 +337,10 @@ struct compat_xt_entry_match char name[XT_FUNCTION_MAXNAMELEN - 1]; u_int8_t revision; } user; + struct { + u_int16_t match_size; + compat_uptr_t match; + } kernel; u_int16_t match_size; } u; unsigned char data[0]; @@ -350,6 +354,10 @@ struct compat_xt_entry_target char name[XT_FUNCTION_MAXNAMELEN - 1]; u_int8_t revision; } user; + struct { + u_int16_t target_size; + compat_uptr_t target; + } kernel; u_int16_t target_size; } u; unsigned char data[0]; diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c index d25ac8ba6eba..6d1c11563943 100644 --- a/net/ipv4/netfilter/ip_tables.c +++ b/net/ipv4/netfilter/ip_tables.c @@ -956,15 +956,16 @@ struct compat_ipt_standard_target compat_int_t verdict; }; -#define IPT_ST_OFFSET (sizeof(struct ipt_standard_target) - \ - sizeof(struct compat_ipt_standard_target)) - struct compat_ipt_standard { struct compat_ipt_entry entry; struct compat_ipt_standard_target target; }; +#define IPT_ST_LEN XT_ALIGN(sizeof(struct ipt_standard_target)) +#define IPT_ST_COMPAT_LEN COMPAT_XT_ALIGN(sizeof(struct compat_ipt_standard_target)) +#define IPT_ST_OFFSET (IPT_ST_LEN - IPT_ST_COMPAT_LEN) + static int compat_ipt_standard_fn(void *target, void **dstptr, int *size, int convert) { @@ -975,35 +976,29 @@ static int compat_ipt_standard_fn(void *target, ret = 0; switch (convert) { case COMPAT_TO_USER: - pst = (struct ipt_standard_target *)target; + pst = target; memcpy(&compat_st.target, &pst->target, - sizeof(struct ipt_entry_target)); + sizeof(compat_st.target)); compat_st.verdict = pst->verdict; if (compat_st.verdict > 0) compat_st.verdict -= compat_calc_jump(compat_st.verdict); - compat_st.target.u.user.target_size = - sizeof(struct compat_ipt_standard_target); - if (__copy_to_user(*dstptr, &compat_st, - sizeof(struct compat_ipt_standard_target))) + compat_st.target.u.user.target_size = IPT_ST_COMPAT_LEN; + if (copy_to_user(*dstptr, &compat_st, IPT_ST_COMPAT_LEN)) ret = -EFAULT; *size -= IPT_ST_OFFSET; - *dstptr += sizeof(struct compat_ipt_standard_target); + *dstptr += IPT_ST_COMPAT_LEN; break; case COMPAT_FROM_USER: - pcompat_st = - (struct compat_ipt_standard_target *)target; - memcpy(&st.target, &pcompat_st->target, - sizeof(struct ipt_entry_target)); + pcompat_st = target; + memcpy(&st.target, &pcompat_st->target, IPT_ST_COMPAT_LEN); st.verdict = pcompat_st->verdict; if (st.verdict > 0) st.verdict += compat_calc_jump(st.verdict); - st.target.u.user.target_size = - sizeof(struct ipt_standard_target); - memcpy(*dstptr, &st, - sizeof(struct ipt_standard_target)); + st.target.u.user.target_size = IPT_ST_LEN; + memcpy(*dstptr, &st, IPT_ST_LEN); *size += IPT_ST_OFFSET; - *dstptr += sizeof(struct ipt_standard_target); + *dstptr += IPT_ST_LEN; break; case COMPAT_CALC_SIZE: *size += IPT_ST_OFFSET; -- cgit v1.2.3 From a893b99be71f1d669b74f840e3a683dd077d007b Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 2 May 2006 15:03:27 +0200 Subject: [PATCH] splice: fix page LRU accounting Currently we rely on the PIPE_BUF_FLAG_LRU flag being set correctly to know whether we need to fiddle with page LRU state after stealing it, however for some origins we just don't know if the page is on the LRU list or not. So remove PIPE_BUF_FLAG_LRU and do this check/add manually in pipe_to_file() instead. Signed-off-by: Jens Axboe --- fs/splice.c | 31 +++++++++++++++++++++---------- include/linux/pipe_fs_i.h | 5 ++--- 2 files changed, 23 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/fs/splice.c b/fs/splice.c index b0c157d76948..3318b965c10b 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -78,7 +78,6 @@ static int page_cache_pipe_buf_steal(struct pipe_inode_info *info, return 1; } - buf->flags |= PIPE_BUF_FLAG_LRU; return 0; } @@ -86,8 +85,6 @@ static void page_cache_pipe_buf_release(struct pipe_inode_info *info, struct pipe_buffer *buf) { page_cache_release(buf->page); - buf->page = NULL; - buf->flags &= ~PIPE_BUF_FLAG_LRU; } static int page_cache_pipe_buf_pin(struct pipe_inode_info *info, @@ -570,22 +567,36 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, if ((sd->flags & SPLICE_F_MOVE) && this_len == PAGE_CACHE_SIZE) { /* * If steal succeeds, buf->page is now pruned from the vm - * side (LRU and page cache) and we can reuse it. The page - * will also be looked on successful return. + * side (page cache) and we can reuse it. The page will also + * be locked on successful return. */ if (buf->ops->steal(info, buf)) goto find_page; page = buf->page; + page_cache_get(page); + + /* + * page must be on the LRU for adding to the pagecache. + * Check this without grabbing the zone lock, if it isn't + * the do grab the zone lock, recheck, and add if necessary. + */ + if (!PageLRU(page)) { + struct zone *zone = page_zone(page); + + spin_lock_irq(&zone->lru_lock); + if (!PageLRU(page)) { + SetPageLRU(page); + add_page_to_inactive_list(zone, page); + } + spin_unlock_irq(&zone->lru_lock); + } + if (add_to_page_cache(page, mapping, index, gfp_mask)) { + page_cache_release(page); unlock_page(page); goto find_page; } - - page_cache_get(page); - - if (!(buf->flags & PIPE_BUF_FLAG_LRU)) - lru_cache_add(page); } else { find_page: page = find_lock_page(mapping, index); diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index df4d3fa7d3dc..070954f05947 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -5,9 +5,8 @@ #define PIPE_BUFFERS (16) -#define PIPE_BUF_FLAG_LRU 0x01 /* page is on the LRU */ -#define PIPE_BUF_FLAG_ATOMIC 0x02 /* was atomically mapped */ -#define PIPE_BUF_FLAG_GIFT 0x04 /* page is a gift */ +#define PIPE_BUF_FLAG_ATOMIC 0x01 /* was atomically mapped */ +#define PIPE_BUF_FLAG_GIFT 0x02 /* page is a gift */ struct pipe_buffer { struct page *page; -- cgit v1.2.3 From 330ab71619bacc4d4494227a6cfc9b7f5500403d Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 2 May 2006 15:29:57 +0200 Subject: [PATCH] vmsplice: restrict stealing a little more Apply the same rules as the anon pipe pages, only allow stealing if no one else is using the page. Signed-off-by: Jens Axboe --- fs/pipe.c | 6 +++--- fs/splice.c | 2 +- include/linux/pipe_fs_i.h | 1 + 3 files changed, 5 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/fs/pipe.c b/fs/pipe.c index 3941a7f78b5d..5acd8954aaa0 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -184,8 +184,8 @@ void generic_pipe_buf_unmap(struct pipe_inode_info *pipe, kunmap(buf->page); } -static int anon_pipe_buf_steal(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) +int generic_pipe_buf_steal(struct pipe_inode_info *pipe, + struct pipe_buffer *buf) { struct page *page = buf->page; @@ -213,7 +213,7 @@ static struct pipe_buf_operations anon_pipe_buf_ops = { .unmap = generic_pipe_buf_unmap, .pin = generic_pipe_buf_pin, .release = anon_pipe_buf_release, - .steal = anon_pipe_buf_steal, + .steal = generic_pipe_buf_steal, .get = generic_pipe_buf_get, }; diff --git a/fs/splice.c b/fs/splice.c index 3318b965c10b..7fb04970c72d 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -141,7 +141,7 @@ static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, if (!(buf->flags & PIPE_BUF_FLAG_GIFT)) return 1; - return 0; + return generic_pipe_buf_steal(pipe, buf); } static struct pipe_buf_operations user_page_pipe_buf_ops = { diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 070954f05947..ba73108cbf8b 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -69,6 +69,7 @@ void *generic_pipe_buf_map(struct pipe_inode_info *, struct pipe_buffer *, int); void generic_pipe_buf_unmap(struct pipe_inode_info *, struct pipe_buffer *, void *); void generic_pipe_buf_get(struct pipe_inode_info *, struct pipe_buffer *); int generic_pipe_buf_pin(struct pipe_inode_info *, struct pipe_buffer *); +int generic_pipe_buf_steal(struct pipe_inode_info *, struct pipe_buffer *); /* * splice is tied to pipes as a transport (at least for now), so we'll just -- cgit v1.2.3 From 37be4e7809e0581db85387e126ae4da68c3d6286 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 May 2006 17:24:59 +0100 Subject: [MMC] extend data timeout for writes The CSD contains a "read2write factor" which determines the multiplier to be applied to the read timeout to obtain the write timeout. We were ignoring this parameter, resulting in the possibility for writes being timed out too early. Signed-off-by: Russell King --- drivers/mmc/mmc.c | 2 ++ drivers/mmc/mmc_block.c | 6 ++++++ include/linux/mmc/card.h | 1 + 3 files changed, 9 insertions(+) (limited to 'include') diff --git a/drivers/mmc/mmc.c b/drivers/mmc/mmc.c index da6ddd910fc5..05aa4d6b4f24 100644 --- a/drivers/mmc/mmc.c +++ b/drivers/mmc/mmc.c @@ -549,6 +549,7 @@ static void mmc_decode_csd(struct mmc_card *card) csd->read_partial = UNSTUFF_BITS(resp, 79, 1); csd->write_misalign = UNSTUFF_BITS(resp, 78, 1); csd->read_misalign = UNSTUFF_BITS(resp, 77, 1); + csd->r2w_factor = UNSTUFF_BITS(resp, 26, 3); csd->write_blkbits = UNSTUFF_BITS(resp, 22, 4); csd->write_partial = UNSTUFF_BITS(resp, 21, 1); } else { @@ -583,6 +584,7 @@ static void mmc_decode_csd(struct mmc_card *card) csd->read_partial = UNSTUFF_BITS(resp, 79, 1); csd->write_misalign = UNSTUFF_BITS(resp, 78, 1); csd->read_misalign = UNSTUFF_BITS(resp, 77, 1); + csd->r2w_factor = UNSTUFF_BITS(resp, 26, 3); csd->write_blkbits = UNSTUFF_BITS(resp, 22, 4); csd->write_partial = UNSTUFF_BITS(resp, 21, 1); } diff --git a/drivers/mmc/mmc_block.c b/drivers/mmc/mmc_block.c index 8eb2a2ede64b..06bd1f4cb9b1 100644 --- a/drivers/mmc/mmc_block.c +++ b/drivers/mmc/mmc_block.c @@ -187,6 +187,12 @@ static int mmc_blk_issue_rq(struct mmc_queue *mq, struct request *req) brq.cmd.opcode = MMC_WRITE_BLOCK; brq.data.flags |= MMC_DATA_WRITE; brq.data.blocks = 1; + + /* + * Scale up the timeout by the r2w factor + */ + brq.data.timeout_ns <<= card->csd.r2w_factor; + brq.data.timeout_clks <<= card->csd.r2w_factor; } if (brq.data.blocks > 1) { diff --git a/include/linux/mmc/card.h b/include/linux/mmc/card.h index 30dd978c1ec8..991a37382a22 100644 --- a/include/linux/mmc/card.h +++ b/include/linux/mmc/card.h @@ -28,6 +28,7 @@ struct mmc_csd { unsigned short cmdclass; unsigned short tacc_clks; unsigned int tacc_ns; + unsigned int r2w_factor; unsigned int max_dtr; unsigned int read_blkbits; unsigned int write_blkbits; -- cgit v1.2.3 From 61f5657c50341198ff05e375e6f1fc0476556562 Mon Sep 17 00:00:00 2001 From: Vitaly Bordug Date: Sat, 29 Apr 2006 22:32:44 +0400 Subject: [PATCH] ppc32 CPM_UART: Fixed break send on SCC SCC uart sends a break sequence each time it is stopped with the CPM_CR_STOP_TX command. That means that each time an application closes the serial device, a break is transmitted. To fix this, graceful tx stop is issued for SCC. Signed-off-by: David Jander Signed-off-by: Vitaly Bordug Signed-off-by: Paul Mackerras --- drivers/serial/cpm_uart/cpm_uart_core.c | 6 +++++- include/asm-ppc/commproc.h | 1 + include/asm-ppc/cpm2.h | 2 +- 3 files changed, 7 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/serial/cpm_uart/cpm_uart_core.c b/drivers/serial/cpm_uart/cpm_uart_core.c index ced193bf9e1e..b9d1185df20c 100644 --- a/drivers/serial/cpm_uart/cpm_uart_core.c +++ b/drivers/serial/cpm_uart/cpm_uart_core.c @@ -457,7 +457,11 @@ static void cpm_uart_shutdown(struct uart_port *port) } /* Shut them really down and reinit buffer descriptors */ - cpm_line_cr_cmd(line, CPM_CR_STOP_TX); + if (IS_SMC(pinfo)) + cpm_line_cr_cmd(line, CPM_CR_STOP_TX); + else + cpm_line_cr_cmd(line, CPM_CR_GRA_STOP_TX); + cpm_uart_initbd(pinfo); } } diff --git a/include/asm-ppc/commproc.h b/include/asm-ppc/commproc.h index 973e60908234..31f362966a58 100644 --- a/include/asm-ppc/commproc.h +++ b/include/asm-ppc/commproc.h @@ -35,6 +35,7 @@ #define CPM_CR_INIT_TX ((ushort)0x0002) #define CPM_CR_HUNT_MODE ((ushort)0x0003) #define CPM_CR_STOP_TX ((ushort)0x0004) +#define CPM_CR_GRA_STOP_TX ((ushort)0x0005) #define CPM_CR_RESTART_TX ((ushort)0x0006) #define CPM_CR_CLOSE_RX_BD ((ushort)0x0007) #define CPM_CR_SET_GADDR ((ushort)0x0008) diff --git a/include/asm-ppc/cpm2.h b/include/asm-ppc/cpm2.h index b638b87cebe3..c70344b91049 100644 --- a/include/asm-ppc/cpm2.h +++ b/include/asm-ppc/cpm2.h @@ -69,7 +69,7 @@ #define CPM_CR_INIT_TX ((ushort)0x0002) #define CPM_CR_HUNT_MODE ((ushort)0x0003) #define CPM_CR_STOP_TX ((ushort)0x0004) -#define CPM_CR_GRA_STOP_TX ((ushort)0x0005) +#define CPM_CR_GRA_STOP_TX ((ushort)0x0005) #define CPM_CR_RESTART_TX ((ushort)0x0006) #define CPM_CR_SET_GADDR ((ushort)0x0008) #define CPM_CR_START_IDMA ((ushort)0x0009) -- cgit v1.2.3 From 6e1976961c9bd9a3dc368139fab1883961efc879 Mon Sep 17 00:00:00 2001 From: Vitaly Bordug Date: Sat, 29 Apr 2006 23:06:00 +0400 Subject: [PATCH] ppc32 CPM_UART: fixes and improvements A number of small issues are fixed, and added the header file, missed from the original series. With this, driver should be pretty stable as tested among both platform-device-driven and "old way" boards. Also added missing GPL statement , and updated year field on existing ones to reflect code update. Signed-off-by: Vitaly Bordug Signed-off-by: Paul Mackerras --- arch/ppc/platforms/mpc866ads_setup.c | 2 +- drivers/serial/cpm_uart/cpm_uart.h | 19 ++++++++--- drivers/serial/cpm_uart/cpm_uart_core.c | 31 +++++++++++++---- drivers/serial/cpm_uart/cpm_uart_cpm1.c | 2 ++ drivers/serial/cpm_uart/cpm_uart_cpm2.c | 2 ++ include/linux/fs_uart_pd.h | 60 +++++++++++++++++++++++++++++++++ 6 files changed, 104 insertions(+), 12 deletions(-) create mode 100644 include/linux/fs_uart_pd.h (limited to 'include') diff --git a/arch/ppc/platforms/mpc866ads_setup.c b/arch/ppc/platforms/mpc866ads_setup.c index 6ce3b842defe..d919dab61347 100644 --- a/arch/ppc/platforms/mpc866ads_setup.c +++ b/arch/ppc/platforms/mpc866ads_setup.c @@ -378,7 +378,7 @@ int __init mpc866ads_init(void) ppc_sys_device_setfunc(MPC8xx_CPM_SMC1, PPC_SYS_FUNC_UART); #endif -#ifdef CONFIG_SERIAL_CPM_SMCer +#ifdef CONFIG_SERIAL_CPM_SMC ppc_sys_device_enable(MPC8xx_CPM_SMC2); ppc_sys_device_setfunc(MPC8xx_CPM_SMC2, PPC_SYS_FUNC_UART); #endif diff --git a/drivers/serial/cpm_uart/cpm_uart.h b/drivers/serial/cpm_uart/cpm_uart.h index aa5eb7ddeda9..3b35cb779539 100644 --- a/drivers/serial/cpm_uart/cpm_uart.h +++ b/drivers/serial/cpm_uart/cpm_uart.h @@ -5,6 +5,13 @@ * * Copyright (C) 2004 Freescale Semiconductor, Inc. * + * 2006 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + * */ #ifndef CPM_UART_H #define CPM_UART_H @@ -101,12 +108,13 @@ static inline unsigned long cpu2cpm_addr(void* addr, struct uart_cpm_port *pinfo int offset; u32 val = (u32)addr; /* sane check */ - if ((val >= (u32)pinfo->mem_addr) && + if (likely((val >= (u32)pinfo->mem_addr)) && (val<((u32)pinfo->mem_addr + pinfo->mem_size))) { offset = val - (u32)pinfo->mem_addr; return pinfo->dma_addr+offset; } - printk("%s(): address %x to translate out of range!\n", __FUNCTION__, val); + /* something nasty happened */ + BUG(); return 0; } @@ -115,12 +123,13 @@ static inline void *cpm2cpu_addr(unsigned long addr, struct uart_cpm_port *pinfo int offset; u32 val = addr; /* sane check */ - if ((val >= pinfo->dma_addr) && - (val<(pinfo->dma_addr + pinfo->mem_size))) { + if (likely((val >= pinfo->dma_addr) && + (val<(pinfo->dma_addr + pinfo->mem_size)))) { offset = val - (u32)pinfo->dma_addr; return (void*)(pinfo->mem_addr+offset); } - printk("%s(): address %x to translate out of range!\n", __FUNCTION__, val); + /* something nasty happened */ + BUG(); return 0; } diff --git a/drivers/serial/cpm_uart/cpm_uart_core.c b/drivers/serial/cpm_uart/cpm_uart_core.c index b9d1185df20c..969f94900431 100644 --- a/drivers/serial/cpm_uart/cpm_uart_core.c +++ b/drivers/serial/cpm_uart/cpm_uart_core.c @@ -12,7 +12,8 @@ * * Copyright (C) 2004 Freescale Semiconductor, Inc. * (C) 2004 Intracom, S.A. - * (C) 2005 MontaVista Software, Inc. by Vitaly Bordug + * (C) 2005-2006 MontaVista Software, Inc. + * Vitaly Bordug * * 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 @@ -81,7 +82,7 @@ early_uart_get_pdev(int index) } -void cpm_uart_count(void) +static void cpm_uart_count(void) { cpm_uart_nr = 0; #ifdef CONFIG_SERIAL_CPM_SMC1 @@ -104,6 +105,21 @@ void cpm_uart_count(void) #endif } +/* Get UART number by its id */ +static int cpm_uart_id2nr(int id) +{ + int i; + if (id < UART_NR) { + for (i=0; ifs_no; line++); + line = cpm_uart_id2nr(idx); + if(line < 0) { + printk(KERN_ERR"%s(): port %d is not registered", __FUNCTION__, idx); + return -1; + } pinfo = (struct uart_cpm_port *) &cpm_uart_ports[idx]; @@ -1245,8 +1265,7 @@ static int cpm_uart_drv_probe(struct device *dev) } pdata = pdev->dev.platform_data; - pr_debug("cpm_uart_drv_probe: Adding CPM UART %d\n", - cpm_uart_port_map[pdata->fs_no]); + pr_debug("cpm_uart_drv_probe: Adding CPM UART %d\n", cpm_uart_id2nr(pdata->fs_no)); if ((ret = cpm_uart_drv_get_platform_data(pdev, 0))) return ret; @@ -1265,7 +1284,7 @@ static int cpm_uart_drv_remove(struct device *dev) struct fs_uart_platform_info *pdata = pdev->dev.platform_data; pr_debug("cpm_uart_drv_remove: Removing CPM UART %d\n", - cpm_uart_port_map[pdata->fs_no]); + cpm_uart_id2nr(pdata->fs_no)); uart_remove_one_port(&cpm_reg, &cpm_uart_ports[pdata->fs_no].port); return 0; diff --git a/drivers/serial/cpm_uart/cpm_uart_cpm1.c b/drivers/serial/cpm_uart/cpm_uart_cpm1.c index a5a30622637a..17406a05ce1f 100644 --- a/drivers/serial/cpm_uart/cpm_uart_cpm1.c +++ b/drivers/serial/cpm_uart/cpm_uart_cpm1.c @@ -8,6 +8,8 @@ * * Copyright (C) 2004 Freescale Semiconductor, Inc. * (C) 2004 Intracom, S.A. + * (C) 2006 MontaVista Software, Inc. + * Vitaly Bordug * * 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/serial/cpm_uart/cpm_uart_cpm2.c b/drivers/serial/cpm_uart/cpm_uart_cpm2.c index 7c6b07aeea92..4b2de08f46d0 100644 --- a/drivers/serial/cpm_uart/cpm_uart_cpm2.c +++ b/drivers/serial/cpm_uart/cpm_uart_cpm2.c @@ -8,6 +8,8 @@ * * Copyright (C) 2004 Freescale Semiconductor, Inc. * (C) 2004 Intracom, S.A. + * (C) 2006 MontaVista Software, Inc. + * Vitaly Bordug * * 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/include/linux/fs_uart_pd.h b/include/linux/fs_uart_pd.h new file mode 100644 index 000000000000..f5975126b712 --- /dev/null +++ b/include/linux/fs_uart_pd.h @@ -0,0 +1,60 @@ +/* + * Platform information definitions for the CPM Uart driver. + * + * 2006 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#ifndef FS_UART_PD_H +#define FS_UART_PD_H + +#include +#include + +enum fs_uart_id { + fsid_smc1_uart, + fsid_smc2_uart, + fsid_scc1_uart, + fsid_scc2_uart, + fsid_scc3_uart, + fsid_scc4_uart, + fs_uart_nr, +}; + +static inline int fs_uart_id_scc2fsid(int id) +{ + return fsid_scc1_uart + id - 1; +} + +static inline int fs_uart_id_fsid2scc(int id) +{ + return id - fsid_scc1_uart + 1; +} + +static inline int fs_uart_id_smc2fsid(int id) +{ + return fsid_smc1_uart + id - 1; +} + +static inline int fs_uart_id_fsid2smc(int id) +{ + return id - fsid_smc1_uart + 1; +} + +struct fs_uart_platform_info { + void(*init_ioports)(void); + /* device specific information */ + int fs_no; /* controller index */ + u32 uart_clk; + u8 tx_num_fifo; + u8 tx_buf_size; + u8 rx_num_fifo; + u8 rx_buf_size; + u8 brg; +}; + +#endif -- cgit v1.2.3 From 6bfd93c32a5065d0e858780b3beb0b667081601c Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Wed, 3 May 2006 23:02:04 +1000 Subject: powerpc: Fix incorrect might_sleep in __get_user/__put_user on kernel addresses We have a case where __get_user and __put_user can validly be used on kernel addresses in interrupt context - namely, the alignment exception handler, as our get/put_unaligned just do a single access and rely on the alignment exception handler to fix things up in the rare cases where the cpu can't handle it in hardware. Thus we can get alignment exceptions in the network stack at interrupt level. The alignment exception handler does a __get_user to read the instruction and blows up in might_sleep(). Since a __get_user on a kernel address won't actually ever sleep, this makes the might_sleep conditional on the address being less than PAGE_OFFSET. Signed-off-by: Paul Mackerras --- include/asm-powerpc/uaccess.h | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/asm-powerpc/uaccess.h b/include/asm-powerpc/uaccess.h index 3872e924cdd6..d83fc29c2bbf 100644 --- a/include/asm-powerpc/uaccess.h +++ b/include/asm-powerpc/uaccess.h @@ -7,6 +7,7 @@ #include #include #include +#include #define VERIFY_READ 0 #define VERIFY_WRITE 1 @@ -179,9 +180,11 @@ do { \ #define __put_user_nocheck(x, ptr, size) \ ({ \ long __pu_err; \ - might_sleep(); \ + __typeof__(*(ptr)) __user *__pu_addr = (ptr); \ + if (!is_kernel_addr((unsigned long)__pu_addr)) \ + might_sleep(); \ __chk_user_ptr(ptr); \ - __put_user_size((x), (ptr), (size), __pu_err); \ + __put_user_size((x), __pu_addr, (size), __pu_err); \ __pu_err; \ }) @@ -258,9 +261,11 @@ do { \ ({ \ long __gu_err; \ unsigned long __gu_val; \ + const __typeof__(*(ptr)) __user *__gu_addr = (ptr); \ __chk_user_ptr(ptr); \ - might_sleep(); \ - __get_user_size(__gu_val, (ptr), (size), __gu_err); \ + if (!is_kernel_addr((unsigned long)__gu_addr)) \ + might_sleep(); \ + __get_user_size(__gu_val, __gu_addr, (size), __gu_err); \ (x) = (__typeof__(*(ptr)))__gu_val; \ __gu_err; \ }) @@ -270,9 +275,11 @@ do { \ ({ \ long __gu_err; \ long long __gu_val; \ + const __typeof__(*(ptr)) __user *__gu_addr = (ptr); \ __chk_user_ptr(ptr); \ - might_sleep(); \ - __get_user_size(__gu_val, (ptr), (size), __gu_err); \ + if (!is_kernel_addr((unsigned long)__gu_addr)) \ + might_sleep(); \ + __get_user_size(__gu_val, __gu_addr, (size), __gu_err); \ (x) = (__typeof__(*(ptr)))__gu_val; \ __gu_err; \ }) -- cgit v1.2.3 From 8c45112b823972e9ff7bbca77dc592ca2224951b Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 3 May 2006 13:55:46 -0700 Subject: [SPARC]: Hook up vmsplice into syscall tables. Signed-off-by: David S. Miller --- arch/sparc/kernel/systbls.S | 2 +- arch/sparc64/kernel/sys32.S | 1 + arch/sparc64/kernel/systbls.S | 4 ++-- include/asm-sparc/unistd.h | 2 +- include/asm-sparc64/unistd.h | 2 +- 5 files changed, 6 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/systbls.S b/arch/sparc/kernel/systbls.S index db8faa75f94d..6e1135cc03b0 100644 --- a/arch/sparc/kernel/systbls.S +++ b/arch/sparc/kernel/systbls.S @@ -23,7 +23,7 @@ sys_call_table: /*10*/ .long sys_unlink, sunos_execv, sys_chdir, sys_chown16, sys_mknod /*15*/ .long sys_chmod, sys_lchown16, sparc_brk, sys_nis_syscall, sys_lseek /*20*/ .long sys_getpid, sys_capget, sys_capset, sys_setuid16, sys_getuid16 -/*25*/ .long sys_time, sys_ptrace, sys_alarm, sys_sigaltstack, sys_pause +/*25*/ .long sys_vmsplice, sys_ptrace, sys_alarm, sys_sigaltstack, sys_pause /*30*/ .long sys_utime, sys_lchown, sys_fchown, sys_access, sys_nice /*35*/ .long sys_chown, sys_sync, sys_kill, sys_newstat, sys_sendfile /*40*/ .long sys_newlstat, sys_dup, sys_pipe, sys_times, sys_getuid diff --git a/arch/sparc64/kernel/sys32.S b/arch/sparc64/kernel/sys32.S index f9b75760163c..bdf1f4d02e3f 100644 --- a/arch/sparc64/kernel/sys32.S +++ b/arch/sparc64/kernel/sys32.S @@ -139,6 +139,7 @@ SIGN3(sys32_ioprio_set, sys_ioprio_set, %o0, %o1, %o2) SIGN2(sys32_splice, sys_splice, %o0, %o1) SIGN2(sys32_sync_file_range, compat_sync_file_range, %o0, %o5) SIGN2(sys32_tee, sys_tee, %o0, %o1) +SIGN1(sys32_vmsplice, compat_sys_vmsplice, %o0) .globl sys32_mmap2 sys32_mmap2: diff --git a/arch/sparc64/kernel/systbls.S b/arch/sparc64/kernel/systbls.S index 62672cd92eca..d4b39cd30310 100644 --- a/arch/sparc64/kernel/systbls.S +++ b/arch/sparc64/kernel/systbls.S @@ -25,7 +25,7 @@ sys_call_table32: /*10*/ .word sys_unlink, sunos_execv, sys_chdir, sys32_chown16, sys32_mknod /*15*/ .word sys_chmod, sys32_lchown16, sparc_brk, sys32_perfctr, sys32_lseek /*20*/ .word sys_getpid, sys_capget, sys_capset, sys32_setuid16, sys32_getuid16 -/*25*/ .word compat_sys_time, sys_ptrace, sys_alarm, sys32_sigaltstack, sys32_pause +/*25*/ .word sys32_vmsplice, sys_ptrace, sys_alarm, sys32_sigaltstack, sys32_pause /*30*/ .word compat_sys_utime, sys_lchown, sys_fchown, sys32_access, sys32_nice .word sys_chown, sys_sync, sys32_kill, compat_sys_newstat, sys32_sendfile /*40*/ .word compat_sys_newlstat, sys_dup, sys_pipe, compat_sys_times, sys_getuid @@ -94,7 +94,7 @@ sys_call_table: /*10*/ .word sys_unlink, sys_nis_syscall, sys_chdir, sys_chown, sys_mknod /*15*/ .word sys_chmod, sys_lchown, sparc_brk, sys_perfctr, sys_lseek /*20*/ .word sys_getpid, sys_capget, sys_capset, sys_setuid, sys_getuid -/*25*/ .word sys_nis_syscall, sys_ptrace, sys_alarm, sys_sigaltstack, sys_nis_syscall +/*25*/ .word sys_vmsplice, sys_ptrace, sys_alarm, sys_sigaltstack, sys_nis_syscall /*30*/ .word sys_utime, sys_nis_syscall, sys_nis_syscall, sys_access, sys_nice .word sys_nis_syscall, sys_sync, sys_kill, sys_newstat, sys_sendfile64 /*40*/ .word sys_newlstat, sys_dup, sys_pipe, sys_times, sys_nis_syscall diff --git a/include/asm-sparc/unistd.h b/include/asm-sparc/unistd.h index 32a48f623e2b..f5611a721fbd 100644 --- a/include/asm-sparc/unistd.h +++ b/include/asm-sparc/unistd.h @@ -41,7 +41,7 @@ #define __NR_capset 22 /* Linux Specific */ #define __NR_setuid 23 /* Implemented via setreuid in SunOS */ #define __NR_getuid 24 /* Common */ -/* #define __NR_time alias 25 ENOSYS under SunOS */ +#define __NR_vmsplice 25 /* ENOSYS under SunOS */ #define __NR_ptrace 26 /* Common */ #define __NR_alarm 27 /* Implemented via setitimer in SunOS */ #define __NR_sigaltstack 28 /* Common */ diff --git a/include/asm-sparc64/unistd.h b/include/asm-sparc64/unistd.h index ca80e8aca128..68705748bec0 100644 --- a/include/asm-sparc64/unistd.h +++ b/include/asm-sparc64/unistd.h @@ -41,7 +41,7 @@ #define __NR_capset 22 /* Linux Specific */ #define __NR_setuid 23 /* Implemented via setreuid in SunOS */ #define __NR_getuid 24 /* Common */ -/* #define __NR_time alias 25 ENOSYS under SunOS */ +#define __NR_vmsplice 25 /* ENOSYS under SunOS */ #define __NR_ptrace 26 /* Common */ #define __NR_alarm 27 /* Implemented via setitimer in SunOS */ #define __NR_sigaltstack 28 /* Common */ -- cgit v1.2.3 From 1432873af7ae29d4bb3c56114c05b539d078ca62 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Wed, 3 May 2006 10:35:26 +0200 Subject: [PATCH] splice: LRU fixups Nick says that the current construct isn't safe. This goes back to the original, but sets PIPE_BUF_FLAG_LRU on user pages as well as they all seem to be on the LRU in the first place. Signed-off-by: Jens Axboe --- fs/splice.c | 33 +++++++++++---------------------- include/linux/pipe_fs_i.h | 5 +++-- 2 files changed, 14 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/fs/splice.c b/fs/splice.c index 27f5e3738a7b..0b202425b0b5 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -78,6 +78,7 @@ static int page_cache_pipe_buf_steal(struct pipe_inode_info *info, return 1; } + buf->flags |= PIPE_BUF_FLAG_LRU; return 0; } @@ -85,6 +86,7 @@ static void page_cache_pipe_buf_release(struct pipe_inode_info *info, struct pipe_buffer *buf) { page_cache_release(buf->page); + buf->flags &= ~PIPE_BUF_FLAG_LRU; } static int page_cache_pipe_buf_pin(struct pipe_inode_info *info, @@ -141,6 +143,7 @@ static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, if (!(buf->flags & PIPE_BUF_FLAG_GIFT)) return 1; + buf->flags |= PIPE_BUF_FLAG_LRU; return generic_pipe_buf_steal(pipe, buf); } @@ -566,37 +569,23 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, */ if ((sd->flags & SPLICE_F_MOVE) && this_len == PAGE_CACHE_SIZE) { /* - * If steal succeeds, buf->page is now pruned from the vm - * side (page cache) and we can reuse it. The page will also - * be locked on successful return. + * If steal succeeds, buf->page is now pruned from the + * pagecache and we can reuse it. The page will also be + * locked on successful return. */ if (buf->ops->steal(info, buf)) goto find_page; page = buf->page; - page_cache_get(page); - - /* - * page must be on the LRU for adding to the pagecache. - * Check this without grabbing the zone lock, if it isn't - * the do grab the zone lock, recheck, and add if necessary. - */ - if (!PageLRU(page)) { - struct zone *zone = page_zone(page); - - spin_lock_irq(&zone->lru_lock); - if (!PageLRU(page)) { - SetPageLRU(page); - add_page_to_inactive_list(zone, page); - } - spin_unlock_irq(&zone->lru_lock); - } - if (add_to_page_cache(page, mapping, index, gfp_mask)) { - page_cache_release(page); unlock_page(page); goto find_page; } + + page_cache_get(page); + + if (!(buf->flags & PIPE_BUF_FLAG_LRU)) + lru_cache_add(page); } else { find_page: page = find_lock_page(mapping, index); diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index ba73108cbf8b..ea4f7cd7bfd8 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -5,8 +5,9 @@ #define PIPE_BUFFERS (16) -#define PIPE_BUF_FLAG_ATOMIC 0x01 /* was atomically mapped */ -#define PIPE_BUF_FLAG_GIFT 0x02 /* page is a gift */ +#define PIPE_BUF_FLAG_LRU 0x01 /* page is on the LRU */ +#define PIPE_BUF_FLAG_ATOMIC 0x02 /* was atomically mapped */ +#define PIPE_BUF_FLAG_GIFT 0x04 /* page is a gift */ struct pipe_buffer { struct page *page; -- cgit v1.2.3 From 7582e9d17edbabab6cbe59467c5d1b5e8c04fca8 Mon Sep 17 00:00:00 2001 From: Jing Min Zhao Date: Wed, 3 May 2006 23:19:59 -0700 Subject: [NETFILTER]: H.323 helper: Change author's email address Signed-off-by: Jing Min Zhao Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller --- include/linux/netfilter_ipv4/ip_conntrack_helper_h323_asn1.h | 2 +- net/ipv4/netfilter/ip_conntrack_helper_h323_asn1.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter_ipv4/ip_conntrack_helper_h323_asn1.h b/include/linux/netfilter_ipv4/ip_conntrack_helper_h323_asn1.h index 0bd828081c0c..c6e9a0b6d30b 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_helper_h323_asn1.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_helper_h323_asn1.h @@ -2,7 +2,7 @@ * ip_conntrack_helper_h323_asn1.h - BER and PER decoding library for H.323 * conntrack/NAT module. * - * Copyright (c) 2006 by Jing Min Zhao + * Copyright (c) 2006 by Jing Min Zhao * * This source code is licensed under General Public License version 2. * diff --git a/net/ipv4/netfilter/ip_conntrack_helper_h323_asn1.c b/net/ipv4/netfilter/ip_conntrack_helper_h323_asn1.c index f52d2c45a6ae..355a53a5b6cd 100644 --- a/net/ipv4/netfilter/ip_conntrack_helper_h323_asn1.c +++ b/net/ipv4/netfilter/ip_conntrack_helper_h323_asn1.c @@ -2,7 +2,7 @@ * ip_conntrack_helper_h323_asn1.c - BER and PER decoding library for H.323 * conntrack/NAT module. * - * Copyright (c) 2006 by Jing Min Zhao + * Copyright (c) 2006 by Jing Min Zhao * * This source code is licensed under General Public License version 2. * -- cgit v1.2.3 From e1fdb5b39656ea2be8cadde565e543649a988af9 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Wed, 3 May 2006 23:27:16 -0700 Subject: [AX.25]: Eleminate HZ from AX.25 kernel interfaces Convert all AX.25 sysctl time values from jiffies to ms as units. Signed-off-by: Ralf Baechle Signed-off-by: David S. Miller --- include/net/ax25.h | 10 +++---- net/ax25/af_ax25.c | 73 ++++++++++++++++++++++++++-------------------- net/ax25/ax25_ds_timer.c | 3 +- net/ax25/sysctl_net_ax25.c | 10 +++---- 4 files changed, 53 insertions(+), 43 deletions(-) (limited to 'include') diff --git a/include/net/ax25.h b/include/net/ax25.h index d052b221dbcd..5bd997487054 100644 --- a/include/net/ax25.h +++ b/include/net/ax25.h @@ -145,14 +145,14 @@ enum { #define AX25_DEF_CONMODE 2 /* Connected mode allowed */ #define AX25_DEF_WINDOW 2 /* Window=2 */ #define AX25_DEF_EWINDOW 32 /* Module-128 Window=32 */ -#define AX25_DEF_T1 (10 * HZ) /* T1=10s */ -#define AX25_DEF_T2 (3 * HZ) /* T2=3s */ -#define AX25_DEF_T3 (300 * HZ) /* T3=300s */ +#define AX25_DEF_T1 10000 /* T1=10s */ +#define AX25_DEF_T2 3000 /* T2=3s */ +#define AX25_DEF_T3 300000 /* T3=300s */ #define AX25_DEF_N2 10 /* N2=10 */ -#define AX25_DEF_IDLE (0 * 60 * HZ) /* Idle=None */ +#define AX25_DEF_IDLE 0 /* Idle=None */ #define AX25_DEF_PACLEN 256 /* Paclen=256 */ #define AX25_DEF_PROTOCOL AX25_PROTO_STD_SIMPLEX /* Standard AX.25 */ -#define AX25_DEF_DS_TIMEOUT (3 * 60 * HZ) /* DAMA timeout 3 minutes */ +#define AX25_DEF_DS_TIMEOUT 180000 /* DAMA timeout 3 minutes */ typedef struct ax25_uid_assoc { struct hlist_node uid_node; diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c index a9f13dfde07e..a2e0dd047e9f 100644 --- a/net/ax25/af_ax25.c +++ b/net/ax25/af_ax25.c @@ -426,6 +426,26 @@ static int ax25_ctl_ioctl(const unsigned int cmd, void __user *arg) return 0; } +static void ax25_fillin_cb_from_dev(ax25_cb *ax25, ax25_dev *ax25_dev) +{ + ax25->rtt = msecs_to_jiffies(ax25_dev->values[AX25_VALUES_T1]) / 2; + ax25->t1 = msecs_to_jiffies(ax25_dev->values[AX25_VALUES_T1]); + ax25->t2 = msecs_to_jiffies(ax25_dev->values[AX25_VALUES_T2]); + ax25->t3 = msecs_to_jiffies(ax25_dev->values[AX25_VALUES_T3]); + ax25->n2 = ax25_dev->values[AX25_VALUES_N2]; + ax25->paclen = ax25_dev->values[AX25_VALUES_PACLEN]; + ax25->idle = msecs_to_jiffies(ax25_dev->values[AX25_VALUES_IDLE]); + ax25->backoff = ax25_dev->values[AX25_VALUES_BACKOFF]; + + if (ax25_dev->values[AX25_VALUES_AXDEFMODE]) { + ax25->modulus = AX25_EMODULUS; + ax25->window = ax25_dev->values[AX25_VALUES_EWINDOW]; + } else { + ax25->modulus = AX25_MODULUS; + ax25->window = ax25_dev->values[AX25_VALUES_WINDOW]; + } +} + /* * Fill in a created AX.25 created control block with the default * values for a particular device. @@ -435,39 +455,28 @@ void ax25_fillin_cb(ax25_cb *ax25, ax25_dev *ax25_dev) ax25->ax25_dev = ax25_dev; if (ax25->ax25_dev != NULL) { - ax25->rtt = ax25_dev->values[AX25_VALUES_T1] / 2; - ax25->t1 = ax25_dev->values[AX25_VALUES_T1]; - ax25->t2 = ax25_dev->values[AX25_VALUES_T2]; - ax25->t3 = ax25_dev->values[AX25_VALUES_T3]; - ax25->n2 = ax25_dev->values[AX25_VALUES_N2]; - ax25->paclen = ax25_dev->values[AX25_VALUES_PACLEN]; - ax25->idle = ax25_dev->values[AX25_VALUES_IDLE]; - ax25->backoff = ax25_dev->values[AX25_VALUES_BACKOFF]; - - if (ax25_dev->values[AX25_VALUES_AXDEFMODE]) { - ax25->modulus = AX25_EMODULUS; - ax25->window = ax25_dev->values[AX25_VALUES_EWINDOW]; - } else { - ax25->modulus = AX25_MODULUS; - ax25->window = ax25_dev->values[AX25_VALUES_WINDOW]; - } + ax25_fillin_cb_from_dev(ax25, ax25_dev); + return; + } + + /* + * No device, use kernel / AX.25 spec default values + */ + ax25->rtt = msecs_to_jiffies(AX25_DEF_T1) / 2; + ax25->t1 = msecs_to_jiffies(AX25_DEF_T1); + ax25->t2 = msecs_to_jiffies(AX25_DEF_T2); + ax25->t3 = msecs_to_jiffies(AX25_DEF_T3); + ax25->n2 = AX25_DEF_N2; + ax25->paclen = AX25_DEF_PACLEN; + ax25->idle = msecs_to_jiffies(AX25_DEF_IDLE); + ax25->backoff = AX25_DEF_BACKOFF; + + if (AX25_DEF_AXDEFMODE) { + ax25->modulus = AX25_EMODULUS; + ax25->window = AX25_DEF_EWINDOW; } else { - ax25->rtt = AX25_DEF_T1 / 2; - ax25->t1 = AX25_DEF_T1; - ax25->t2 = AX25_DEF_T2; - ax25->t3 = AX25_DEF_T3; - ax25->n2 = AX25_DEF_N2; - ax25->paclen = AX25_DEF_PACLEN; - ax25->idle = AX25_DEF_IDLE; - ax25->backoff = AX25_DEF_BACKOFF; - - if (AX25_DEF_AXDEFMODE) { - ax25->modulus = AX25_EMODULUS; - ax25->window = AX25_DEF_EWINDOW; - } else { - ax25->modulus = AX25_MODULUS; - ax25->window = AX25_DEF_WINDOW; - } + ax25->modulus = AX25_MODULUS; + ax25->window = AX25_DEF_WINDOW; } } diff --git a/net/ax25/ax25_ds_timer.c b/net/ax25/ax25_ds_timer.c index 061083efc1dc..5961459935eb 100644 --- a/net/ax25/ax25_ds_timer.c +++ b/net/ax25/ax25_ds_timer.c @@ -61,7 +61,8 @@ void ax25_ds_set_timer(ax25_dev *ax25_dev) return; del_timer(&ax25_dev->dama.slave_timer); - ax25_dev->dama.slave_timeout = ax25_dev->values[AX25_VALUES_DS_TIMEOUT] / 10; + ax25_dev->dama.slave_timeout = + msecs_to_jiffies(ax25_dev->values[AX25_VALUES_DS_TIMEOUT]) / 10; ax25_ds_add_timer(ax25_dev); } diff --git a/net/ax25/sysctl_net_ax25.c b/net/ax25/sysctl_net_ax25.c index 894a22558d9d..bdb64c36df12 100644 --- a/net/ax25/sysctl_net_ax25.c +++ b/net/ax25/sysctl_net_ax25.c @@ -18,14 +18,14 @@ static int min_backoff[1], max_backoff[] = {2}; static int min_conmode[1], max_conmode[] = {2}; static int min_window[] = {1}, max_window[] = {7}; static int min_ewindow[] = {1}, max_ewindow[] = {63}; -static int min_t1[] = {1}, max_t1[] = {30 * HZ}; -static int min_t2[] = {1}, max_t2[] = {20 * HZ}; -static int min_t3[1], max_t3[] = {3600 * HZ}; -static int min_idle[1], max_idle[] = {65535 * HZ}; +static int min_t1[] = {1}, max_t1[] = {30000}; +static int min_t2[] = {1}, max_t2[] = {20000}; +static int min_t3[1], max_t3[] = {3600000}; +static int min_idle[1], max_idle[] = {65535000}; static int min_n2[] = {1}, max_n2[] = {31}; static int min_paclen[] = {1}, max_paclen[] = {512}; static int min_proto[1], max_proto[] = { AX25_PROTO_MAX }; -static int min_ds_timeout[1], max_ds_timeout[] = {65535 * HZ}; +static int min_ds_timeout[1], max_ds_timeout[] = {65535000}; static struct ctl_table_header *ax25_table_header; -- cgit v1.2.3 From 4d8937d0b113e8ec39f7d18cf13804f3b5fb8fd4 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Wed, 3 May 2006 23:27:47 -0700 Subject: [NETROM]: Eleminate HZ from NET/ROM kernel interfaces Convert all NET/ROM sysctl time values from jiffies to ms as units. Signed-off-by: Ralf Baechle Signed-off-by: David S. Miller --- include/net/netrom.h | 8 ++++---- net/netrom/af_netrom.c | 15 ++++++++++----- 2 files changed, 14 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/net/netrom.h b/include/net/netrom.h index a5ee53bce62f..e0ca112024a3 100644 --- a/include/net/netrom.h +++ b/include/net/netrom.h @@ -42,11 +42,11 @@ enum { #define NR_COND_PEER_RX_BUSY 0x04 #define NR_COND_OWN_RX_BUSY 0x08 -#define NR_DEFAULT_T1 (120 * HZ) /* Outstanding frames - 120 seconds */ -#define NR_DEFAULT_T2 (5 * HZ) /* Response delay - 5 seconds */ +#define NR_DEFAULT_T1 120000 /* Outstanding frames - 120 seconds */ +#define NR_DEFAULT_T2 5000 /* Response delay - 5 seconds */ #define NR_DEFAULT_N2 3 /* Number of Retries - 3 */ -#define NR_DEFAULT_T4 (180 * HZ) /* Busy Delay - 180 seconds */ -#define NR_DEFAULT_IDLE (0 * 60 * HZ) /* No Activity Timeout - none */ +#define NR_DEFAULT_T4 180000 /* Busy Delay - 180 seconds */ +#define NR_DEFAULT_IDLE 0 /* No Activity Timeout - none */ #define NR_DEFAULT_WINDOW 4 /* Default Window Size - 4 */ #define NR_DEFAULT_OBS 6 /* Default Obsolescence Count - 6 */ #define NR_DEFAULT_QUAL 10 /* Default Neighbour Quality - 10 */ diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c index d44981f5a619..ecd288beca7c 100644 --- a/net/netrom/af_netrom.c +++ b/net/netrom/af_netrom.c @@ -425,11 +425,16 @@ static int nr_create(struct socket *sock, int protocol) nr_init_timers(sk); - nr->t1 = sysctl_netrom_transport_timeout; - nr->t2 = sysctl_netrom_transport_acknowledge_delay; - nr->n2 = sysctl_netrom_transport_maximum_tries; - nr->t4 = sysctl_netrom_transport_busy_delay; - nr->idle = sysctl_netrom_transport_no_activity_timeout; + nr->t1 = + msecs_to_jiffies(sysctl_netrom_transport_timeout); + nr->t2 = + msecs_to_jiffies(sysctl_netrom_transport_acknowledge_delay); + nr->n2 = + msecs_to_jiffies(sysctl_netrom_transport_maximum_tries); + nr->t4 = + msecs_to_jiffies(sysctl_netrom_transport_busy_delay); + nr->idle = + msecs_to_jiffies(sysctl_netrom_transport_no_activity_timeout); nr->window = sysctl_netrom_transport_requested_window_size; nr->bpqext = 1; -- cgit v1.2.3 From 82e84249f0ee098e004c8bd6d90a1640bd56cfbb Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Wed, 3 May 2006 23:28:20 -0700 Subject: [ROSE]: Eleminate HZ from ROSE kernel interfaces Convert all ROSE sysctl time values from jiffies to ms as units. Signed-off-by: Ralf Baechle Signed-off-by: David S. Miller --- include/net/rose.h | 14 +++++++------- net/rose/af_rose.c | 10 +++++----- net/rose/rose_link.c | 6 ++++-- 3 files changed, 16 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/net/rose.h b/include/net/rose.h index 3249b979605a..012b09ed2401 100644 --- a/include/net/rose.h +++ b/include/net/rose.h @@ -49,14 +49,14 @@ enum { ROSE_STATE_5 /* Deferred Call Acceptance */ }; -#define ROSE_DEFAULT_T0 (180 * HZ) /* Default T10 T20 value */ -#define ROSE_DEFAULT_T1 (200 * HZ) /* Default T11 T21 value */ -#define ROSE_DEFAULT_T2 (180 * HZ) /* Default T12 T22 value */ -#define ROSE_DEFAULT_T3 (180 * HZ) /* Default T13 T23 value */ -#define ROSE_DEFAULT_HB (5 * HZ) /* Default Holdback value */ -#define ROSE_DEFAULT_IDLE (0 * 60 * HZ) /* No Activity Timeout - none */ +#define ROSE_DEFAULT_T0 180000 /* Default T10 T20 value */ +#define ROSE_DEFAULT_T1 200000 /* Default T11 T21 value */ +#define ROSE_DEFAULT_T2 180000 /* Default T12 T22 value */ +#define ROSE_DEFAULT_T3 180000 /* Default T13 T23 value */ +#define ROSE_DEFAULT_HB 5000 /* Default Holdback value */ +#define ROSE_DEFAULT_IDLE 0 /* No Activity Timeout - none */ #define ROSE_DEFAULT_ROUTING 1 /* Default routing flag */ -#define ROSE_DEFAULT_FAIL_TIMEOUT (120 * HZ) /* Time until link considered usable */ +#define ROSE_DEFAULT_FAIL_TIMEOUT 120000 /* Time until link considered usable */ #define ROSE_DEFAULT_MAXVC 50 /* Maximum number of VCs per neighbour */ #define ROSE_DEFAULT_WINDOW_SIZE 7 /* Default window size */ diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c index ea65396d1619..ef4538ac84f0 100644 --- a/net/rose/af_rose.c +++ b/net/rose/af_rose.c @@ -518,11 +518,11 @@ static int rose_create(struct socket *sock, int protocol) init_timer(&rose->timer); init_timer(&rose->idletimer); - rose->t1 = sysctl_rose_call_request_timeout; - rose->t2 = sysctl_rose_reset_request_timeout; - rose->t3 = sysctl_rose_clear_request_timeout; - rose->hb = sysctl_rose_ack_hold_back_timeout; - rose->idle = sysctl_rose_no_activity_timeout; + rose->t1 = msecs_to_jiffies(sysctl_rose_call_request_timeout); + rose->t2 = msecs_to_jiffies(sysctl_rose_reset_request_timeout); + rose->t3 = msecs_to_jiffies(sysctl_rose_clear_request_timeout); + rose->hb = msecs_to_jiffies(sysctl_rose_ack_hold_back_timeout); + rose->idle = msecs_to_jiffies(sysctl_rose_no_activity_timeout); rose->state = ROSE_STATE_0; diff --git a/net/rose/rose_link.c b/net/rose/rose_link.c index 09e9e9d04d92..bd86a63960ce 100644 --- a/net/rose/rose_link.c +++ b/net/rose/rose_link.c @@ -40,7 +40,8 @@ void rose_start_ftimer(struct rose_neigh *neigh) neigh->ftimer.data = (unsigned long)neigh; neigh->ftimer.function = &rose_ftimer_expiry; - neigh->ftimer.expires = jiffies + sysctl_rose_link_fail_timeout; + neigh->ftimer.expires = + jiffies + msecs_to_jiffies(sysctl_rose_link_fail_timeout); add_timer(&neigh->ftimer); } @@ -51,7 +52,8 @@ static void rose_start_t0timer(struct rose_neigh *neigh) neigh->t0timer.data = (unsigned long)neigh; neigh->t0timer.function = &rose_t0timer_expiry; - neigh->t0timer.expires = jiffies + sysctl_rose_restart_request_timeout; + neigh->t0timer.expires = + jiffies + msecs_to_jiffies(sysctl_rose_restart_request_timeout); add_timer(&neigh->t0timer); } -- cgit v1.2.3 From 5b802344357338a4d645beac8ca97470bcbe3542 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 4 May 2006 14:07:42 +0100 Subject: [ARM] 3490/1: i.MX: move uart resources to board files Patch from Sascha Hauer This patch moves the i.MX uart resources and the gpio pin setup to the board files. This allows the boards to decide how many internal uarts are connected to the outside world and whether they use rts/cts or not. Signed-off-by: Sascha Hauer Signed-off-by: Russell King --- arch/arm/mach-imx/generic.c | 52 -------------------------- arch/arm/mach-imx/mx1ads.c | 74 +++++++++++++++++++++++++++++++++++++ drivers/serial/imx.c | 40 ++++++++------------ include/asm-arm/arch-imx/imx-uart.h | 10 +++++ 4 files changed, 100 insertions(+), 76 deletions(-) create mode 100644 include/asm-arm/arch-imx/imx-uart.h (limited to 'include') diff --git a/arch/arm/mach-imx/generic.c b/arch/arm/mach-imx/generic.c index 9d8331be2b58..12ea58a3b84f 100644 --- a/arch/arm/mach-imx/generic.c +++ b/arch/arm/mach-imx/generic.c @@ -195,56 +195,6 @@ void __init imx_set_mmc_info(struct imxmmc_platform_data *info) } EXPORT_SYMBOL(imx_set_mmc_info); -static struct resource imx_uart1_resources[] = { - [0] = { - .start = 0x00206000, - .end = 0x002060FF, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = (UART1_MINT_RX), - .end = (UART1_MINT_RX), - .flags = IORESOURCE_IRQ, - }, - [2] = { - .start = (UART1_MINT_TX), - .end = (UART1_MINT_TX), - .flags = IORESOURCE_IRQ, - }, -}; - -static struct platform_device imx_uart1_device = { - .name = "imx-uart", - .id = 0, - .num_resources = ARRAY_SIZE(imx_uart1_resources), - .resource = imx_uart1_resources, -}; - -static struct resource imx_uart2_resources[] = { - [0] = { - .start = 0x00207000, - .end = 0x002070FF, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = (UART2_MINT_RX), - .end = (UART2_MINT_RX), - .flags = IORESOURCE_IRQ, - }, - [2] = { - .start = (UART2_MINT_TX), - .end = (UART2_MINT_TX), - .flags = IORESOURCE_IRQ, - }, -}; - -static struct platform_device imx_uart2_device = { - .name = "imx-uart", - .id = 1, - .num_resources = ARRAY_SIZE(imx_uart2_resources), - .resource = imx_uart2_resources, -}; - static struct imxfb_mach_info imx_fb_info; void __init set_imx_fb_info(struct imxfb_mach_info *hard_imx_fb_info) @@ -283,8 +233,6 @@ static struct platform_device imxfb_device = { static struct platform_device *devices[] __initdata = { &imx_mmc_device, &imxfb_device, - &imx_uart1_device, - &imx_uart2_device, }; static struct map_desc imx_io_desc[] __initdata = { diff --git a/arch/arm/mach-imx/mx1ads.c b/arch/arm/mach-imx/mx1ads.c index e34d0df90aed..e1f6c0bbe1e7 100644 --- a/arch/arm/mach-imx/mx1ads.c +++ b/arch/arm/mach-imx/mx1ads.c @@ -26,6 +26,7 @@ #include #include +#include #include #include "generic.h" @@ -48,8 +49,70 @@ static struct platform_device cs89x0_device = { .resource = cs89x0_resources, }; +static struct imxuart_platform_data uart_pdata = { + .flags = IMXUART_HAVE_RTSCTS, +}; + +static struct resource imx_uart1_resources[] = { + [0] = { + .start = 0x00206000, + .end = 0x002060FF, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = (UART1_MINT_RX), + .end = (UART1_MINT_RX), + .flags = IORESOURCE_IRQ, + }, + [2] = { + .start = (UART1_MINT_TX), + .end = (UART1_MINT_TX), + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device imx_uart1_device = { + .name = "imx-uart", + .id = 0, + .num_resources = ARRAY_SIZE(imx_uart1_resources), + .resource = imx_uart1_resources, + .dev = { + .platform_data = &uart_pdata, + } +}; + +static struct resource imx_uart2_resources[] = { + [0] = { + .start = 0x00207000, + .end = 0x002070FF, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = (UART2_MINT_RX), + .end = (UART2_MINT_RX), + .flags = IORESOURCE_IRQ, + }, + [2] = { + .start = (UART2_MINT_TX), + .end = (UART2_MINT_TX), + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device imx_uart2_device = { + .name = "imx-uart", + .id = 1, + .num_resources = ARRAY_SIZE(imx_uart2_resources), + .resource = imx_uart2_resources, + .dev = { + .platform_data = &uart_pdata, + } +}; + static struct platform_device *devices[] __initdata = { &cs89x0_device, + &imx_uart1_device, + &imx_uart2_device, }; #ifdef CONFIG_MMC_IMX @@ -75,6 +138,17 @@ mx1ads_init(void) imx_gpio_mode(GPIO_PORTB | GPIO_GIUS | GPIO_IN | 20); imx_set_mmc_info(&mx1ads_mmc_info); #endif + + imx_gpio_mode(PC9_PF_UART1_CTS); + imx_gpio_mode(PC10_PF_UART1_RTS); + imx_gpio_mode(PC11_PF_UART1_TXD); + imx_gpio_mode(PC12_PF_UART1_RXD); + + imx_gpio_mode(PB28_PF_UART2_CTS); + imx_gpio_mode(PB29_PF_UART2_RTS); + imx_gpio_mode(PB30_PF_UART2_TXD); + imx_gpio_mode(PB31_PF_UART2_RXD); + platform_add_devices(devices, ARRAY_SIZE(devices)); } diff --git a/drivers/serial/imx.c b/drivers/serial/imx.c index c3b7a6673e9c..d202eb4f3848 100644 --- a/drivers/serial/imx.c +++ b/drivers/serial/imx.c @@ -45,6 +45,7 @@ #include #include #include +#include /* We've been assigned a range on the "Low-density serial ports" major */ #define SERIAL_IMX_MAJOR 204 @@ -73,7 +74,8 @@ struct imx_port { struct uart_port port; struct timer_list timer; unsigned int old_status; - int txirq,rxirq,rtsirq; + int txirq,rxirq,rtsirq; + int have_rtscts:1; }; /* @@ -491,8 +493,12 @@ imx_set_termios(struct uart_port *port, struct termios *termios, ucr2 = UCR2_SRST | UCR2_IRTS; if (termios->c_cflag & CRTSCTS) { - ucr2 &= ~UCR2_IRTS; - ucr2 |= UCR2_CTSC; + if( sport->have_rtscts ) { + ucr2 &= ~UCR2_IRTS; + ucr2 |= UCR2_CTSC; + } else { + termios->c_cflag &= ~CRTSCTS; + } } if (termios->c_cflag & CSTOPB) @@ -719,27 +725,6 @@ static void __init imx_init_ports(void) imx_ports[i].timer.function = imx_timeout; imx_ports[i].timer.data = (unsigned long)&imx_ports[i]; } - - imx_gpio_mode(PC9_PF_UART1_CTS); - imx_gpio_mode(PC10_PF_UART1_RTS); - imx_gpio_mode(PC11_PF_UART1_TXD); - imx_gpio_mode(PC12_PF_UART1_RXD); - imx_gpio_mode(PB28_PF_UART2_CTS); - imx_gpio_mode(PB29_PF_UART2_RTS); - - imx_gpio_mode(PB30_PF_UART2_TXD); - imx_gpio_mode(PB31_PF_UART2_RXD); - -#if 0 /* We don't need these, on the mx1 the _modem_ side of the uart - * is implemented. - */ - imx_gpio_mode(PD7_AF_UART2_DTR); - imx_gpio_mode(PD8_AF_UART2_DCD); - imx_gpio_mode(PD9_AF_UART2_RI); - imx_gpio_mode(PD10_AF_UART2_DSR); -#endif - - } #ifdef CONFIG_SERIAL_IMX_CONSOLE @@ -932,7 +917,14 @@ static int serial_imx_resume(struct platform_device *dev) static int serial_imx_probe(struct platform_device *dev) { + struct imxuart_platform_data *pdata; + imx_ports[dev->id].port.dev = &dev->dev; + + pdata = (struct imxuart_platform_data *)dev->dev.platform_data; + if(pdata && (pdata->flags & IMXUART_HAVE_RTSCTS)) + imx_ports[dev->id].have_rtscts = 1; + uart_add_one_port(&imx_reg, &imx_ports[dev->id].port); platform_set_drvdata(dev, &imx_ports[dev->id]); return 0; diff --git a/include/asm-arm/arch-imx/imx-uart.h b/include/asm-arm/arch-imx/imx-uart.h new file mode 100644 index 000000000000..3a685e1780ea --- /dev/null +++ b/include/asm-arm/arch-imx/imx-uart.h @@ -0,0 +1,10 @@ +#ifndef ASMARM_ARCH_UART_H +#define ASMARM_ARCH_UART_H + +#define IMXUART_HAVE_RTSCTS (1<<0) + +struct imxuart_platform_data { + unsigned int flags; +}; + +#endif -- cgit v1.2.3 From ff10952a547dad934d9ed9afc5cf579ed1ccb53a Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Fri, 5 May 2006 15:11:14 +0100 Subject: [ARM] 3494/1: asm-arm/bug.h needs linux/stddef.h Patch from Nicolas Pitre ... for the definition of NULL. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King --- include/asm-arm/bug.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/asm-arm/bug.h b/include/asm-arm/bug.h index 7fb02138f585..5ab8216f5204 100644 --- a/include/asm-arm/bug.h +++ b/include/asm-arm/bug.h @@ -2,6 +2,7 @@ #define _ASMARM_BUG_H #include +#include #ifdef CONFIG_BUG #ifdef CONFIG_DEBUG_BUGVERBOSE -- cgit v1.2.3 From 913ed41eb5c948d2f8b5deffd29c2638eceef3d7 Mon Sep 17 00:00:00 2001 From: Jon Mason Date: Wed, 3 May 2006 17:26:58 -0500 Subject: [IA64] remove asm-ia64/bitops.h self-inclusion asm-ia64/bitops.h includes itself. The #ifndef _ASM_IA64_BITOPS_H prevents this from being an issue, but it should still be removed. Signed-off-by: Jon Mason Signed-off-by: Tony Luck --- include/asm-ia64/bitops.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/asm-ia64/bitops.h b/include/asm-ia64/bitops.h index 90921e162793..6cc517e212a9 100644 --- a/include/asm-ia64/bitops.h +++ b/include/asm-ia64/bitops.h @@ -11,7 +11,6 @@ #include #include -#include #include /** -- cgit v1.2.3 From d57336e3f2dd7c2d1fbe4a8323029869fb6e1f00 Mon Sep 17 00:00:00 2001 From: Daniel Drake Date: Sun, 30 Apr 2006 22:09:07 +0100 Subject: [PATCH] softmac: make non-operational after being stopped zd1211 with softmac and wpa_supplicant revealed an issue with softmac and the use of workqueues. Some of the work functions actually reschedule themselves, so this meant that there could still be pending work after flush_scheduled_work() had been called during ieee80211softmac_stop(). This patch introduces a "running" flag which is used to ensure that rescheduling does not happen in this situation. I also used this flag to ensure that softmac's hooks into ieee80211 are non-operational once the stop operation has been started. This simply makes softmac a little more robust, because I could crash it easily by receiving frames in the short timeframe after shutting down softmac and before turning off the ZD1211 radio. (ZD1211 is now fixed as well!) Signed-off-by: Daniel Drake Acked-by: Johannes Berg Signed-off-by: John W. Linville --- include/net/ieee80211softmac.h | 3 ++- net/ieee80211/softmac/ieee80211softmac_assoc.c | 17 +++++++++++++++-- net/ieee80211/softmac/ieee80211softmac_auth.c | 11 +++++++++++ net/ieee80211/softmac/ieee80211softmac_module.c | 4 ++++ net/ieee80211/softmac/ieee80211softmac_scan.c | 8 ++++++++ 5 files changed, 40 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/net/ieee80211softmac.h b/include/net/ieee80211softmac.h index b1ebfbae397f..052ed596a4e4 100644 --- a/include/net/ieee80211softmac.h +++ b/include/net/ieee80211softmac.h @@ -204,7 +204,8 @@ struct ieee80211softmac_device { /* couple of flags */ u8 scanning:1, /* protects scanning from being done multiple times at once */ - associated:1; + associated:1, + running:1; struct ieee80211softmac_scaninfo *scaninfo; struct ieee80211softmac_assoc_info associnfo; diff --git a/net/ieee80211/softmac/ieee80211softmac_assoc.c b/net/ieee80211/softmac/ieee80211softmac_assoc.c index fb79ce7d6439..57ea9f6f465c 100644 --- a/net/ieee80211/softmac/ieee80211softmac_assoc.c +++ b/net/ieee80211/softmac/ieee80211softmac_assoc.c @@ -51,11 +51,12 @@ ieee80211softmac_assoc(struct ieee80211softmac_device *mac, struct ieee80211soft spin_lock_irqsave(&mac->lock, flags); mac->associnfo.associating = 1; mac->associated = 0; /* just to make sure */ - spin_unlock_irqrestore(&mac->lock, flags); /* Set a timer for timeout */ /* FIXME: make timeout configurable */ - schedule_delayed_work(&mac->associnfo.timeout, 5 * HZ); + if (likely(mac->running)) + schedule_delayed_work(&mac->associnfo.timeout, 5 * HZ); + spin_unlock_irqrestore(&mac->lock, flags); } void @@ -319,6 +320,9 @@ ieee80211softmac_handle_assoc_response(struct net_device * dev, u16 status = le16_to_cpup(&resp->status); struct ieee80211softmac_network *network = NULL; unsigned long flags; + + if (unlikely(!mac->running)) + return -ENODEV; spin_lock_irqsave(&mac->lock, flags); @@ -377,10 +381,16 @@ ieee80211softmac_handle_disassoc(struct net_device * dev, { struct ieee80211softmac_device *mac = ieee80211_priv(dev); unsigned long flags; + + if (unlikely(!mac->running)) + return -ENODEV; + if (memcmp(disassoc->header.addr2, mac->associnfo.bssid, ETH_ALEN)) return 0; + if (memcmp(disassoc->header.addr1, mac->dev->dev_addr, ETH_ALEN)) return 0; + dprintk(KERN_INFO PFX "got disassoc frame\n"); netif_carrier_off(dev); spin_lock_irqsave(&mac->lock, flags); @@ -400,6 +410,9 @@ ieee80211softmac_handle_reassoc_req(struct net_device * dev, struct ieee80211softmac_device *mac = ieee80211_priv(dev); struct ieee80211softmac_network *network; + if (unlikely(!mac->running)) + return -ENODEV; + network = ieee80211softmac_get_network_by_bssid(mac, resp->header.addr3); if (!network) { dprintkl(KERN_INFO PFX "reassoc request from unknown network\n"); diff --git a/net/ieee80211/softmac/ieee80211softmac_auth.c b/net/ieee80211/softmac/ieee80211softmac_auth.c index d6a04f3ab86c..06e332624665 100644 --- a/net/ieee80211/softmac/ieee80211softmac_auth.c +++ b/net/ieee80211/softmac/ieee80211softmac_auth.c @@ -86,6 +86,11 @@ ieee80211softmac_auth_queue(void *data) /* Lock and set flags */ spin_lock_irqsave(&mac->lock, flags); + if (unlikely(!mac->running)) { + /* Prevent reschedule on workqueue flush */ + spin_unlock_irqrestore(&mac->lock, flags); + return; + } net->authenticated = 0; net->authenticating = 1; /* add a timeout call so we eventually give up waiting for an auth reply */ @@ -124,6 +129,9 @@ ieee80211softmac_auth_resp(struct net_device *dev, struct ieee80211_auth *auth) unsigned long flags; u8 * data; + if (unlikely(!mac->running)) + return -ENODEV; + /* Find correct auth queue item */ spin_lock_irqsave(&mac->lock, flags); list_for_each(list_ptr, &mac->auth_queue) { @@ -336,6 +344,9 @@ ieee80211softmac_deauth_resp(struct net_device *dev, struct ieee80211_deauth *de struct ieee80211softmac_network *net = NULL; struct ieee80211softmac_device *mac = ieee80211_priv(dev); + if (unlikely(!mac->running)) + return -ENODEV; + if (!deauth) { dprintk("deauth without deauth packet. eek!\n"); return 0; diff --git a/net/ieee80211/softmac/ieee80211softmac_module.c b/net/ieee80211/softmac/ieee80211softmac_module.c index be83bdc1644a..6252be2c0db9 100644 --- a/net/ieee80211/softmac/ieee80211softmac_module.c +++ b/net/ieee80211/softmac/ieee80211softmac_module.c @@ -89,6 +89,8 @@ ieee80211softmac_clear_pending_work(struct ieee80211softmac_device *sm) ieee80211softmac_wait_for_scan(sm); spin_lock_irqsave(&sm->lock, flags); + sm->running = 0; + /* Free all pending assoc work items */ cancel_delayed_work(&sm->associnfo.work); @@ -204,6 +206,8 @@ void ieee80211softmac_start(struct net_device *dev) assert(0); if (mac->txrates_change) mac->txrates_change(dev, change, &oldrates); + + mac->running = 1; } EXPORT_SYMBOL_GPL(ieee80211softmac_start); diff --git a/net/ieee80211/softmac/ieee80211softmac_scan.c b/net/ieee80211/softmac/ieee80211softmac_scan.c index 2b9e7edfa3ce..d31cf77498c4 100644 --- a/net/ieee80211/softmac/ieee80211softmac_scan.c +++ b/net/ieee80211/softmac/ieee80211softmac_scan.c @@ -115,7 +115,15 @@ void ieee80211softmac_scan(void *d) // TODO: is this if correct, or should we do this only if scanning from assoc request? if (sm->associnfo.req_essid.len) ieee80211softmac_send_mgt_frame(sm, &sm->associnfo.req_essid, IEEE80211_STYPE_PROBE_REQ, 0); + + spin_lock_irqsave(&sm->lock, flags); + if (unlikely(!sm->running)) { + /* Prevent reschedule on workqueue flush */ + spin_unlock_irqrestore(&sm->lock, flags); + break; + } schedule_delayed_work(&si->softmac_scan, IEEE80211SOFTMAC_PROBE_DELAY); + spin_unlock_irqrestore(&sm->lock, flags); return; } else { dprintk(PFX "Not probing Channel %d (not allowed here)\n", si->channels[current_channel_idx].channel); -- cgit v1.2.3 From f21709d70ad6d7ad50288f7056c3a368138b017c Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Thu, 4 May 2006 19:47:19 +0200 Subject: [PATCH] ieee80211: Fix A band channel count (resent) The channel count for 802.11a is still not right. We better compute it from the min and max channel numbers, rather than hardcoding it. Signed-off-by: Jean Delvare Signed-off-by: John W. Linville --- include/net/ieee80211.h | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/net/ieee80211.h b/include/net/ieee80211.h index 4725ff861c57..d5926bfb1fc9 100644 --- a/include/net/ieee80211.h +++ b/include/net/ieee80211.h @@ -955,11 +955,13 @@ enum ieee80211_state { #define IEEE80211_24GHZ_MIN_CHANNEL 1 #define IEEE80211_24GHZ_MAX_CHANNEL 14 -#define IEEE80211_24GHZ_CHANNELS 14 +#define IEEE80211_24GHZ_CHANNELS (IEEE80211_24GHZ_MAX_CHANNEL - \ + IEEE80211_24GHZ_MIN_CHANNEL + 1) #define IEEE80211_52GHZ_MIN_CHANNEL 34 #define IEEE80211_52GHZ_MAX_CHANNEL 165 -#define IEEE80211_52GHZ_CHANNELS 131 +#define IEEE80211_52GHZ_CHANNELS (IEEE80211_52GHZ_MAX_CHANNEL - \ + IEEE80211_52GHZ_MIN_CHANNEL + 1) enum { IEEE80211_CH_PASSIVE_ONLY = (1 << 0), -- cgit v1.2.3 From 99532559dc7a8e686b2cef14c780a7ad5dbd4a31 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Fri, 5 May 2006 22:32:24 +0100 Subject: [ARM] 3500/1: fix PXA27x DMA allocation priority Patch from Nicolas Pitre Intel PXA27x developers manual section 5.4.1.1 lists a priority distribution for the DMA channels differently than what the code currently assumes. This patch fixes that. Noticed by Simon Vogl Signed-off-by: Nicolas Pitre Signed-off-by: Russell King --- arch/arm/mach-pxa/dma.c | 17 +++++------------ include/asm-arm/arch-pxa/dma.h | 26 ++++++++++++++------------ 2 files changed, 19 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/arch/arm/mach-pxa/dma.c b/arch/arm/mach-pxa/dma.c index 458112b21e25..7d8c85486c66 100644 --- a/arch/arm/mach-pxa/dma.c +++ b/arch/arm/mach-pxa/dma.c @@ -45,23 +45,16 @@ int pxa_request_dma (char *name, pxa_dma_prio prio, local_irq_save(flags); - /* try grabbing a DMA channel with the requested priority */ - for (i = prio; i < prio + PXA_DMA_NBCH(prio); i++) { - if (!dma_channels[i].name) { - found = 1; - break; - } - } - - if (!found) { - /* requested prio group is full, try hier priorities */ - for (i = prio-1; i >= 0; i--) { + do { + /* try grabbing a DMA channel with the requested priority */ + pxa_for_each_dma_prio (i, prio) { if (!dma_channels[i].name) { found = 1; break; } } - } + /* if requested prio group is full, try a hier priority */ + } while (!found && prio--); if (found) { DCSR(i) = DCSR_STARTINTR|DCSR_ENDINTR|DCSR_BUSERR; diff --git a/include/asm-arm/arch-pxa/dma.h b/include/asm-arm/arch-pxa/dma.h index 3e88a2a02a0f..a008150abc59 100644 --- a/include/asm-arm/arch-pxa/dma.h +++ b/include/asm-arm/arch-pxa/dma.h @@ -24,27 +24,29 @@ typedef struct pxa_dma_desc { volatile u32 dcmd; /* DCMD value for the current transfer */ } pxa_dma_desc; +typedef enum { + DMA_PRIO_HIGH = 0, + DMA_PRIO_MEDIUM = 1, + DMA_PRIO_LOW = 2 +} pxa_dma_prio; + #if defined(CONFIG_PXA27x) #define PXA_DMA_CHANNELS 32 -#define PXA_DMA_NBCH(prio) ((prio == DMA_PRIO_LOW) ? 16 : 8) -typedef enum { - DMA_PRIO_HIGH = 0, - DMA_PRIO_MEDIUM = 8, - DMA_PRIO_LOW = 16 -} pxa_dma_prio; +#define pxa_for_each_dma_prio(ch, prio) \ +for ( \ + ch = prio * 4; \ + ch != (4 << prio) + 16; \ + ch = (ch + 1 == (4 << prio)) ? (prio * 4 + 16) : (ch + 1) \ +) #elif defined(CONFIG_PXA25x) #define PXA_DMA_CHANNELS 16 -#define PXA_DMA_NBCH(prio) ((prio == DMA_PRIO_LOW) ? 8 : 4) -typedef enum { - DMA_PRIO_HIGH = 0, - DMA_PRIO_MEDIUM = 4, - DMA_PRIO_LOW = 8 -} pxa_dma_prio; +#define pxa_for_each_dma_prio(ch, prio) \ + for (ch = prio * 4; ch != (4 << prio); ch++) #endif -- cgit v1.2.3 From 568cb09b9d889b6f2852ede19772b8e9eed36c1e Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Fri, 5 May 2006 22:35:05 +0100 Subject: [ARM] 3495/1: EABI: undefine removed syscalls, but... Patch from Nicolas Pitre ... but only for user space. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King --- include/asm-arm/unistd.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/asm-arm/unistd.h b/include/asm-arm/unistd.h index 26f2f4828e03..cbf39a56dbe7 100644 --- a/include/asm-arm/unistd.h +++ b/include/asm-arm/unistd.h @@ -363,7 +363,7 @@ /* * The following syscalls are obsolete and no longer available for EABI. */ -#if defined(__ARM_EABI__) +#if defined(__ARM_EABI__) && !defined(__KERNEL__) #undef __NR_time #undef __NR_umount #undef __NR_stime -- cgit v1.2.3 From 7c3ceb4fb9667f34f1599a062efecf4cdc4a4ce5 Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Fri, 5 May 2006 17:02:09 -0700 Subject: [SCTP]: Allow spillover of receive buffer to avoid deadlock. This patch fixes a deadlock situation in the receive path by allowing temporary spillover of the receive buffer. - If the chunk we receive has a tsn that immediately follows the ctsn, accept it even if we run out of receive buffer space and renege data with higher TSNs. - Once we accept one chunk in a packet, accept all the remaining chunks even if we run out of receive buffer space. Signed-off-by: Neil Horman Acked-by: Mark Butler Acked-by: Vlad Yasevich Signed-off-by: Sridhar Samudrala Signed-off-by: David S. Miller --- include/net/sctp/structs.h | 1 + net/sctp/inqueue.c | 1 + net/sctp/sm_statefuns.c | 46 ++++++++++++++++++++++++++++++++++++---------- 3 files changed, 38 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index eba99f375517..7f4fea173fb1 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -712,6 +712,7 @@ struct sctp_chunk { __u8 tsn_gap_acked; /* Is this chunk acked by a GAP ACK? */ __s8 fast_retransmit; /* Is this chunk fast retransmitted? */ __u8 tsn_missing_report; /* Data chunk missing counter. */ + __u8 data_accepted; /* At least 1 chunk in this packet accepted */ }; void sctp_chunk_hold(struct sctp_chunk *); diff --git a/net/sctp/inqueue.c b/net/sctp/inqueue.c index 297b8951463e..cf0c767d43ae 100644 --- a/net/sctp/inqueue.c +++ b/net/sctp/inqueue.c @@ -149,6 +149,7 @@ struct sctp_chunk *sctp_inq_pop(struct sctp_inq *queue) /* This is the first chunk in the packet. */ chunk->singleton = 1; ch = (sctp_chunkhdr_t *) chunk->skb->data; + chunk->data_accepted = 0; } chunk->chunk_hdr = ch; diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index 2b9a832b29a7..f5d131f52a70 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -5151,7 +5151,9 @@ static int sctp_eat_data(const struct sctp_association *asoc, int tmp; __u32 tsn; int account_value; + struct sctp_tsnmap *map = (struct sctp_tsnmap *)&asoc->peer.tsn_map; struct sock *sk = asoc->base.sk; + int rcvbuf_over = 0; data_hdr = chunk->subh.data_hdr = (sctp_datahdr_t *)chunk->skb->data; skb_pull(chunk->skb, sizeof(sctp_datahdr_t)); @@ -5162,10 +5164,16 @@ static int sctp_eat_data(const struct sctp_association *asoc, /* ASSERT: Now skb->data is really the user data. */ /* - * if we are established, and we have used up our receive - * buffer memory, drop the frame - */ - if (asoc->state == SCTP_STATE_ESTABLISHED) { + * If we are established, and we have used up our receive buffer + * memory, think about droping the frame. + * Note that we have an opportunity to improve performance here. + * If we accept one chunk from an skbuff, we have to keep all the + * memory of that skbuff around until the chunk is read into user + * space. Therefore, once we accept 1 chunk we may as well accept all + * remaining chunks in the skbuff. The data_accepted flag helps us do + * that. + */ + if ((asoc->state == SCTP_STATE_ESTABLISHED) && (!chunk->data_accepted)) { /* * If the receive buffer policy is 1, then each * association can allocate up to sk_rcvbuf bytes @@ -5176,9 +5184,25 @@ static int sctp_eat_data(const struct sctp_association *asoc, account_value = atomic_read(&asoc->rmem_alloc); else account_value = atomic_read(&sk->sk_rmem_alloc); - - if (account_value > sk->sk_rcvbuf) - return SCTP_IERROR_IGNORE_TSN; + if (account_value > sk->sk_rcvbuf) { + /* + * We need to make forward progress, even when we are + * under memory pressure, so we always allow the + * next tsn after the ctsn ack point to be accepted. + * This lets us avoid deadlocks in which we have to + * drop frames that would otherwise let us drain the + * receive queue. + */ + if ((sctp_tsnmap_get_ctsn(map) + 1) != tsn) + return SCTP_IERROR_IGNORE_TSN; + + /* + * We're going to accept the frame but we should renege + * to make space for it. This will send us down that + * path later in this function. + */ + rcvbuf_over = 1; + } } /* Process ECN based congestion. @@ -5226,6 +5250,7 @@ static int sctp_eat_data(const struct sctp_association *asoc, datalen -= sizeof(sctp_data_chunk_t); deliver = SCTP_CMD_CHUNK_ULP; + chunk->data_accepted = 1; /* Think about partial delivery. */ if ((datalen >= asoc->rwnd) && (!asoc->ulpq.pd_mode)) { @@ -5242,7 +5267,8 @@ static int sctp_eat_data(const struct sctp_association *asoc, * large spill over. */ if (!asoc->rwnd || asoc->rwnd_over || - (datalen > asoc->rwnd + asoc->frag_point)) { + (datalen > asoc->rwnd + asoc->frag_point) || + rcvbuf_over) { /* If this is the next TSN, consider reneging to make * room. Note: Playing nice with a confused sender. A @@ -5250,8 +5276,8 @@ static int sctp_eat_data(const struct sctp_association *asoc, * space and in the future we may want to detect and * do more drastic reneging. */ - if (sctp_tsnmap_has_gap(&asoc->peer.tsn_map) && - (sctp_tsnmap_get_ctsn(&asoc->peer.tsn_map) + 1) == tsn) { + if (sctp_tsnmap_has_gap(map) && + (sctp_tsnmap_get_ctsn(map) + 1) == tsn) { SCTP_DEBUG_PRINTK("Reneging for tsn:%u\n", tsn); deliver = SCTP_CMD_RENEGE; } else { -- cgit v1.2.3 From 1498221d51a43d5fa1a580618591497d90f957d9 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Sat, 6 May 2006 17:55:11 -0700 Subject: [CLASS DEVICE]: add attribute_group creation Extend the support of attribute groups in class_device's to allow groups to be created as part of the registration process. This allows network device's to avoid race between registration and creating groups. Note that unlike attributes that are a property of the class object, the groups are a property of the class_device object. This is done because there are different types of network devices (wireless for example). Signed-off-by: Stephen Hemminger Signed-off-by: Greg Kroah-Hartman Signed-off-by: David S. Miller --- drivers/base/class.c | 32 ++++++++++++++++++++++++++++++++ include/linux/device.h | 2 ++ 2 files changed, 34 insertions(+) (limited to 'include') diff --git a/drivers/base/class.c b/drivers/base/class.c index 0e71dff327cd..b1ea4df85c7d 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -456,6 +456,35 @@ static void class_device_remove_attrs(struct class_device * cd) } } +static int class_device_add_groups(struct class_device * cd) +{ + int i; + int error = 0; + + if (cd->groups) { + for (i = 0; cd->groups[i]; i++) { + error = sysfs_create_group(&cd->kobj, cd->groups[i]); + if (error) { + while (--i >= 0) + sysfs_remove_group(&cd->kobj, cd->groups[i]); + goto out; + } + } + } +out: + return error; +} + +static void class_device_remove_groups(struct class_device * cd) +{ + int i; + if (cd->groups) { + for (i = 0; cd->groups[i]; i++) { + sysfs_remove_group(&cd->kobj, cd->groups[i]); + } + } +} + static ssize_t show_dev(struct class_device *class_dev, char *buf) { return print_dev_t(buf, class_dev->devt); @@ -559,6 +588,8 @@ int class_device_add(struct class_device *class_dev) class_name); } + class_device_add_groups(class_dev); + kobject_uevent(&class_dev->kobj, KOBJ_ADD); /* notify any interfaces this device is now here */ @@ -672,6 +703,7 @@ void class_device_del(struct class_device *class_dev) if (class_dev->devt_attr) class_device_remove_file(class_dev, class_dev->devt_attr); class_device_remove_attrs(class_dev); + class_device_remove_groups(class_dev); kobject_uevent(&class_dev->kobj, KOBJ_REMOVE); kobject_del(&class_dev->kobj); diff --git a/include/linux/device.h b/include/linux/device.h index f6e72a65a3f2..e8e53b9accc6 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -200,6 +200,7 @@ extern int class_device_create_file(struct class_device *, * @node: for internal use by the driver core only. * @kobj: for internal use by the driver core only. * @devt_attr: for internal use by the driver core only. + * @groups: optional additional groups to be created * @dev: if set, a symlink to the struct device is created in the sysfs * directory for this struct class device. * @class_data: pointer to whatever you want to store here for this struct @@ -228,6 +229,7 @@ struct class_device { struct device * dev; /* not necessary, but nice to have */ void * class_data; /* class-specific data */ struct class_device *parent; /* parent of this child device, if there is one */ + struct attribute_group ** groups; /* optional groups */ void (*release)(struct class_device *dev); int (*uevent)(struct class_device *dev, char **envp, -- cgit v1.2.3 From fe9925b551a95fae6ec61470c79f8b701a2fe928 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Sat, 6 May 2006 17:56:03 -0700 Subject: [NET]: Create netdev attribute_groups with class_device_add Atomically create attributes when class device is added. This avoids the race between registering class_device (which generates hotplug event), and the creation of attribute groups. Signed-off-by: Stephen Hemminger Signed-off-by: Greg Kroah-Hartman Signed-off-by: David S. Miller --- include/linux/netdevice.h | 2 ++ net/core/dev.c | 2 +- net/core/net-sysfs.c | 49 ++++++++++++----------------------------------- 3 files changed, 15 insertions(+), 38 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 01db7b88a2b1..309f9190a922 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -506,6 +506,8 @@ struct net_device /* class/net/name entry */ struct class_device class_dev; + /* space for optional statistics and wireless sysfs groups */ + struct attribute_group *sysfs_groups[3]; }; #define NETDEV_ALIGN 32 diff --git a/net/core/dev.c b/net/core/dev.c index 3bad1afc89fa..9ab3cfa58466 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3043,11 +3043,11 @@ void netdev_run_todo(void) switch(dev->reg_state) { case NETREG_REGISTERING: - dev->reg_state = NETREG_REGISTERED; err = netdev_register_sysfs(dev); if (err) printk(KERN_ERR "%s: failed sysfs registration (%d)\n", dev->name, err); + dev->reg_state = NETREG_REGISTERED; break; case NETREG_UNREGISTERING: diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index c12990c9c603..47a6fceb6771 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -29,7 +29,7 @@ static const char fmt_ulong[] = "%lu\n"; static inline int dev_isalive(const struct net_device *dev) { - return dev->reg_state == NETREG_REGISTERED; + return dev->reg_state <= NETREG_REGISTERED; } /* use same locking rules as GIF* ioctl's */ @@ -445,58 +445,33 @@ static struct class net_class = { void netdev_unregister_sysfs(struct net_device * net) { - struct class_device * class_dev = &(net->class_dev); - - if (net->get_stats) - sysfs_remove_group(&class_dev->kobj, &netstat_group); - -#ifdef WIRELESS_EXT - if (net->get_wireless_stats || (net->wireless_handlers && - net->wireless_handlers->get_wireless_stats)) - sysfs_remove_group(&class_dev->kobj, &wireless_group); -#endif - class_device_del(class_dev); - + class_device_del(&(net->class_dev)); } /* Create sysfs entries for network device. */ int netdev_register_sysfs(struct net_device *net) { struct class_device *class_dev = &(net->class_dev); - int ret; + struct attribute_group **groups = net->sysfs_groups; + class_device_initialize(class_dev); class_dev->class = &net_class; class_dev->class_data = net; + class_dev->groups = groups; + BUILD_BUG_ON(BUS_ID_SIZE < IFNAMSIZ); strlcpy(class_dev->class_id, net->name, BUS_ID_SIZE); - if ((ret = class_device_register(class_dev))) - goto out; - if (net->get_stats && - (ret = sysfs_create_group(&class_dev->kobj, &netstat_group))) - goto out_unreg; + if (net->get_stats) + *groups++ = &netstat_group; #ifdef WIRELESS_EXT - if (net->get_wireless_stats || (net->wireless_handlers && - net->wireless_handlers->get_wireless_stats)) { - ret = sysfs_create_group(&class_dev->kobj, &wireless_group); - if (ret) - goto out_cleanup; - } - return 0; -out_cleanup: - if (net->get_stats) - sysfs_remove_group(&class_dev->kobj, &netstat_group); -#else - return 0; + if (net->get_wireless_stats + || (net->wireless_handlers && net->wireless_handlers->get_wireless_stats)) + *groups++ = &wireless_group; #endif -out_unreg: - printk(KERN_WARNING "%s: sysfs attribute registration failed %d\n", - net->name, ret); - class_device_unregister(class_dev); -out: - return ret; + return class_device_add(class_dev); } int netdev_sysfs_init(void) -- cgit v1.2.3 From 216251cff98838f2b79c53fc8a9e76884944be7d Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Sun, 7 May 2006 18:56:27 +0100 Subject: [ARM] 3501/1: i.MX: fix lowlevel debug macros Patch from Sascha Hauer This patch fixes the addruart macro to work with both mmu enabled and disabled. Signed-off-by: Sascha Hauer Signed-off-by: Russell King --- arch/arm/mach-imx/mx1ads.c | 2 +- include/asm-arm/arch-imx/debug-macro.S | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/arm/mach-imx/mx1ads.c b/arch/arm/mach-imx/mx1ads.c index e1f6c0bbe1e7..da893c80d471 100644 --- a/arch/arm/mach-imx/mx1ads.c +++ b/arch/arm/mach-imx/mx1ads.c @@ -161,7 +161,7 @@ mx1ads_map_io(void) MACHINE_START(MX1ADS, "Motorola MX1ADS") /* Maintainer: Sascha Hauer, Pengutronix */ .phys_io = 0x00200000, - .io_pg_offst = ((0xe0200000) >> 18) & 0xfffc, + .io_pg_offst = ((0xe0000000) >> 18) & 0xfffc, .boot_params = 0x08000100, .map_io = mx1ads_map_io, .init_irq = imx_init_irq, diff --git a/include/asm-arm/arch-imx/debug-macro.S b/include/asm-arm/arch-imx/debug-macro.S index 83f552f7bcc1..c611871643a2 100644 --- a/include/asm-arm/arch-imx/debug-macro.S +++ b/include/asm-arm/arch-imx/debug-macro.S @@ -16,7 +16,7 @@ tst \rx, #1 @ MMU enabled? moveq \rx, #0x00000000 @ physical movne \rx, #0xe0000000 @ virtual - orr \rx, \rx, #0x00200000 + orreq \rx, \rx, #0x00200000 @ physical orr \rx, \rx, #0x00006000 @ UART1 offset .endm -- cgit v1.2.3 From 201be92a4243e58bcc6c0878489bcc2aaaf51c80 Mon Sep 17 00:00:00 2001 From: Bellido Nicolas Date: Sun, 7 May 2006 22:49:22 +0100 Subject: [ARM] 3505/1: aaec2000: entry-macro.S needs asm/arch/irqs.h Patch from Bellido Nicolas Since git commit 2b78838842346da390e8547cd37035184376d506, entry-macro.S needs to include asm/arch/irqs.h Signed-off-by: Nicolas Bellido Signed-off-by: Russell King --- include/asm-arm/arch-aaec2000/entry-macro.S | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/asm-arm/arch-aaec2000/entry-macro.S b/include/asm-arm/arch-aaec2000/entry-macro.S index df31313ab07e..1eb3503bd16e 100644 --- a/include/asm-arm/arch-aaec2000/entry-macro.S +++ b/include/asm-arm/arch-aaec2000/entry-macro.S @@ -10,6 +10,7 @@ * published by the Free Software Foundation. * */ +#include .macro disable_fiq .endm -- cgit v1.2.3 From 9a708becafe99fa32211e8c53dbacefdb4b11718 Mon Sep 17 00:00:00 2001 From: Bellido Nicolas Date: Sun, 7 May 2006 22:49:23 +0100 Subject: [ARM] 3506/1: aaec2000: debug-macro.S needs hardware.h Patch from Bellido Nicolas Include hardware.h in debug-macro.S, otherwise io_p2v is undefined. Signed-off-by: Nicolas Bellido Signed-off-by: Russell King --- include/asm-arm/arch-aaec2000/debug-macro.S | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/asm-arm/arch-aaec2000/debug-macro.S b/include/asm-arm/arch-aaec2000/debug-macro.S index e4f1fa539a74..7b1fce021d8a 100644 --- a/include/asm-arm/arch-aaec2000/debug-macro.S +++ b/include/asm-arm/arch-aaec2000/debug-macro.S @@ -9,6 +9,7 @@ * published by the Free Software Foundation. */ +#include "hardware.h" .macro addruart,rx mrc p15, 0, \rx, c1, c0 tst \rx, #1 @ MMU enabled? -- cgit v1.2.3 From e0c1e9bf81badc7ba59e120d6218101903d5d103 Mon Sep 17 00:00:00 2001 From: Kimball Murray Date: Mon, 8 May 2006 15:17:16 +0200 Subject: [PATCH] x86_64: avoid IRQ0 ioapic pin collision The patch addresses a problem with ACPI SCI interrupt entry, which gets re-used, and the IRQ is assigned to another unrelated device. The patch corrects the code such that SCI IRQ is skipped and duplicate entry is avoided. Second issue came up with VIA chipset, the problem was caused by original patch assigning IRQs starting 16 and up. The VIA chipset uses 4-bit IRQ register for internal interrupt routing, and therefore cannot handle IRQ numbers assigned to its devices. The patch corrects this problem by allowing PCI IRQs below 16. Cc: len.brown@intel.com Signed-off by: Natalie Protasevich Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/io_apic.c | 5 +++++ arch/i386/kernel/mpparse.c | 12 +++++++++++- arch/x86_64/kernel/io_apic.c | 5 +++++ arch/x86_64/kernel/mpparse.c | 12 +++++++++++- include/asm-i386/io_apic.h | 1 + include/asm-x86_64/io_apic.h | 1 + 6 files changed, 34 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/arch/i386/kernel/io_apic.c b/arch/i386/kernel/io_apic.c index f8f132aa5472..d70f2ade5cde 100644 --- a/arch/i386/kernel/io_apic.c +++ b/arch/i386/kernel/io_apic.c @@ -2238,6 +2238,8 @@ static inline void unlock_ExtINT_logic(void) spin_unlock_irqrestore(&ioapic_lock, flags); } +int timer_uses_ioapic_pin_0; + /* * This code may look a bit paranoid, but it's supposed to cooperate with * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ @@ -2274,6 +2276,9 @@ static inline void check_timer(void) pin2 = ioapic_i8259.pin; apic2 = ioapic_i8259.apic; + if (pin1 == 0) + timer_uses_ioapic_pin_0 = 1; + printk(KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", vector, apic1, pin1, apic2, pin2); diff --git a/arch/i386/kernel/mpparse.c b/arch/i386/kernel/mpparse.c index 34d21e21e012..6b1392d33ed5 100644 --- a/arch/i386/kernel/mpparse.c +++ b/arch/i386/kernel/mpparse.c @@ -1130,7 +1130,17 @@ int mp_register_gsi (u32 gsi, int triggering, int polarity) */ int irq = gsi; if (gsi < MAX_GSI_NUM) { - if (gsi > 15) + /* + * Retain the VIA chipset work-around (gsi > 15), but + * avoid a problem where the 8254 timer (IRQ0) is setup + * via an override (so it's not on pin 0 of the ioapic), + * and at the same time, the pin 0 interrupt is a PCI + * type. The gsi > 15 test could cause these two pins + * to be shared as IRQ0, and they are not shareable. + * So test for this condition, and if necessary, avoid + * the pin collision. + */ + if (gsi > 15 || (gsi == 0 && !timer_uses_ioapic_pin_0)) gsi = pci_irq++; /* * Don't assign IRQ used by ACPI SCI diff --git a/arch/x86_64/kernel/io_apic.c b/arch/x86_64/kernel/io_apic.c index 77b4c608cca0..0de3ea938830 100644 --- a/arch/x86_64/kernel/io_apic.c +++ b/arch/x86_64/kernel/io_apic.c @@ -1777,6 +1777,8 @@ static inline void unlock_ExtINT_logic(void) spin_unlock_irqrestore(&ioapic_lock, flags); } +int timer_uses_ioapic_pin_0; + /* * This code may look a bit paranoid, but it's supposed to cooperate with * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ @@ -1814,6 +1816,9 @@ static inline void check_timer(void) pin2 = ioapic_i8259.pin; apic2 = ioapic_i8259.apic; + if (pin1 == 0) + timer_uses_ioapic_pin_0 = 1; + apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", vector, apic1, pin1, apic2, pin2); diff --git a/arch/x86_64/kernel/mpparse.c b/arch/x86_64/kernel/mpparse.c index b17cf3eba359..083da7e606b1 100644 --- a/arch/x86_64/kernel/mpparse.c +++ b/arch/x86_64/kernel/mpparse.c @@ -968,7 +968,17 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) */ int irq = gsi; if (gsi < MAX_GSI_NUM) { - if (gsi > 15) + /* + * Retain the VIA chipset work-around (gsi > 15), but + * avoid a problem where the 8254 timer (IRQ0) is setup + * via an override (so it's not on pin 0 of the ioapic), + * and at the same time, the pin 0 interrupt is a PCI + * type. The gsi > 15 test could cause these two pins + * to be shared as IRQ0, and they are not shareable. + * So test for this condition, and if necessary, avoid + * the pin collision. + */ + if (gsi > 15 || (gsi == 0 && !timer_uses_ioapic_pin_0)) gsi = pci_irq++; /* * Don't assign IRQ used by ACPI SCI diff --git a/include/asm-i386/io_apic.h b/include/asm-i386/io_apic.h index 51c4e5fe6062..d92e253f7f6f 100644 --- a/include/asm-i386/io_apic.h +++ b/include/asm-i386/io_apic.h @@ -200,6 +200,7 @@ extern int io_apic_get_unique_id (int ioapic, int apic_id); extern int io_apic_get_version (int ioapic); extern int io_apic_get_redir_entries (int ioapic); extern int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low); +extern int timer_uses_ioapic_pin_0; #endif /* CONFIG_ACPI */ extern int (*ioapic_renumber_irq)(int ioapic, int irq); diff --git a/include/asm-x86_64/io_apic.h b/include/asm-x86_64/io_apic.h index ee1bc69aec9c..52484e82c641 100644 --- a/include/asm-x86_64/io_apic.h +++ b/include/asm-x86_64/io_apic.h @@ -205,6 +205,7 @@ extern int skip_ioapic_setup; extern int io_apic_get_version (int ioapic); extern int io_apic_get_redir_entries (int ioapic); extern int io_apic_set_pci_routing (int ioapic, int pin, int irq, int, int); +extern int timer_uses_ioapic_pin_0; #endif extern int sis_apic_bug; /* dummy */ -- cgit v1.2.3 From ac71d12c990526b01ef6cfe50907ef8530a30331 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Mon, 8 May 2006 15:17:28 +0200 Subject: [PATCH] x86_64: Avoid EBDA area in early boot allocator Based on analysis&patch from Robert Hentosch Observed on a Dell PE6850 with 16GB The problem occurs very early on, when the kernel allocates space for the temporary memory map called bootmap. The bootmap overlaps the EBDA region. EBDA region is not historically reserved in the e820 mapping. When the bootmap is freed it marks the EBDA region as usable. If you notice in setup.c there is already code to work around the EBDA in reserve_ebda_region(), this check however occurs after the bootmap is allocated and doesn't prevent the bootmap from using this range. AK: I redid the original patch. Thanks also to Jan Beulich for spotting some mistakes. Cc: Robert_Hentosch@dell.com Cc: jbeulich@novell.com Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/e820.c | 6 ++++++ arch/x86_64/kernel/setup.c | 30 ++++++++++++++++++++++-------- include/asm-x86_64/e820.h | 2 ++ 3 files changed, 30 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/arch/x86_64/kernel/e820.c b/arch/x86_64/kernel/e820.c index 62776c07cff1..222b5b46d2b2 100644 --- a/arch/x86_64/kernel/e820.c +++ b/arch/x86_64/kernel/e820.c @@ -76,6 +76,12 @@ static inline int bad_addr(unsigned long *addrp, unsigned long size) *addrp = __pa_symbol(&_end); return 1; } + + if (last >= ebda_addr && addr < ebda_addr + ebda_size) { + *addrp = ebda_addr + ebda_size; + return 1; + } + /* XXX ramdisk image here? */ return 0; } diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c index ebc3c33b1c6c..f0870bef24d1 100644 --- a/arch/x86_64/kernel/setup.c +++ b/arch/x86_64/kernel/setup.c @@ -571,17 +571,28 @@ static inline void copy_edd(void) #endif #define EBDA_ADDR_POINTER 0x40E -static void __init reserve_ebda_region(void) + +unsigned __initdata ebda_addr; +unsigned __initdata ebda_size; + +static void discover_ebda(void) { - unsigned int addr; - /** + /* * there is a real-mode segmented pointer pointing to the * 4K EBDA area at 0x40E */ - addr = *(unsigned short *)phys_to_virt(EBDA_ADDR_POINTER); - addr <<= 4; - if (addr) - reserve_bootmem_generic(addr, PAGE_SIZE); + ebda_addr = *(unsigned short *)EBDA_ADDR_POINTER; + ebda_addr <<= 4; + + ebda_size = *(unsigned short *)(unsigned long)ebda_addr; + + /* Round EBDA up to pages */ + if (ebda_size == 0) + ebda_size = 1; + ebda_size <<= 10; + ebda_size = round_up(ebda_size + (ebda_addr & ~PAGE_MASK), PAGE_SIZE); + if (ebda_size > 64*1024) + ebda_size = 64*1024; } void __init setup_arch(char **cmdline_p) @@ -627,6 +638,8 @@ void __init setup_arch(char **cmdline_p) check_efer(); + discover_ebda(); + init_memory_mapping(0, (end_pfn_map << PAGE_SHIFT)); dmi_scan_machine(); @@ -669,7 +682,8 @@ void __init setup_arch(char **cmdline_p) reserve_bootmem_generic(0, PAGE_SIZE); /* reserve ebda region */ - reserve_ebda_region(); + if (ebda_addr) + reserve_bootmem_generic(ebda_addr, ebda_size); #ifdef CONFIG_SMP /* diff --git a/include/asm-x86_64/e820.h b/include/asm-x86_64/e820.h index 93b51df51687..670a3388e70a 100644 --- a/include/asm-x86_64/e820.h +++ b/include/asm-x86_64/e820.h @@ -59,6 +59,8 @@ extern void __init parse_memopt(char *p, char **end); extern void __init parse_memmapopt(char *p, char **end); extern struct e820map e820; + +extern unsigned ebda_addr, ebda_size; #endif/*!__ASSEMBLY__*/ #endif/*__E820_HEADER*/ -- cgit v1.2.3 From d324031245abbb54e4e0321004430826052b6c37 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 8 May 2006 15:11:26 -0700 Subject: sky2: backout NAPI reschedule This is a backout of earlier patch. The whole rescheduling hack was a bad idea. It doesn't really solve the problem and it makes the code more complicated for no good reason. Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 20 ++++---------------- include/linux/netdevice.h | 18 ++++++++---------- 2 files changed, 12 insertions(+), 26 deletions(-) (limited to 'include') diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index 227df9876a2c..76da74fbe859 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -2105,7 +2105,6 @@ static int sky2_poll(struct net_device *dev0, int *budget) int work_done = 0; u32 status = sky2_read32(hw, B0_Y2_SP_EISR); - restart_poll: if (unlikely(status & ~Y2_IS_STAT_BMU)) { if (status & Y2_IS_HW_ERR) sky2_hw_intr(hw); @@ -2136,7 +2135,7 @@ static int sky2_poll(struct net_device *dev0, int *budget) } if (status & Y2_IS_STAT_BMU) { - work_done += sky2_status_intr(hw, work_limit - work_done); + work_done = sky2_status_intr(hw, work_limit); *budget -= work_done; dev0->quota -= work_done; @@ -2148,22 +2147,9 @@ static int sky2_poll(struct net_device *dev0, int *budget) mod_timer(&hw->idle_timer, jiffies + HZ); - local_irq_disable(); - __netif_rx_complete(dev0); + netif_rx_complete(dev0); status = sky2_read32(hw, B0_Y2_SP_LISR); - - if (unlikely(status)) { - /* More work pending, try and keep going */ - if (__netif_rx_schedule_prep(dev0)) { - __netif_rx_reschedule(dev0, work_done); - status = sky2_read32(hw, B0_Y2_SP_EISR); - local_irq_enable(); - goto restart_poll; - } - } - - local_irq_enable(); return 0; } @@ -2181,6 +2167,8 @@ static irqreturn_t sky2_intr(int irq, void *dev_id, struct pt_regs *regs) prefetch(&hw->st_le[hw->st_idx]); if (likely(__netif_rx_schedule_prep(dev0))) __netif_rx_schedule(dev0); + else + printk(KERN_DEBUG PFX "irq race detected\n"); return IRQ_HANDLED; } diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 309f9190a922..a461b51d6076 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -831,21 +831,19 @@ static inline void netif_rx_schedule(struct net_device *dev) __netif_rx_schedule(dev); } - -static inline void __netif_rx_reschedule(struct net_device *dev, int undo) -{ - dev->quota += undo; - list_add_tail(&dev->poll_list, &__get_cpu_var(softnet_data).poll_list); - __raise_softirq_irqoff(NET_RX_SOFTIRQ); -} - -/* Try to reschedule poll. Called by dev->poll() after netif_rx_complete(). */ +/* Try to reschedule poll. Called by dev->poll() after netif_rx_complete(). + * Do not inline this? + */ static inline int netif_rx_reschedule(struct net_device *dev, int undo) { if (netif_rx_schedule_prep(dev)) { unsigned long flags; + + dev->quota += undo; + local_irq_save(flags); - __netif_rx_reschedule(dev, undo); + list_add_tail(&dev->poll_list, &__get_cpu_var(softnet_data).poll_list); + __raise_softirq_irqoff(NET_RX_SOFTIRQ); local_irq_restore(flags); return 1; } -- cgit v1.2.3 From e4de00215c3af02116db3d486bf53700dfe10619 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Tue, 9 May 2006 16:00:59 +1000 Subject: powerpc/32: Define an is_kernel_addr() to fix ARCH=ppc compilation My commit 6bfd93c32a5065d0e858780b3beb0b667081601c broke the ARCH=ppc compilation by using the is_kernel_addr() macro in asm/uaccess.h. This fixes it by defining a suitable is_kernel_addr() for ARCH=ppc. Signed-off-by: Paul Mackerras --- include/asm-ppc/page.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/asm-ppc/page.h b/include/asm-ppc/page.h index a70ba2ee552d..0fb68a0b0181 100644 --- a/include/asm-ppc/page.h +++ b/include/asm-ppc/page.h @@ -20,6 +20,7 @@ /* This must match what is in arch/ppc/Makefile */ #define PAGE_OFFSET CONFIG_KERNEL_START #define KERNELBASE PAGE_OFFSET +#define is_kernel_addr(x) ((x) >= PAGE_OFFSET) #ifndef __ASSEMBLY__ -- cgit v1.2.3 From b17a7c179dd3ce7d04373fddf660eda21efc9db9 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 10 May 2006 13:21:17 -0700 Subject: [NET]: Do sysfs registration as part of register_netdevice. The last step of netdevice registration was being done by a delayed call, but because it was delayed, it was impossible to return any error code if the class_device registration failed. Side effects: * one state in registration process is unnecessary. * register_netdevice can sleep inside class_device registration/hotplug * code in netdev_run_todo only does unregistration so it is simpler. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- include/linux/netdevice.h | 3 +-- net/core/dev.c | 63 ++++++++++++++++++++--------------------------- 2 files changed, 28 insertions(+), 38 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index a461b51d6076..f4169bbb60eb 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -433,8 +433,7 @@ struct net_device /* register/unregister state machine */ enum { NETREG_UNINITIALIZED=0, - NETREG_REGISTERING, /* called register_netdevice */ - NETREG_REGISTERED, /* completed register todo */ + NETREG_REGISTERED, /* completed register_netdevice */ NETREG_UNREGISTERING, /* called unregister_netdevice */ NETREG_UNREGISTERED, /* completed unregister todo */ NETREG_RELEASED, /* called free_netdev */ diff --git a/net/core/dev.c b/net/core/dev.c index ced57430f6d8..2dce673a039b 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2777,6 +2777,8 @@ int register_netdevice(struct net_device *dev) BUG_ON(dev_boot_phase); ASSERT_RTNL(); + might_sleep(); + /* When net_device's are persistent, this will be fatal. */ BUG_ON(dev->reg_state != NETREG_UNINITIALIZED); @@ -2863,6 +2865,11 @@ int register_netdevice(struct net_device *dev) if (!dev->rebuild_header) dev->rebuild_header = default_rebuild_header; + ret = netdev_register_sysfs(dev); + if (ret) + goto out_err; + dev->reg_state = NETREG_REGISTERED; + /* * Default initial state at registry is that the * device is present. @@ -2878,14 +2885,11 @@ int register_netdevice(struct net_device *dev) hlist_add_head(&dev->name_hlist, head); hlist_add_head(&dev->index_hlist, dev_index_hash(dev->ifindex)); dev_hold(dev); - dev->reg_state = NETREG_REGISTERING; write_unlock_bh(&dev_base_lock); /* Notify protocols, that a new device appeared. */ raw_notifier_call_chain(&netdev_chain, NETDEV_REGISTER, dev); - /* Finish registration after unlock */ - net_set_todo(dev); ret = 0; out: @@ -3008,7 +3012,7 @@ static void netdev_wait_allrefs(struct net_device *dev) * * We are invoked by rtnl_unlock() after it drops the semaphore. * This allows us to deal with problems: - * 1) We can create/delete sysfs objects which invoke hotplug + * 1) We can delete sysfs objects which invoke hotplug * without deadlocking with linkwatch via keventd. * 2) Since we run with the RTNL semaphore not held, we can sleep * safely in order to wait for the netdev refcnt to drop to zero. @@ -3017,8 +3021,6 @@ static DEFINE_MUTEX(net_todo_run_mutex); void netdev_run_todo(void) { struct list_head list = LIST_HEAD_INIT(list); - int err; - /* Need to guard against multiple cpu's getting out of order. */ mutex_lock(&net_todo_run_mutex); @@ -3041,40 +3043,29 @@ void netdev_run_todo(void) = list_entry(list.next, struct net_device, todo_list); list_del(&dev->todo_list); - switch(dev->reg_state) { - case NETREG_REGISTERING: - err = netdev_register_sysfs(dev); - if (err) - printk(KERN_ERR "%s: failed sysfs registration (%d)\n", - dev->name, err); - dev->reg_state = NETREG_REGISTERED; - break; - - case NETREG_UNREGISTERING: - netdev_unregister_sysfs(dev); - dev->reg_state = NETREG_UNREGISTERED; - - netdev_wait_allrefs(dev); + if (unlikely(dev->reg_state != NETREG_UNREGISTERING)) { + printk(KERN_ERR "network todo '%s' but state %d\n", + dev->name, dev->reg_state); + dump_stack(); + continue; + } - /* paranoia */ - BUG_ON(atomic_read(&dev->refcnt)); - BUG_TRAP(!dev->ip_ptr); - BUG_TRAP(!dev->ip6_ptr); - BUG_TRAP(!dev->dn_ptr); + netdev_unregister_sysfs(dev); + dev->reg_state = NETREG_UNREGISTERED; + netdev_wait_allrefs(dev); - /* It must be the very last action, - * after this 'dev' may point to freed up memory. - */ - if (dev->destructor) - dev->destructor(dev); - break; + /* paranoia */ + BUG_ON(atomic_read(&dev->refcnt)); + BUG_TRAP(!dev->ip_ptr); + BUG_TRAP(!dev->ip6_ptr); + BUG_TRAP(!dev->dn_ptr); - default: - printk(KERN_ERR "network todo '%s' but state %d\n", - dev->name, dev->reg_state); - break; - } + /* It must be the very last action, + * after this 'dev' may point to freed up memory. + */ + if (dev->destructor) + dev->destructor(dev); } out: -- cgit v1.2.3 From 4c1b46226ce4424a93b8ac544e37afb26c8a72c6 Mon Sep 17 00:00:00 2001 From: Francois Romieu Date: Wed, 10 May 2006 12:48:57 -0700 Subject: dl2k: use DMA_48BIT_MASK constant Typo will be harder with this one. Signed-off-by: Francois Romieu Signed-off-by: Stephen Hemminger --- drivers/net/dl2k.c | 12 ++++++------ include/linux/dma-mapping.h | 1 + 2 files changed, 7 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/drivers/net/dl2k.c b/drivers/net/dl2k.c index 1f3627470c95..1ddefd281213 100644 --- a/drivers/net/dl2k.c +++ b/drivers/net/dl2k.c @@ -765,7 +765,7 @@ rio_free_tx (struct net_device *dev, int irq) break; skb = np->tx_skbuff[entry]; pci_unmap_single (np->pdev, - np->tx_ring[entry].fraginfo & 0xffffffffffff, + np->tx_ring[entry].fraginfo & DMA_48BIT_MASK, skb->len, PCI_DMA_TODEVICE); if (irq) dev_kfree_skb_irq (skb); @@ -893,7 +893,7 @@ receive_packet (struct net_device *dev) /* Small skbuffs for short packets */ if (pkt_len > copy_thresh) { pci_unmap_single (np->pdev, - desc->fraginfo & 0xffffffffffff, + desc->fraginfo & DMA_48BIT_MASK, np->rx_buf_sz, PCI_DMA_FROMDEVICE); skb_put (skb = np->rx_skbuff[entry], pkt_len); @@ -901,7 +901,7 @@ receive_packet (struct net_device *dev) } else if ((skb = dev_alloc_skb (pkt_len + 2)) != NULL) { pci_dma_sync_single_for_cpu(np->pdev, desc->fraginfo & - 0xffffffffffff, + DMA_48BIT_MASK, np->rx_buf_sz, PCI_DMA_FROMDEVICE); skb->dev = dev; @@ -913,7 +913,7 @@ receive_packet (struct net_device *dev) skb_put (skb, pkt_len); pci_dma_sync_single_for_device(np->pdev, desc->fraginfo & - 0xffffffffffff, + DMA_48BIT_MASK, np->rx_buf_sz, PCI_DMA_FROMDEVICE); } @@ -1800,7 +1800,7 @@ rio_close (struct net_device *dev) skb = np->rx_skbuff[i]; if (skb) { pci_unmap_single(np->pdev, - np->rx_ring[i].fraginfo & 0xffffffffffff, + np->rx_ring[i].fraginfo & DMA_48BIT_MASK, skb->len, PCI_DMA_FROMDEVICE); dev_kfree_skb (skb); np->rx_skbuff[i] = NULL; @@ -1810,7 +1810,7 @@ rio_close (struct net_device *dev) skb = np->tx_skbuff[i]; if (skb) { pci_unmap_single(np->pdev, - np->tx_ring[i].fraginfo & 0xffffffffffff, + np->tx_ring[i].fraginfo & DMA_48BIT_MASK, skb->len, PCI_DMA_TODEVICE); dev_kfree_skb (skb); np->tx_skbuff[i] = NULL; diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index ff61817082fa..635690cf3e3d 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -14,6 +14,7 @@ enum dma_data_direction { }; #define DMA_64BIT_MASK 0xffffffffffffffffULL +#define DMA_48BIT_MASK 0x0000ffffffffffffULL #define DMA_40BIT_MASK 0x000000ffffffffffULL #define DMA_39BIT_MASK 0x0000007fffffffffULL #define DMA_32BIT_MASK 0x00000000ffffffffULL -- cgit v1.2.3 From bd89efc532fe41f867f848144cc8b42054ddf6f9 Mon Sep 17 00:00:00 2001 From: Simon Kelley Date: Fri, 12 May 2006 14:56:08 -0700 Subject: [NEIGH]: Fix IP-over-ATM and ARP interaction. The classical IP over ATM code maintains its own IPv4 <-> ARP table, using the standard neighbour-table code. The neigh_table_init function adds this neighbour table to a linked list of all neighbor tables which is used by the functions neigh_delete() neigh_add() and neightbl_set(), all called by the netlink code. Once the ATM neighbour table is added to the list, there are two tables with family == AF_INET there, and ARP entries sent via netlink go into the first table with matching family. This is indeterminate and often wrong. To see the bug, on a kernel with CLIP enabled, create a standard IPv4 ARP entry by pinging an unused address on a local subnet. Then attempt to complete that entry by doing ip neigh replace lladdr nud reachable Looking at the ARP tables by using ip neigh show will reveal two ARP entries for the same address. One of these can be found in /proc/net/arp, and the other in /proc/net/atm/arp. This patch adds a new function, neigh_table_init_no_netlink() which does everything the neigh_table_init() does, except add the table to the netlink all-arp-tables chain. In addition neigh_table_init() has a check that all tables on the chain have a distinct address family. The init call in clip.c is changed to call neigh_table_init_no_netlink(). Since ATM ARP tables are rather more complicated than can currently be handled by the available rtattrs in the netlink protocol, no functionality is lost by this patch, and non-ATM ARP manipulation via netlink is rescued. A more complete solution would involve a rtattr for ATM ARP entries and some way for the netlink code to give neigh_add and friends more information than just address family with which to find the correct ARP table. [ I've changed the assertion checking in neigh_table_init() to not use BUG_ON() while holding neigh_tbl_lock. Instead we remember that we found an existing tbl with the same family, and after dropping the lock we'll give a diagnostic kernel log message and a stack dump. -DaveM ] Signed-off-by: Simon Kelley Signed-off-by: David S. Miller --- include/net/neighbour.h | 1 + net/atm/clip.c | 2 +- net/core/neighbour.c | 21 +++++++++++++++++++-- 3 files changed, 21 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/net/neighbour.h b/include/net/neighbour.h index b0666d66293f..4901ee446879 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -211,6 +211,7 @@ struct neigh_table #define NEIGH_UPDATE_F_ADMIN 0x80000000 extern void neigh_table_init(struct neigh_table *tbl); +extern void neigh_table_init_no_netlink(struct neigh_table *tbl); extern int neigh_table_clear(struct neigh_table *tbl); extern struct neighbour * neigh_lookup(struct neigh_table *tbl, const void *pkey, diff --git a/net/atm/clip.c b/net/atm/clip.c index 1a786bfaa416..72d852982664 100644 --- a/net/atm/clip.c +++ b/net/atm/clip.c @@ -963,7 +963,7 @@ static struct file_operations arp_seq_fops = { static int __init atm_clip_init(void) { struct proc_dir_entry *p; - neigh_table_init(&clip_tbl); + neigh_table_init_no_netlink(&clip_tbl); clip_tbl_hook = &clip_tbl; register_atm_ioctl(&clip_ioctl_ops); diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 4cf878efdb49..50a8c73caf97 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -1326,8 +1326,7 @@ void neigh_parms_destroy(struct neigh_parms *parms) kfree(parms); } - -void neigh_table_init(struct neigh_table *tbl) +void neigh_table_init_no_netlink(struct neigh_table *tbl) { unsigned long now = jiffies; unsigned long phsize; @@ -1383,10 +1382,27 @@ void neigh_table_init(struct neigh_table *tbl) tbl->last_flush = now; tbl->last_rand = now + tbl->parms.reachable_time * 20; +} + +void neigh_table_init(struct neigh_table *tbl) +{ + struct neigh_table *tmp; + + neigh_table_init_no_netlink(tbl); write_lock(&neigh_tbl_lock); + for (tmp = neigh_tables; tmp; tmp = tmp->next) { + if (tmp->family == tbl->family) + break; + } tbl->next = neigh_tables; neigh_tables = tbl; write_unlock(&neigh_tbl_lock); + + if (unlikely(tmp)) { + printk(KERN_ERR "NEIGH: Registering multiple tables for " + "family %d\n", tbl->family); + dump_stack(); + } } int neigh_table_clear(struct neigh_table *tbl) @@ -2657,6 +2673,7 @@ EXPORT_SYMBOL(neigh_rand_reach_time); EXPORT_SYMBOL(neigh_resolve_output); EXPORT_SYMBOL(neigh_table_clear); EXPORT_SYMBOL(neigh_table_init); +EXPORT_SYMBOL(neigh_table_init_no_netlink); EXPORT_SYMBOL(neigh_update); EXPORT_SYMBOL(neigh_update_hhs); EXPORT_SYMBOL(pneigh_enqueue); -- cgit v1.2.3