From ca5ecddfa8fcbd948c95530e7e817cee9fb43a3d Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 28 Apr 2010 14:39:09 -0700 Subject: rcu: define __rcu address space modifier for sparse This commit provides definitions for the __rcu annotation defined earlier. This annotation permits sparse to check for correct use of RCU-protected pointers. If a pointer that is annotated with __rcu is accessed directly (as opposed to via rcu_dereference(), rcu_assign_pointer(), or one of their variants), sparse can be made to complain. To enable such complaints, use the new default-disabled CONFIG_SPARSE_RCU_POINTER kernel configuration option. Please note that these sparse complaints are intended to be a debugging aid, -not- a code-style-enforcement mechanism. There are special rcu_dereference_protected() and rcu_access_pointer() accessors for use when RCU read-side protection is not required, for example, when no other CPU has access to the data structure in question or while the current CPU hold the update-side lock. This patch also updates a number of docbook comments that were showing their age. Signed-off-by: Arnd Bergmann Signed-off-by: Paul E. McKenney Cc: Christopher Li Reviewed-by: Josh Triplett --- lib/Kconfig.debug | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1b4afd2e6ca0..12465f2ef766 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -539,6 +539,19 @@ config PROVE_RCU_REPEATEDLY disabling, allowing multiple RCU-lockdep warnings to be printed on a single reboot. +config SPARSE_RCU_POINTER + bool "RCU debugging: sparse-based checks for pointer usage" + default n + help + This feature enables the __rcu sparse annotation for + RCU-protected pointers. This annotation will cause sparse + to flag any non-RCU used of annotated pointers. This can be + helpful when debugging RCU usage. Please note that this feature + is not intended to enforce code cleanliness; it is instead merely + a debugging aid. + + Say Y to make sparse flag questionable use of RCU-protected pointers + Say N if you are unsure. config LOCKDEP -- cgit v1.2.2 From b163760e37047781b37c412cde54d146ac4b651f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 2 Jun 2010 16:21:38 -0700 Subject: rcu: make CPU stall warning timeout configurable Also set the default to 60 seconds, up from the previous hard-coded timeout of 10 seconds. This allows people who care to set short timeouts, while avoiding people with unusual configurations (make randconfig!!!) from being bothered with spurious CPU stall warnings. Signed-off-by: Paul E. McKenney Reviewed-by: Josh Triplett --- lib/Kconfig.debug | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 12465f2ef766..b5005b762103 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -845,6 +845,17 @@ config RCU_CPU_STALL_DETECTOR Say Y if you are unsure. +config RCU_CPU_STALL_TIMEOUT + int "RCU CPU stall timeout in seconds" + depends on RCU_CPU_STALL_DETECTOR + range 3 300 + default 60 + help + If a given RCU grace period extends more than the specified + number of seconds, a CPU stall warning is printed. If the + RCU grace period persists, additional CPU stall warnings are + printed at more widely spaced intervals. + config RCU_CPU_STALL_VERBOSE bool "Print additional per-task information for RCU_CPU_STALL_DETECTOR" depends on RCU_CPU_STALL_DETECTOR && TREE_PREEMPT_RCU -- cgit v1.2.2 From a1115570b31091f3e3ab9e6cf7ee8d320a42be84 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 25 Feb 2010 23:43:52 +0100 Subject: radix-tree: __rcu annotations Signed-off-by: Arnd Bergmann Signed-off-by: Paul E. McKenney Cc: Nick Piggin Reviewed-by: Josh Triplett --- lib/radix-tree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/radix-tree.c b/lib/radix-tree.c index e907858498a6..899fb750946f 100644 --- a/lib/radix-tree.c +++ b/lib/radix-tree.c @@ -49,7 +49,7 @@ struct radix_tree_node { unsigned int height; /* Height from the bottom */ unsigned int count; struct rcu_head rcu_head; - void *slots[RADIX_TREE_MAP_SIZE]; + void __rcu *slots[RADIX_TREE_MAP_SIZE]; unsigned long tags[RADIX_TREE_MAX_TAGS][RADIX_TREE_TAG_LONGS]; }; -- cgit v1.2.2 From 910b1b7e19a292ff685001caf1bf1a9775b771a1 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 21 Jul 2010 08:05:56 -0700 Subject: rcu: Allow RCU CPU stall warnings to be off at boot, but manually enablable Currently, if RCU CPU stall warnings are enabled, they are enabled immediately upon boot. They can be manually disabled via /sys (and also re-enabled via /sys), and are automatically disabled upon panic. However, some users need RCU CPU stalls to be disabled at boot time, but to be enabled without rebuilding/rebooting. For example, someone running a real-time application in production might not want the additional latency of RCU CPU stall detection in normal operation, but might need to enable it at any point for fault isolation purposes. This commit therefore provides a new CONFIG_RCU_CPU_STALL_DETECTOR_RUNNABLE kernel configuration parameter that maintains the current behavior (enable at boot) by default, but allows a kernel to be configured with RCU CPU stall detection built into the kernel, but disabled at boot time. Requested-by: Clark Williams Requested-by: John Kacur Signed-off-by: Paul E. McKenney --- lib/Kconfig.debug | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index b5005b762103..52c2172dff16 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -856,6 +856,19 @@ config RCU_CPU_STALL_TIMEOUT RCU grace period persists, additional CPU stall warnings are printed at more widely spaced intervals. +config RCU_CPU_STALL_DETECTOR_RUNNABLE + bool "RCU CPU stall checking starts automatically at boot" + depends on RCU_CPU_STALL_DETECTOR + default y + help + If set, start checking for RCU CPU stalls immediately on + boot. Otherwise, RCU CPU stall checking must be manually + enabled. + + Say Y if you are unsure. + + Say N if you wish to suppress RCU CPU stall checking during boot. + config RCU_CPU_STALL_VERBOSE bool "Print additional per-task information for RCU_CPU_STALL_DETECTOR" depends on RCU_CPU_STALL_DETECTOR && TREE_PREEMPT_RCU -- cgit v1.2.2 From ea24ea850bcd7dd5f0994de2cf99ace10c4484cc Mon Sep 17 00:00:00 2001 From: Naohiro Aota Date: Tue, 31 Aug 2010 00:37:03 +0900 Subject: idr: fix kernel-doc warnings. Fix the following kernel-doc warnings. % perl scripts/kernel-doc lib/idr.c > /dev/null Warning(lib/idr.c:300): No description found for parameter 'starting_id' Warning(lib/idr.c:300): Excess function parameter 'start_id' description in 'idr_get_new_above' Warning(lib/idr.c:485): No description found for parameter 'idp' Warning(lib/idr.c:596): No description found for parameter 'nextidp' Warning(lib/idr.c:596): Excess function parameter 'id' description in 'idr_get_next' Warning(lib/idr.c:774): No description found for parameter 'starting_id' Warning(lib/idr.c:774): Excess function parameter 'staring_id' description in 'ida_get_new_above' Warning(lib/idr.c:918): No description found for parameter 'ida' Signed-off-by: Naohiro Aota Acked-by: Tejun Heo Signed-off-by: Jiri Kosina --- lib/idr.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'lib') diff --git a/lib/idr.c b/lib/idr.c index 7f1a4f0acf50..fb86a675dcff 100644 --- a/lib/idr.c +++ b/lib/idr.c @@ -284,7 +284,7 @@ static int idr_get_new_above_int(struct idr *idp, void *ptr, int starting_id) * idr_get_new_above - allocate new idr entry above or equal to a start id * @idp: idr handle * @ptr: pointer you want associated with the id - * @start_id: id to start search at + * @starting_id: id to start search at * @id: pointer to the allocated handle * * This is the allocate id function. It should be called with any @@ -479,7 +479,7 @@ EXPORT_SYMBOL(idr_remove_all); /** * idr_destroy - release all cached layers within an idr tree - * idp: idr handle + * @idp: idr handle */ void idr_destroy(struct idr *idp) { @@ -586,7 +586,7 @@ EXPORT_SYMBOL(idr_for_each); /** * idr_get_next - lookup next object of id to given id. * @idp: idr handle - * @id: pointer to lookup key + * @nextidp: pointer to lookup key * * Returns pointer to registered object with id, which is next number to * given id. @@ -758,7 +758,7 @@ EXPORT_SYMBOL(ida_pre_get); /** * ida_get_new_above - allocate new ID above or equal to a start id * @ida: ida handle - * @staring_id: id to start search at + * @starting_id: id to start search at * @p_id: pointer to the allocated handle * * Allocate new ID above or equal to @ida. It should be called with @@ -912,7 +912,7 @@ EXPORT_SYMBOL(ida_remove); /** * ida_destroy - release all cached layers within an ida tree - * ida: ida handle + * @ida: ida handle */ void ida_destroy(struct ida *ida) { -- cgit v1.2.2 From 1458ce166c1b333ecbaf8caa9f4f54eab3a522a7 Mon Sep 17 00:00:00 2001 From: Naohiro Aota Date: Fri, 27 Aug 2010 17:43:46 +0900 Subject: idr: describe how nextidp works in idr_get_next(). It was unclear in original kernel-doc how nextidp worked in idr_get_next(). Let's describe it. Signed-off-by: Naohiro Aota Acked-by: Tejun Heo Signed-off-by: Jiri Kosina --- lib/idr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/idr.c b/lib/idr.c index fb86a675dcff..5e0966be0f7c 100644 --- a/lib/idr.c +++ b/lib/idr.c @@ -589,7 +589,8 @@ EXPORT_SYMBOL(idr_for_each); * @nextidp: pointer to lookup key * * Returns pointer to registered object with id, which is next number to - * given id. + * given id. After being looked up, *@nextidp will be updated for the next + * iteration. */ void *idr_get_next(struct idr *idp, int *nextidp) -- cgit v1.2.2 From 46b93b74fc022885ab69dfc8e2af0562dd626638 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 31 Aug 2010 16:35:20 -0400 Subject: tracing/lockdep: Fix dependency of TRACE_IRQFLAGS When CONFIG_IRQSOFF_TRACER is set and CONFIG_PROVE_LOCKING is not, we get the following error: $ make oldconfig scripts/kconfig/conf --oldconfig arch/x86/Kconfig warning: (IRQSOFF_TRACER && TRACING_SUPPORT && FTRACE && TRACE_IRQFLAGS_SUPPORT && !ARCH_USES_GETTIMEOFFSET) selects TRACE_IRQFLAGS which has unmet direct dependencies (DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && PROVE_LOCKING) warning: (IRQSOFF_TRACER && TRACING_SUPPORT && FTRACE && TRACE_IRQFLAGS_SUPPORT && !ARCH_USES_GETTIMEOFFSET) selects TRACE_IRQFLAGS which has unmet direct dependencies (DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && PROVE_LOCKING) This is because IRQSOFF_TRACER selects TRACE_IRQFLAGS but TRACE_IRQFLAGS has PROVE_LOCKING as a dependency. This code is incorrect, and this patch changes the TRACE_IRQFLAGS to be just a simple bool that does not depend or select anything. Instead both IRQSOFF_TRACER and PROVE_LOCKING select it. Reported-by: Richard Kennedy Signed-off-by: Steven Rostedt --- lib/Kconfig.debug | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1b4afd2e6ca0..e85d549b6eac 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -482,6 +482,7 @@ config PROVE_LOCKING select DEBUG_SPINLOCK select DEBUG_MUTEXES select DEBUG_LOCK_ALLOC + select TRACE_IRQFLAGS default n help This feature enables the kernel to prove that all locking @@ -579,11 +580,10 @@ config DEBUG_LOCKDEP of more runtime overhead. config TRACE_IRQFLAGS - depends on DEBUG_KERNEL bool - default y - depends on TRACE_IRQFLAGS_SUPPORT - depends on PROVE_LOCKING + help + Enables hooks to interrupt enabling and disabling for + either tracing or lock debugging. config DEBUG_SPINLOCK_SLEEP bool "Spinlock debugging: sleep-inside-spinlock checking" -- cgit v1.2.2 From 52159d98be6f26c48f5e02c7ab3c9848a85979b5 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Fri, 17 Sep 2010 11:09:17 -0400 Subject: jump label: Convert dynamic debug to use jump labels Convert the 'dynamic debug' infrastructure to use jump labels. Signed-off-by: Jason Baron LKML-Reference: Signed-off-by: Steven Rostedt --- lib/dynamic_debug.c | 42 +++--------------------------------------- 1 file changed, 3 insertions(+), 39 deletions(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index 02afc2533728..e925c7b960f1 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -26,19 +26,11 @@ #include #include #include +#include extern struct _ddebug __start___verbose[]; extern struct _ddebug __stop___verbose[]; -/* dynamic_debug_enabled, and dynamic_debug_enabled2 are bitmasks in which - * bit n is set to 1 if any modname hashes into the bucket n, 0 otherwise. They - * use independent hash functions, to reduce the chance of false positives. - */ -long long dynamic_debug_enabled; -EXPORT_SYMBOL_GPL(dynamic_debug_enabled); -long long dynamic_debug_enabled2; -EXPORT_SYMBOL_GPL(dynamic_debug_enabled2); - struct ddebug_table { struct list_head link; char *mod_name; @@ -87,26 +79,6 @@ static char *ddebug_describe_flags(struct _ddebug *dp, char *buf, return buf; } -/* - * must be called with ddebug_lock held - */ - -static int disabled_hash(char hash, bool first_table) -{ - struct ddebug_table *dt; - char table_hash_value; - - list_for_each_entry(dt, &ddebug_tables, link) { - if (first_table) - table_hash_value = dt->ddebugs->primary_hash; - else - table_hash_value = dt->ddebugs->secondary_hash; - if (dt->num_enabled && (hash == table_hash_value)) - return 0; - } - return 1; -} - /* * Search the tables for _ddebug's which match the given * `query' and apply the `flags' and `mask' to them. Tells @@ -170,17 +142,9 @@ static void ddebug_change(const struct ddebug_query *query, dt->num_enabled++; dp->flags = newflags; if (newflags) { - dynamic_debug_enabled |= - (1LL << dp->primary_hash); - dynamic_debug_enabled2 |= - (1LL << dp->secondary_hash); + enable_jump_label(&dp->enabled); } else { - if (disabled_hash(dp->primary_hash, true)) - dynamic_debug_enabled &= - ~(1LL << dp->primary_hash); - if (disabled_hash(dp->secondary_hash, false)) - dynamic_debug_enabled2 &= - ~(1LL << dp->secondary_hash); + disable_jump_label(&dp->enabled); } if (verbose) printk(KERN_INFO -- cgit v1.2.2 From 2dfbf4dfbe47a484bae20456c12b40763b9b6af7 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 15 Sep 2010 23:30:48 -0700 Subject: rcu: Add advice to PROVE_RCU_REPEATEDLY kernel config parameter The PROVE_RCU_REPEATEDLY has no "Say Y"/"Say N" advice, so this commit adds it. Reported-by: Johannes Berg Signed-off-by: Paul E. McKenney --- lib/Kconfig.debug | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 52c2172dff16..9886cf5365ba 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -539,6 +539,10 @@ config PROVE_RCU_REPEATEDLY disabling, allowing multiple RCU-lockdep warnings to be printed on a single reboot. + Say Y to allow multiple RCU-lockdep warnings per boot. + + Say N if you are unsure. + config SPARSE_RCU_POINTER bool "RCU debugging: sparse-based checks for pointer usage" default n -- cgit v1.2.2 From ab4d5ed5eeda4f57c50d14131ce1b1da75d0c938 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Tue, 5 Oct 2010 13:57:26 -0500 Subject: slub: Enable sysfs support for !CONFIG_SLUB_DEBUG Currently disabling CONFIG_SLUB_DEBUG also disabled SYSFS support meaning that the slabs cannot be tuned without DEBUG. Make SYSFS support independent of CONFIG_SLUB_DEBUG Signed-off-by: Christoph Lameter Signed-off-by: Pekka Enberg --- lib/Kconfig.debug | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1b4afd2e6ca0..b6263651a955 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -353,7 +353,7 @@ config SLUB_DEBUG_ON config SLUB_STATS default n bool "Enable SLUB performance statistics" - depends on SLUB && SLUB_DEBUG && SYSFS + depends on SLUB && SYSFS help SLUB statistics are useful to debug SLUBs allocation behavior in order find ways to optimize the allocator. This should never be -- cgit v1.2.2 From 400fb7f6a0cfe13025cb0296fdb4737da7025a8a Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Thu, 7 Oct 2010 15:25:04 -0700 Subject: move async raid6 test to lib/Kconfig.debug The prompt for "Self test for hardware accelerated raid6 recovery" does not belong in the top level configuration menu. All the options in crypto/async_tx/Kconfig are selected and do not depend on CRYPTO. Kconfig.debug seems like a reasonable fit. Cc: Herbert Xu Cc: David Woodhouse Signed-off-by: Dan Williams --- lib/Kconfig.debug | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1b4afd2e6ca0..0deaf81ac111 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -1150,6 +1150,19 @@ config ATOMIC64_SELFTEST If unsure, say N. +config ASYNC_RAID6_TEST + tristate "Self test for hardware accelerated raid6 recovery" + depends on ASYNC_RAID6_RECOV + select ASYNC_MEMCPY + ---help--- + This is a one-shot self test that permutes through the + recovery of all the possible two disk failure scenarios for a + N-disk array. Recovery is performed with the asynchronous + raid6 recovery routines, and will optionally use an offload + engine if one is available. + + If unsure, say N. + source "samples/Kconfig" source "lib/Kconfig.kgdb" -- cgit v1.2.2 From 03620b2d7545fee29ecd3bb270f206b9cff75e58 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Mon, 2 Aug 2010 23:48:06 +0900 Subject: swiotlb: make io_tlb_overflow static We don't need to export io_tlb_overflow_buffer. I'll remove io_tlb_overflow_buffer completely in the long term though. Signed-off-by: FUJITA Tomonori Signed-off-by: Konrad Rzeszutek Wilk --- lib/swiotlb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/swiotlb.c b/lib/swiotlb.c index 34e3082632d8..b237e649cd5d 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -70,7 +70,7 @@ static unsigned long io_tlb_nslabs; */ static unsigned long io_tlb_overflow = 32*1024; -void *io_tlb_overflow_buffer; +static void *io_tlb_overflow_buffer; /* * This is a free list describing the number of free entries available from -- cgit v1.2.2 From e79f86b2ef9c0a8c47225217c1018b7d3d90101c Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 11 Oct 2010 10:40:25 -0700 Subject: swiotlb: Use page alignment for early buffer allocation We could call free_bootmem_late() if swiotlb is not used, and it will shrink to page alignment. So alloc them with page alignment at first, to avoid lose two pages before patch: [ 0.000000] memblock_x86_reserve_range: [00d3600000, 00d7600000] swiotlb buffer [ 0.000000] memblock_x86_reserve_range: [00d7e7ef40, 00d7e9ef40] swiotlb list [ 0.000000] memblock_x86_reserve_range: [00d7e3ef40, 00d7e7ef40] swiotlb orig_ad [ 0.000000] memblock_x86_reserve_range: [000008a000, 0000092000] swiotlb overflo after patch will get [ 0.000000] memblock_x86_reserve_range: [00d3600000, 00d7600000] swiotlb buffer [ 0.000000] memblock_x86_reserve_range: [00d7e7e000, 00d7e9e000] swiotlb list [ 0.000000] memblock_x86_reserve_range: [00d7e3e000, 00d7e7e000] swiotlb orig_ad [ 0.000000] memblock_x86_reserve_range: [000008a000, 0000092000] swiotlb overflo Signed-off-by: Yinghai Lu Acked-by: FUJITA Tomonori Cc: Becky Bruce Signed-off-by: Konrad Rzeszutek Wilk --- lib/swiotlb.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) (limited to 'lib') diff --git a/lib/swiotlb.c b/lib/swiotlb.c index b237e649cd5d..7c06ee51a29a 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -147,16 +147,16 @@ void __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose) * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE * between io_tlb_start and io_tlb_end. */ - io_tlb_list = alloc_bootmem(io_tlb_nslabs * sizeof(int)); + io_tlb_list = alloc_bootmem_pages(PAGE_ALIGN(io_tlb_nslabs * sizeof(int))); for (i = 0; i < io_tlb_nslabs; i++) io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE); io_tlb_index = 0; - io_tlb_orig_addr = alloc_bootmem(io_tlb_nslabs * sizeof(phys_addr_t)); + io_tlb_orig_addr = alloc_bootmem_pages(PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t))); /* * Get the overflow emergency buffer */ - io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow); + io_tlb_overflow_buffer = alloc_bootmem_low_pages(PAGE_ALIGN(io_tlb_overflow)); if (!io_tlb_overflow_buffer) panic("Cannot allocate SWIOTLB overflow buffer!\n"); if (verbose) @@ -182,7 +182,7 @@ swiotlb_init_with_default_size(size_t default_size, int verbose) /* * Get IO TLB memory from the low pages */ - io_tlb_start = alloc_bootmem_low_pages(bytes); + io_tlb_start = alloc_bootmem_low_pages(PAGE_ALIGN(bytes)); if (!io_tlb_start) panic("Cannot allocate SWIOTLB buffer"); @@ -308,13 +308,13 @@ void __init swiotlb_free(void) get_order(io_tlb_nslabs << IO_TLB_SHIFT)); } else { free_bootmem_late(__pa(io_tlb_overflow_buffer), - io_tlb_overflow); + PAGE_ALIGN(io_tlb_overflow)); free_bootmem_late(__pa(io_tlb_orig_addr), - io_tlb_nslabs * sizeof(phys_addr_t)); + PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t))); free_bootmem_late(__pa(io_tlb_list), - io_tlb_nslabs * sizeof(int)); + PAGE_ALIGN(io_tlb_nslabs * sizeof(int))); free_bootmem_late(__pa(io_tlb_start), - io_tlb_nslabs << IO_TLB_SHIFT); + PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); } } -- cgit v1.2.2 From 6b945df74233386aab526cddd3593ed4b854f574 Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Thu, 14 Oct 2010 14:54:47 -0400 Subject: kmemleak: add TILE to the list of supported architectures. All the necessary functionality was already there; we just need to make it possible to select the config option. Signed-off-by: Chris Metcalf --- lib/Kconfig.debug | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1b4afd2e6ca0..9f211b2642f5 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -366,7 +366,7 @@ config SLUB_STATS config DEBUG_KMEMLEAK bool "Kernel memory leak detector" depends on DEBUG_KERNEL && EXPERIMENTAL && !MEMORY_HOTPLUG && \ - (X86 || ARM || PPC || S390 || SPARC64 || SUPERH || MICROBLAZE) + (X86 || ARM || PPC || S390 || SPARC64 || SUPERH || MICROBLAZE || TILE) select DEBUG_FS if SYSFS select STACKTRACE if STACKTRACE_SUPPORT -- cgit v1.2.2 From 6038f373a3dc1f1c26496e60b6c40b164716f07e Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sun, 15 Aug 2010 18:52:59 +0200 Subject: llseek: automatically add .llseek fop All file_operations should get a .llseek operation so we can make nonseekable_open the default for future file operations without a .llseek pointer. The three cases that we can automatically detect are no_llseek, seq_lseek and default_llseek. For cases where we can we can automatically prove that the file offset is always ignored, we use noop_llseek, which maintains the current behavior of not returning an error from a seek. New drivers should normally not use noop_llseek but instead use no_llseek and call nonseekable_open at open time. Existing drivers can be converted to do the same when the maintainer knows for certain that no user code relies on calling seek on the device file. The generated code is often incorrectly indented and right now contains comments that clarify for each added line why a specific variant was chosen. In the version that gets submitted upstream, the comments will be gone and I will manually fix the indentation, because there does not seem to be a way to do that using coccinelle. Some amount of new code is currently sitting in linux-next that should get the same modifications, which I will do at the end of the merge window. Many thanks to Julia Lawall for helping me learn to write a semantic patch that does all this. ===== begin semantic patch ===== // This adds an llseek= method to all file operations, // as a preparation for making no_llseek the default. // // The rules are // - use no_llseek explicitly if we do nonseekable_open // - use seq_lseek for sequential files // - use default_llseek if we know we access f_pos // - use noop_llseek if we know we don't access f_pos, // but we still want to allow users to call lseek // @ open1 exists @ identifier nested_open; @@ nested_open(...) { <+... nonseekable_open(...) ...+> } @ open exists@ identifier open_f; identifier i, f; identifier open1.nested_open; @@ int open_f(struct inode *i, struct file *f) { <+... ( nonseekable_open(...) | nested_open(...) ) ...+> } @ read disable optional_qualifier exists @ identifier read_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; expression E; identifier func; @@ ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off) { <+... ( *off = E | *off += E | func(..., off, ...) | E = *off ) ...+> } @ read_no_fpos disable optional_qualifier exists @ identifier read_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; @@ ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off) { ... when != off } @ write @ identifier write_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; expression E; identifier func; @@ ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off) { <+... ( *off = E | *off += E | func(..., off, ...) | E = *off ) ...+> } @ write_no_fpos @ identifier write_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; @@ ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off) { ... when != off } @ fops0 @ identifier fops; @@ struct file_operations fops = { ... }; @ has_llseek depends on fops0 @ identifier fops0.fops; identifier llseek_f; @@ struct file_operations fops = { ... .llseek = llseek_f, ... }; @ has_read depends on fops0 @ identifier fops0.fops; identifier read_f; @@ struct file_operations fops = { ... .read = read_f, ... }; @ has_write depends on fops0 @ identifier fops0.fops; identifier write_f; @@ struct file_operations fops = { ... .write = write_f, ... }; @ has_open depends on fops0 @ identifier fops0.fops; identifier open_f; @@ struct file_operations fops = { ... .open = open_f, ... }; // use no_llseek if we call nonseekable_open //////////////////////////////////////////// @ nonseekable1 depends on !has_llseek && has_open @ identifier fops0.fops; identifier nso ~= "nonseekable_open"; @@ struct file_operations fops = { ... .open = nso, ... +.llseek = no_llseek, /* nonseekable */ }; @ nonseekable2 depends on !has_llseek @ identifier fops0.fops; identifier open.open_f; @@ struct file_operations fops = { ... .open = open_f, ... +.llseek = no_llseek, /* open uses nonseekable */ }; // use seq_lseek for sequential files ///////////////////////////////////// @ seq depends on !has_llseek @ identifier fops0.fops; identifier sr ~= "seq_read"; @@ struct file_operations fops = { ... .read = sr, ... +.llseek = seq_lseek, /* we have seq_read */ }; // use default_llseek if there is a readdir /////////////////////////////////////////// @ fops1 depends on !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier readdir_e; @@ // any other fop is used that changes pos struct file_operations fops = { ... .readdir = readdir_e, ... +.llseek = default_llseek, /* readdir is present */ }; // use default_llseek if at least one of read/write touches f_pos ///////////////////////////////////////////////////////////////// @ fops2 depends on !fops1 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read.read_f; @@ // read fops use offset struct file_operations fops = { ... .read = read_f, ... +.llseek = default_llseek, /* read accesses f_pos */ }; @ fops3 depends on !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier write.write_f; @@ // write fops use offset struct file_operations fops = { ... .write = write_f, ... + .llseek = default_llseek, /* write accesses f_pos */ }; // Use noop_llseek if neither read nor write accesses f_pos /////////////////////////////////////////////////////////// @ fops4 depends on !fops1 && !fops2 && !fops3 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read_no_fpos.read_f; identifier write_no_fpos.write_f; @@ // write fops use offset struct file_operations fops = { ... .write = write_f, .read = read_f, ... +.llseek = noop_llseek, /* read and write both use no f_pos */ }; @ depends on has_write && !has_read && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier write_no_fpos.write_f; @@ struct file_operations fops = { ... .write = write_f, ... +.llseek = noop_llseek, /* write uses no f_pos */ }; @ depends on has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read_no_fpos.read_f; @@ struct file_operations fops = { ... .read = read_f, ... +.llseek = noop_llseek, /* read uses no f_pos */ }; @ depends on !has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; @@ struct file_operations fops = { ... +.llseek = noop_llseek, /* no read or write fn */ }; ===== End semantic patch ===== Signed-off-by: Arnd Bergmann Cc: Julia Lawall Cc: Christoph Hellwig --- lib/dma-debug.c | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/dma-debug.c b/lib/dma-debug.c index 01e64270e246..4bfb0471f106 100644 --- a/lib/dma-debug.c +++ b/lib/dma-debug.c @@ -590,6 +590,7 @@ out_unlock: static const struct file_operations filter_fops = { .read = filter_read, .write = filter_write, + .llseek = default_llseek, }; static int dma_debug_fs_init(void) -- cgit v1.2.2 From 3b6e901f839f42afb40f614418df82c08b01320a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 14 Oct 2010 21:10:38 +0200 Subject: jump_label: Use more consistent naming Now that there's still only a few users around, rename things to make them more consistent. Signed-off-by: Peter Zijlstra LKML-Reference: <20101014203625.448565169@chello.nl> Signed-off-by: Ingo Molnar --- lib/dynamic_debug.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index e925c7b960f1..7bd6df781ce5 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -142,9 +142,9 @@ static void ddebug_change(const struct ddebug_query *query, dt->num_enabled++; dp->flags = newflags; if (newflags) { - enable_jump_label(&dp->enabled); + jump_label_enable(&dp->enabled); } else { - disable_jump_label(&dp->enabled); + jump_label_disable(&dp->enabled); } if (verbose) printk(KERN_INFO -- cgit v1.2.2 From 6de5bd128d381ad88ac6d419a5e597048eb468cf Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sat, 11 Sep 2010 18:00:57 +0200 Subject: BKL: introduce CONFIG_BKL. With all the patches we have queued in the BKL removal tree, only a few dozen modules are left that actually rely on the BKL, and even there are lots of low-hanging fruit. We need to decide what to do about them, this patch illustrates one of the options: Every user of the BKL is marked as 'depends on BKL' in Kconfig, and the CONFIG_BKL becomes a user-visible option. If it gets disabled, no BKL using module can be built any more and the BKL code itself is compiled out. The one exception is file locking, which is practically always enabled and does a 'select BKL' instead. This effectively forces CONFIG_BKL to be enabled until we have solved the fs/lockd mess and can apply the patch that removes the BKL from fs/locks.c. Signed-off-by: Arnd Bergmann --- lib/Kconfig.debug | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1b4afd2e6ca0..088eea1c2bef 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -461,6 +461,15 @@ config DEBUG_MUTEXES This feature allows mutex semantics violations to be detected and reported. +config BKL + bool "Big Kernel Lock" if (SMP || PREEMPT) + default y + help + This is the traditional lock that is used in old code instead + of proper locking. All drivers that use the BKL should depend + on this symbol. + Say Y here unless you are working on removing the BKL. + config DEBUG_LOCK_ALLOC bool "Lock debugging: detect incorrect freeing of live locks" depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT -- cgit v1.2.2 From fd89cfb8718753459fcea3fe6103d19de5e86c9b Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 6 Aug 2010 16:11:01 +0200 Subject: Dynamic Debug: Split out query string parsing/setup from proc_write The parsing and applying of dynamic debug strings is not only useful for /sys/../dynamic_debug/control write access, but can also be used for boot parameter parsing. The boot parameter is introduced in a follow up patch. Signed-off-by: Thomas Renninger Acked-by: jbaron@redhat.com Acked-by: Pekka Enberg Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 40 +++++++++++++++++++++++++--------------- 1 file changed, 25 insertions(+), 15 deletions(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index 02afc2533728..84d103c474e4 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -429,6 +429,27 @@ static int ddebug_parse_flags(const char *str, unsigned int *flagsp, return 0; } +static int ddebug_exec_query(char *query_string) +{ + unsigned int flags = 0, mask = 0; + struct ddebug_query query; +#define MAXWORDS 9 + int nwords; + char *words[MAXWORDS]; + + nwords = ddebug_tokenize(query_string, words, MAXWORDS); + if (nwords <= 0) + return -EINVAL; + if (ddebug_parse_query(words, nwords-1, &query)) + return -EINVAL; + if (ddebug_parse_flags(words[nwords-1], &flags, &mask)) + return -EINVAL; + + /* actually go and implement the change */ + ddebug_change(&query, flags, mask); + return 0; +} + /* * File_ops->write method for /dynamic_debug/conrol. Gathers the * command text from userspace, parses and executes it. @@ -436,12 +457,8 @@ static int ddebug_parse_flags(const char *str, unsigned int *flagsp, static ssize_t ddebug_proc_write(struct file *file, const char __user *ubuf, size_t len, loff_t *offp) { - unsigned int flags = 0, mask = 0; - struct ddebug_query query; -#define MAXWORDS 9 - int nwords; - char *words[MAXWORDS]; char tmpbuf[256]; + int ret; if (len == 0) return 0; @@ -455,16 +472,9 @@ static ssize_t ddebug_proc_write(struct file *file, const char __user *ubuf, printk(KERN_INFO "%s: read %d bytes from userspace\n", __func__, (int)len); - nwords = ddebug_tokenize(tmpbuf, words, MAXWORDS); - if (nwords <= 0) - return -EINVAL; - if (ddebug_parse_query(words, nwords-1, &query)) - return -EINVAL; - if (ddebug_parse_flags(words[nwords-1], &flags, &mask)) - return -EINVAL; - - /* actually go and implement the change */ - ddebug_change(&query, flags, mask); + ret = ddebug_exec_query(tmpbuf); + if (ret) + return ret; *offp += len; return len; -- cgit v1.2.2 From a648ec05bb950fae2f35d0490ddd6cf15010af72 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 6 Aug 2010 16:11:02 +0200 Subject: Dynamic Debug: Introduce ddebug_query= boot parameter Dynamic debug lacks the ability to enable debug messages at boot time. One could patch initramfs or service startup scripts to write to /sys/../dynamic_debug/control, but this sucks. This patch makes it possible to pass a query in the same format one can write to /sys/../dynamic_debug/control via boot param. When dynamic debug gets initialized, this query will automatically be applied. Signed-off-by: Thomas Renninger Acked-by: jbaron@redhat.com Acked-by: Pekka Enberg Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index 84d103c474e4..44ce66bdb211 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -450,6 +450,19 @@ static int ddebug_exec_query(char *query_string) return 0; } +static __initdata char ddebug_setup_string[1024]; +static __init int ddebug_setup_query(char *str) +{ + if (strlen(str) >= 1024) { + pr_warning("ddebug boot param string too large\n"); + return 0; + } + strcpy(ddebug_setup_string, str); + return 1; +} + +__setup("ddebug_query=", ddebug_setup_query); + /* * File_ops->write method for /dynamic_debug/conrol. Gathers the * command text from userspace, parses and executes it. @@ -769,6 +782,18 @@ static int __init dynamic_debug_init(void) } ret = ddebug_add_module(iter_start, n, modname); } + + /* ddebug_query boot param got passed -> set it up */ + if (ddebug_setup_string[0] != '\0') { + ret = ddebug_exec_query(ddebug_setup_string); + if (ret) + pr_warning("Invalid ddebug boot param %s", + ddebug_setup_string); + else + pr_info("ddebug initialized with string %s", + ddebug_setup_string); + } + out_free: if (ret) { ddebug_remove_all_tables(); -- cgit v1.2.2 From 6a5c083de2f5fbf89a4b0a251be2c2205434d7ea Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 6 Aug 2010 16:11:03 +0200 Subject: Dynamic Debug: Initialize dynamic debug earlier via arch_initcall Having the ddebug_query= boot parameter it makes sense to set up dynamic debug as soon as possible. I expect sysfs files cannot be set up via an arch_initcall, because this one is even before fs_initcall. Therefore I splitted the dynamic_debug_init function into an early one and a later one providing /sys/../dynamic_debug/control file. Possibly dynamic_debug can be initialized even earlier, not sure whether this still makes sense then. I picked up arch_initcall as it covers quite a lot already. Dynamic debug needs to allocate memory, therefore it's not easily possible to set it up even before the command line gets parsed. Therefore the boot param query string is stored in a temp string which is applied when dynamic debug gets set up. This has been tested with ddebug_query="file ec.c +p" and I could retrieve pr_debug() messages early at boot during ACPI setup: ACPI: EC: Look up EC in DSDT ACPI: EC: ---> status = 0x08 ACPI: EC: transaction start ACPI: EC: <--- command = 0x80 ACPI: EC: ~~~> interrupt ACPI: EC: ---> status = 0x08 ACPI: EC: <--- data = 0xa4 ... ACPI: Interpreter enabled ACPI: (supports S0 S3 S4 S5) ACPI: Using IOAPIC for interrupt routing ACPI: EC: ---> status = 0x00 ACPI: EC: transaction start ACPI: EC: <--- command = 0x80 Signed-off-by: Thomas Renninger Acked-by: jbaron@redhat.com Acked-by: Pekka Enberg CC: linux-acpi@vger.kernel.org Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 33 +++++++++++++++++++++++---------- 1 file changed, 23 insertions(+), 10 deletions(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index 44ce66bdb211..a687d902daab 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -748,13 +748,14 @@ static void ddebug_remove_all_tables(void) mutex_unlock(&ddebug_lock); } -static int __init dynamic_debug_init(void) +static __initdata int ddebug_init_success; + +static int __init dynamic_debug_init_debugfs(void) { struct dentry *dir, *file; - struct _ddebug *iter, *iter_start; - const char *modname = NULL; - int ret = 0; - int n = 0; + + if (!ddebug_init_success) + return -ENODEV; dir = debugfs_create_dir("dynamic_debug", NULL); if (!dir) @@ -765,6 +766,16 @@ static int __init dynamic_debug_init(void) debugfs_remove(dir); return -ENOMEM; } + return 0; +} + +static int __init dynamic_debug_init(void) +{ + struct _ddebug *iter, *iter_start; + const char *modname = NULL; + int ret = 0; + int n = 0; + if (__start___verbose != __stop___verbose) { iter = __start___verbose; modname = iter->modname; @@ -795,11 +806,13 @@ static int __init dynamic_debug_init(void) } out_free: - if (ret) { + if (ret) ddebug_remove_all_tables(); - debugfs_remove(dir); - debugfs_remove(file); - } + else + ddebug_init_success = 1; return 0; } -module_init(dynamic_debug_init); +/* Allow early initialization for boot messages via boot param */ +arch_initcall(dynamic_debug_init); +/* Debugfs setup must be done later */ +module_init(dynamic_debug_init_debugfs); -- cgit v1.2.2 From c25d1dfbd403209025df41a737f82ce8f43d93f5 Mon Sep 17 00:00:00 2001 From: Robin Holt Date: Wed, 29 Sep 2010 14:00:54 -0500 Subject: kobject: Introduce kset_find_obj_hinted. One call chain getting to kset_find_obj is: link_mem_sections() find_mem_section() kset_find_obj() This is done during boot. The memory sections were added in a linearly increasing order and link_mem_sections tends to utilize them in that same linear order. Introduce a kset_find_obj_hinted which is passed the result of the previous kset_find_obj which it uses for a quick "is the next object our desired object" check before falling back to the old behavior. Signed-off-by: Robin Holt To: Robert P. J. Day Reviewed-by: KAMEZAWA Hiroyuki Signed-off-by: Greg Kroah-Hartman --- lib/kobject.c | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) (limited to 'lib') diff --git a/lib/kobject.c b/lib/kobject.c index f07c57252e82..82dc34c095c2 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -745,18 +745,57 @@ void kset_unregister(struct kset *k) * take a reference and return the object. */ struct kobject *kset_find_obj(struct kset *kset, const char *name) +{ + return kset_find_obj_hinted(kset, name, NULL); +} + +/** + * kset_find_obj_hinted - search for object in kset given a predecessor hint. + * @kset: kset we're looking in. + * @name: object's name. + * @hint: hint to possible object's predecessor. + * + * Check the hint's next object and if it is a match return it directly, + * otherwise, fall back to the behavior of kset_find_obj(). Either way + * a reference for the returned object is held and the reference on the + * hinted object is released. + */ +struct kobject *kset_find_obj_hinted(struct kset *kset, const char *name, + struct kobject *hint) { struct kobject *k; struct kobject *ret = NULL; spin_lock(&kset->list_lock); + + if (!hint) + goto slow_search; + + /* end of list detection */ + if (hint->entry.next == kset->list.next) + goto slow_search; + + k = container_of(hint->entry.next, struct kobject, entry); + if (!kobject_name(k) || strcmp(kobject_name(k), name)) + goto slow_search; + + ret = kobject_get(k); + goto unlock_exit; + +slow_search: list_for_each_entry(k, &kset->list, entry) { if (kobject_name(k) && !strcmp(kobject_name(k), name)) { ret = kobject_get(k); break; } } + +unlock_exit: spin_unlock(&kset->list_lock); + + if (hint) + kobject_put(hint); + return ret; } -- cgit v1.2.2 From 8474b591faf3bb0a1e08a60d21d6baac498f15e4 Mon Sep 17 00:00:00 2001 From: Masanori ITOH Date: Tue, 26 Oct 2010 14:21:20 -0700 Subject: percpu: fix list_head init bug in __percpu_counter_init() WARNING: at lib/list_debug.c:26 __list_add+0x3f/0x81() Hardware name: Express5800/B120a [N8400-085] list_add corruption. next->prev should be prev (ffffffff81a7ea00), but was dead000000200200. (next=ffff88080b872d58). Modules linked in: aoe ipt_MASQUERADE iptable_nat nf_nat autofs4 sunrpc bridge 8021q garp stp llc ipv6 cpufreq_ondemand acpi_cpufreq freq_table dm_round_robin dm_multipath kvm_intel kvm uinput lpfc scsi_transport_fc igb ioatdma scsi_tgt i2c_i801 i2c_core dca iTCO_wdt iTCO_vendor_support pcspkr shpchp megaraid_sas [last unloaded: aoe] Pid: 54, comm: events/3 Tainted: G W 2.6.34-vanilla1 #1 Call Trace: [] warn_slowpath_common+0x7c/0x94 [] warn_slowpath_fmt+0x41/0x43 [] __list_add+0x3f/0x81 [] __percpu_counter_init+0x59/0x6b [] bdi_init+0x118/0x17e [] blk_alloc_queue_node+0x79/0x143 [] blk_alloc_queue+0x11/0x13 [] aoeblk_gdalloc+0x8e/0x1c9 [aoe] [] aoecmd_sleepwork+0x25/0xa8 [aoe] [] worker_thread+0x1a9/0x237 [] ? aoecmd_sleepwork+0x0/0xa8 [aoe] [] ? autoremove_wake_function+0x0/0x39 [] ? worker_thread+0x0/0x237 [] kthread+0x7f/0x87 [] kernel_thread_helper+0x4/0x10 [] ? kthread+0x0/0x87 [] ? kernel_thread_helper+0x0/0x10 It's because there is no initialization code for a list_head contained in the struct backing_dev_info under CONFIG_HOTPLUG_CPU, and the bug comes up when block device drivers calling blk_alloc_queue() are used. In case of me, I got them by using aoe. Signed-off-by: Masanori Itoh Cc: Tejun Heo Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/percpu_counter.c | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/percpu_counter.c b/lib/percpu_counter.c index ec9048e74f44..209448e1d2b9 100644 --- a/lib/percpu_counter.c +++ b/lib/percpu_counter.c @@ -76,6 +76,7 @@ int __percpu_counter_init(struct percpu_counter *fbc, s64 amount, if (!fbc->counters) return -ENOMEM; #ifdef CONFIG_HOTPLUG_CPU + INIT_LIST_HEAD(&fbc->list); mutex_lock(&percpu_counters_lock); list_add(&fbc->list, &percpu_counters); mutex_unlock(&percpu_counters_lock); -- cgit v1.2.2 From 5e0579812834ab7fa072db4a15ebdff68d62e2e7 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 26 Oct 2010 14:22:50 -0700 Subject: vsprintf.c: use default pointer field size for "(null)" strings It might be nicer to align the output. For instance, ACPI messages sometimes have "(null)" pointers. $ dmesg | grep "(null)" -A 1 -B 1 [ 0.198733] ACPI: Dynamic OEM Table Load: [ 0.198745] ACPI: SSDT (null) 00239 (v02 PmRef Cpu0Ist 00003000 INTL 20051117) [ 0.199294] ACPI: SSDT 7f596e10 001C7 (v02 PmRef Cpu0Cst 00003001 INTL 20051117) [ 0.200708] ACPI: Dynamic OEM Table Load: [ 0.200721] ACPI: SSDT (null) 001C7 (v02 PmRef Cpu0Cst 00003001 INTL 20051117) [ 0.201950] ACPI: SSDT 7f597f10 000D0 (v02 PmRef Cpu1Ist 00003000 INTL 20051117) [ 0.203386] ACPI: Dynamic OEM Table Load: [ 0.203398] ACPI: SSDT (null) 000D0 (v02 PmRef Cpu1Ist 00003000 INTL 20051117) [ 0.203871] ACPI: SSDT 7f595f10 00083 (v02 PmRef Cpu1Cst 00003000 INTL 20051117) [ 0.205301] ACPI: Dynamic OEM Table Load: [ 0.205315] ACPI: SSDT (null) 00083 (v02 PmRef Cpu1Cst 00003000 INTL 20051117) [akpm@linux-foundation.org: add code comment] Signed-off-by: Joe Perches Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/vsprintf.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 7af9d841c43b..8378c136b6e1 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -988,8 +988,15 @@ static noinline_for_stack char *pointer(const char *fmt, char *buf, char *end, void *ptr, struct printf_spec spec) { - if (!ptr) + if (!ptr) { + /* + * Print (null) with the same width as a pointer so it makes + * tabular output look nice. + */ + if (spec.field_width == -1) + spec.field_width = 2 * sizeof(void *); return string(buf, end, "(null)", spec); + } switch (*fmt) { case 'F': @@ -1031,7 +1038,7 @@ char *pointer(const char *fmt, char *buf, char *end, void *ptr, } spec.flags |= SMALL; if (spec.field_width == -1) { - spec.field_width = 2*sizeof(void *); + spec.field_width = 2 * sizeof(void *); spec.flags |= ZEROPAD; } spec.base = 16; -- cgit v1.2.2 From b903c0b8899b46829a9b80ba55b61079b35940ec Mon Sep 17 00:00:00 2001 From: Changli Gao Date: Tue, 26 Oct 2010 14:22:50 -0700 Subject: lib: fix scnprintf() if @size is == 0 scnprintf() should return 0 if @size is == 0. Update the comment for it, as @size is unsigned. Signed-off-by: Changli Gao Cc: Ingo Molnar Cc: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/vsprintf.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 8378c136b6e1..c150d3dafff4 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -1504,7 +1504,7 @@ EXPORT_SYMBOL(snprintf); * @...: Arguments for the format string * * The return value is the number of characters written into @buf not including - * the trailing '\0'. If @size is <= 0 the function returns 0. + * the trailing '\0'. If @size is == 0 the function returns 0. */ int scnprintf(char *buf, size_t size, const char *fmt, ...) @@ -1516,7 +1516,11 @@ int scnprintf(char *buf, size_t size, const char *fmt, ...) i = vsnprintf(buf, size, fmt, args); va_end(args); - return (i >= size) ? (size - 1) : i; + if (likely(i < size)) + return i; + if (size != 0) + return size - 1; + return 0; } EXPORT_SYMBOL(scnprintf); -- cgit v1.2.2 From 66f1991bc2357436498ac990302b6f5bf403d7ef Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Tue, 26 Oct 2010 14:23:03 -0700 Subject: lib/bitmap.c: use hex_to_bin() Signed-off-by: Andy Shevchenko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/bitmap.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/bitmap.c b/lib/bitmap.c index ffb78c916ccd..741fae905ae3 100644 --- a/lib/bitmap.c +++ b/lib/bitmap.c @@ -359,7 +359,6 @@ EXPORT_SYMBOL(bitmap_find_next_zero_area); #define CHUNKSZ 32 #define nbits_to_hold_value(val) fls(val) -#define unhex(c) (isdigit(c) ? (c - '0') : (toupper(c) - 'A' + 10)) #define BASEDEC 10 /* fancier cpuset lists input in decimal */ /** @@ -466,7 +465,7 @@ int __bitmap_parse(const char *buf, unsigned int buflen, if (chunk & ~((1UL << (CHUNKSZ - 4)) - 1)) return -EOVERFLOW; - chunk = (chunk << 4) | unhex(c); + chunk = (chunk << 4) | hex_to_bin(c); ndigits++; totaldigits++; } if (ndigits == 0) -- cgit v1.2.2 From 066a9be6c0124edc9527088231f03c6236be375d Mon Sep 17 00:00:00 2001 From: Naohiro Aota Date: Tue, 26 Oct 2010 14:23:03 -0700 Subject: idr: fix idr_pre_get() locking description Despite the idr_pre_get() kernel-doc, there are some cases where you can call idr_pre_get() from within locked regions. Add a description for such cases. See also: http://lkml.org/lkml/2010/9/16/462 [akpm@linux-foundation.org: cleanups, grammatical fixes] Signed-off-by: Naohiro Aota Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/idr.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-) (limited to 'lib') diff --git a/lib/idr.c b/lib/idr.c index 5e0966be0f7c..e35850d3004a 100644 --- a/lib/idr.c +++ b/lib/idr.c @@ -110,9 +110,10 @@ static void idr_mark_full(struct idr_layer **pa, int id) * @idp: idr handle * @gfp_mask: memory allocation flags * - * This function should be called prior to locking and calling the - * idr_get_new* functions. It preallocates enough memory to satisfy - * the worst possible allocation. + * This function should be called prior to calling the idr_get_new* functions. + * It preallocates enough memory to satisfy the worst possible allocation. The + * caller should pass in GFP_KERNEL if possible. This of course requires that + * no spinning locks be held. * * If the system is REALLY out of memory this function returns 0, * otherwise 1. @@ -290,9 +291,11 @@ static int idr_get_new_above_int(struct idr *idp, void *ptr, int starting_id) * This is the allocate id function. It should be called with any * required locks. * - * If memory is required, it will return -EAGAIN, you should unlock - * and go back to the idr_pre_get() call. If the idr is full, it will - * return -ENOSPC. + * If allocation from IDR's private freelist fails, idr_get_new_above() will + * return -EAGAIN. The caller should retry the idr_pre_get() call to refill + * IDR's preallocation and then retry the idr_get_new_above() call. + * + * If the idr is full idr_get_new_above() will return -ENOSPC. * * @id returns a value in the range @starting_id ... 0x7fffffff */ @@ -318,12 +321,11 @@ EXPORT_SYMBOL(idr_get_new_above); * @ptr: pointer you want associated with the id * @id: pointer to the allocated handle * - * This is the allocate id function. It should be called with any - * required locks. + * If allocation from IDR's private freelist fails, idr_get_new_above() will + * return -EAGAIN. The caller should retry the idr_pre_get() call to refill + * IDR's preallocation and then retry the idr_get_new_above() call. * - * If memory is required, it will return -EAGAIN, you should unlock - * and go back to the idr_pre_get() call. If the idr is full, it will - * return -ENOSPC. + * If the idr is full idr_get_new_above() will return -ENOSPC. * * @id returns a value in the range 0 ... 0x7fffffff */ -- cgit v1.2.2 From e2852ae825dba5ebc159788720baec1a28a57125 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 26 Oct 2010 14:23:05 -0700 Subject: percpu_counter: add debugobj support All percpu counters are linked to a global list on initialization and removed from it on destruction. The list is walked during CPU up/down. If a percpu counter is freed without being properly destroyed, the system will oops only on the next CPU up/down making it pretty nasty to track down. This patch adds debugobj support for percpu counters so that such problems can be found easily. As percpu counters don't make sense on stack and can't be statically initialized, debugobj support is pretty simple. It's initialized and activated on counter initialization, and deactivatd and destroyed on counter destruction. With this patch applied, the bug fixed by commit 602586a83b719df0fbd94196a1359ed35aeb2df3 (shmem: put_super must percpu_counter_destroy) triggers the following warning on tmpfs unmount and the system won't oops on the next cpu up/down operation. ------------[ cut here ]------------ WARNING: at lib/debugobjects.c:259 debug_print_object+0x5c/0x70() Hardware name: Bochs ODEBUG: free active (active state 0) object type: percpu_counter Modules linked in: Pid: 3999, comm: umount Not tainted 2.6.36-rc2-work+ #5 Call Trace: [] warn_slowpath_common+0x7f/0xc0 [] warn_slowpath_fmt+0x46/0x50 [] debug_print_object+0x5c/0x70 [] debug_check_no_obj_freed+0x125/0x210 [] kfree+0xb3/0x2f0 [] shmem_put_super+0x1d/0x30 [] generic_shutdown_super+0x56/0xe0 [] kill_anon_super+0x16/0x60 [] kill_litter_super+0x27/0x30 [] deactivate_locked_super+0x45/0x60 [] deactivate_super+0x4a/0x70 [] mntput_no_expire+0x86/0xe0 [] sys_umount+0x6f/0x360 [] system_call_fastpath+0x16/0x1b ---[ end trace cce2a341ba3611a7 ]--- Signed-off-by: Tejun Heo Acked-by: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/Kconfig.debug | 8 ++++++++ lib/percpu_counter.c | 48 ++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 56 insertions(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 69a32664c289..0d5c762532a5 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -317,6 +317,14 @@ config DEBUG_OBJECTS_RCU_HEAD help Enable this to turn on debugging of RCU list heads (call_rcu() usage). +config DEBUG_OBJECTS_PERCPU_COUNTER + bool "Debug percpu counter objects" + depends on DEBUG_OBJECTS + help + If you say Y here, additional code will be inserted into the + percpu counter routines to track the life time of percpu counter + objects and validate the percpu counter operations. + config DEBUG_OBJECTS_ENABLE_DEFAULT int "debug_objects bootup default value (0-1)" range 0 1 diff --git a/lib/percpu_counter.c b/lib/percpu_counter.c index 209448e1d2b9..1d954ea72331 100644 --- a/lib/percpu_counter.c +++ b/lib/percpu_counter.c @@ -8,10 +8,53 @@ #include #include #include +#include static LIST_HEAD(percpu_counters); static DEFINE_MUTEX(percpu_counters_lock); +#ifdef CONFIG_DEBUG_OBJECTS_PERCPU_COUNTER + +static struct debug_obj_descr percpu_counter_debug_descr; + +static int percpu_counter_fixup_free(void *addr, enum debug_obj_state state) +{ + struct percpu_counter *fbc = addr; + + switch (state) { + case ODEBUG_STATE_ACTIVE: + percpu_counter_destroy(fbc); + debug_object_free(fbc, &percpu_counter_debug_descr); + return 1; + default: + return 0; + } +} + +static struct debug_obj_descr percpu_counter_debug_descr = { + .name = "percpu_counter", + .fixup_free = percpu_counter_fixup_free, +}; + +static inline void debug_percpu_counter_activate(struct percpu_counter *fbc) +{ + debug_object_init(fbc, &percpu_counter_debug_descr); + debug_object_activate(fbc, &percpu_counter_debug_descr); +} + +static inline void debug_percpu_counter_deactivate(struct percpu_counter *fbc) +{ + debug_object_deactivate(fbc, &percpu_counter_debug_descr); + debug_object_free(fbc, &percpu_counter_debug_descr); +} + +#else /* CONFIG_DEBUG_OBJECTS_PERCPU_COUNTER */ +static inline void debug_percpu_counter_activate(struct percpu_counter *fbc) +{ } +static inline void debug_percpu_counter_deactivate(struct percpu_counter *fbc) +{ } +#endif /* CONFIG_DEBUG_OBJECTS_PERCPU_COUNTER */ + void percpu_counter_set(struct percpu_counter *fbc, s64 amount) { int cpu; @@ -75,6 +118,9 @@ int __percpu_counter_init(struct percpu_counter *fbc, s64 amount, fbc->counters = alloc_percpu(s32); if (!fbc->counters) return -ENOMEM; + + debug_percpu_counter_activate(fbc); + #ifdef CONFIG_HOTPLUG_CPU INIT_LIST_HEAD(&fbc->list); mutex_lock(&percpu_counters_lock); @@ -90,6 +136,8 @@ void percpu_counter_destroy(struct percpu_counter *fbc) if (!fbc->counters) return; + debug_percpu_counter_deactivate(fbc); + #ifdef CONFIG_HOTPLUG_CPU mutex_lock(&percpu_counters_lock); list_del(&fbc->list); -- cgit v1.2.2 From 6d411e6c01608cefb7b9ea6712110538a1432f9f Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 26 Oct 2010 14:23:05 -0700 Subject: lib/Kconfig.debug: add list_sort debugging switch While hunting a non-existing bug in 'list_sort()', I've improved the 'list_sort_test()' function which tests the 'list_sort()' library call. Although at the end I found a bug in my code, but not in 'list_sort()', I think my clean-ups and improvements are worth merging because they make the test function better. This patch: Make the self-tests selectable via Kconfig rather than by manual enabling of DEBUG_LIST_SORT. Signed-off-by: Artem Bityutskiy Cc: Don Mullis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/Kconfig.debug | 9 +++++++++ lib/list_sort.c | 4 ++-- 2 files changed, 11 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 0d5c762532a5..95bda87a3e84 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -748,6 +748,15 @@ config DEBUG_LIST If unsure, say N. +config TEST_LIST_SORT + bool "Linked list sorting test" + depends on DEBUG_KERNEL + help + Enable this to turn on 'list_sort()' function test. This test is + executed only once during system boot, so affects only boot time. + + If unsure, say N. + config DEBUG_SG bool "Debug SG table operations" depends on DEBUG_KERNEL diff --git a/lib/list_sort.c b/lib/list_sort.c index a7616fa3162e..827794016bfb 100644 --- a/lib/list_sort.c +++ b/lib/list_sort.c @@ -141,7 +141,7 @@ void list_sort(void *priv, struct list_head *head, } EXPORT_SYMBOL(list_sort); -#ifdef DEBUG_LIST_SORT +#ifdef CONFIG_TEST_LIST_SORT struct debug_el { struct list_head l_h; int value; @@ -214,4 +214,4 @@ static int __init list_sort_test(void) return 0; } module_init(list_sort_test); -#endif +#endif /* CONFIG_TEST_LIST_SORT */ -- cgit v1.2.2 From bb2ab10fa693110cffa7087ffe2749d6e9a27d5f Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 26 Oct 2010 14:23:06 -0700 Subject: lib/list_sort: test: use more reasonable printk levels I do not see any reason to use KERN_WARN for normal messages and KERN_EMERG for error messages in the lib_sort testing routine. Let's use more reasonable KERN_NORM and KERN_ERR levels. Signed-off-by: Artem Bityutskiy Cc: Don Mullis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/list_sort.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) (limited to 'lib') diff --git a/lib/list_sort.c b/lib/list_sort.c index 827794016bfb..679b3a060e7e 100644 --- a/lib/list_sort.c +++ b/lib/list_sort.c @@ -166,7 +166,7 @@ static int __init list_sort_test(void) struct list_head *head = kmalloc(sizeof(*head), GFP_KERNEL); struct list_head *cur; - printk(KERN_WARNING "testing list_sort()\n"); + printk(KERN_DEBUG "testing list_sort()\n"); cur = head; for (i = 0; i < LIST_SORT_TEST_LENGTH; i++) { @@ -189,17 +189,17 @@ static int __init list_sort_test(void) struct debug_el *el = container_of(cur, struct debug_el, l_h); int cmp_result = cmp(NULL, cur, cur->next); if (cur->next->prev != cur) { - printk(KERN_EMERG "list_sort() returned " - "a corrupted list!\n"); + printk(KERN_ERR "list_sort() returned " + "a corrupted list!\n"); return 1; } else if (cmp_result > 0) { - printk(KERN_EMERG "list_sort() failed to sort!\n"); + printk(KERN_ERR "list_sort() failed to sort!\n"); return 1; } else if (cmp_result == 0 && el->serial >= container_of(cur->next, struct debug_el, l_h)->serial) { - printk(KERN_EMERG "list_sort() failed to preserve order" - " of equivalent elements!\n"); + printk(KERN_ERR "list_sort() failed to preserve order " + "of equivalent elements!\n"); return 1; } kfree(cur->prev); @@ -207,8 +207,8 @@ static int __init list_sort_test(void) } kfree(cur); if (count != LIST_SORT_TEST_LENGTH) { - printk(KERN_EMERG "list_sort() returned list of" - "different length!\n"); + printk(KERN_ERR "list_sort() returned list of " + "different length!\n"); return 1; } return 0; -- cgit v1.2.2 From eeee9ebb54b76a33a13d2c926ffb018a4aea410f Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 26 Oct 2010 14:23:06 -0700 Subject: lib/list_sort: test: use generic random32 Instead of using own pseudo-random generator, use generic linux 'random32()' function. Presumably, this should improve test coverage. At the same time, do the following changes: o Use shorter macro name for test list length o Do not use strange 'l_h' name for 'struct list_head' element, use 'list', because it is traditional name and thus, makes the code more obvious and readable. Signed-off-by: Artem Bityutskiy Cc: Don Mullis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/list_sort.c | 27 +++++++++++++++------------ 1 file changed, 15 insertions(+), 12 deletions(-) (limited to 'lib') diff --git a/lib/list_sort.c b/lib/list_sort.c index 679b3a060e7e..8f3c24415ae5 100644 --- a/lib/list_sort.c +++ b/lib/list_sort.c @@ -142,42 +142,45 @@ void list_sort(void *priv, struct list_head *head, EXPORT_SYMBOL(list_sort); #ifdef CONFIG_TEST_LIST_SORT + +#include + struct debug_el { - struct list_head l_h; + struct list_head list; int value; unsigned serial; }; static int cmp(void *priv, struct list_head *a, struct list_head *b) { - return container_of(a, struct debug_el, l_h)->value - - container_of(b, struct debug_el, l_h)->value; + return container_of(a, struct debug_el, list)->value + - container_of(b, struct debug_el, list)->value; } /* * The pattern of set bits in the list length determines which cases * are hit in list_sort(). */ -#define LIST_SORT_TEST_LENGTH (512+128+2) /* not including head */ +#define TEST_LIST_LEN (512+128+2) /* not including head */ static int __init list_sort_test(void) { - int i, r = 1, count; + int i, count; struct list_head *head = kmalloc(sizeof(*head), GFP_KERNEL); struct list_head *cur; printk(KERN_DEBUG "testing list_sort()\n"); cur = head; - for (i = 0; i < LIST_SORT_TEST_LENGTH; i++) { + for (i = 0; i < TEST_LIST_LEN; i++) { struct debug_el *el = kmalloc(sizeof(*el), GFP_KERNEL); BUG_ON(!el); /* force some equivalencies */ - el->value = (r = (r * 725861) % 6599) % (LIST_SORT_TEST_LENGTH/3); + el->value = random32() % (TEST_LIST_LEN/3); el->serial = i; - el->l_h.prev = cur; - cur->next = &el->l_h; + el->list.prev = cur; + cur->next = &el->list; cur = cur->next; } head->prev = cur; @@ -186,7 +189,7 @@ static int __init list_sort_test(void) count = 1; for (cur = head->next; cur->next != head; cur = cur->next) { - struct debug_el *el = container_of(cur, struct debug_el, l_h); + struct debug_el *el = container_of(cur, struct debug_el, list); int cmp_result = cmp(NULL, cur, cur->next); if (cur->next->prev != cur) { printk(KERN_ERR "list_sort() returned " @@ -197,7 +200,7 @@ static int __init list_sort_test(void) return 1; } else if (cmp_result == 0 && el->serial >= container_of(cur->next, - struct debug_el, l_h)->serial) { + struct debug_el, list)->serial) { printk(KERN_ERR "list_sort() failed to preserve order " "of equivalent elements!\n"); return 1; @@ -206,7 +209,7 @@ static int __init list_sort_test(void) count++; } kfree(cur); - if (count != LIST_SORT_TEST_LENGTH) { + if (count != TEST_LIST_LEN) { printk(KERN_ERR "list_sort() returned list of " "different length!\n"); return 1; -- cgit v1.2.2 From f3dc0e384248ea6fda0987f909007fa9ab5fb51a Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 26 Oct 2010 14:23:07 -0700 Subject: lib/list_sort: test: improve errors handling The 'lib_sort()' test does not free memory if it fails, and it makes the kernel panic if it cannot allocate memory. This patch fixes the problem. This patch also changes several small things: o use 'list_add()' helper instead of adding manually o introduce temporary 'el1' variable to avoid ugly and unreadalbe "if" statement o make 'head' to be stack variable instead of 'kmalloc()'ed, which simplifies code a bit Overall, this patch is of clean-up type. Signed-off-by: Artem Bityutskiy Cc: Don Mullis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/list_sort.c | 65 +++++++++++++++++++++++++++++++++------------------------ 1 file changed, 38 insertions(+), 27 deletions(-) (limited to 'lib') diff --git a/lib/list_sort.c b/lib/list_sort.c index 8f3c24415ae5..2b99ff80f4be 100644 --- a/lib/list_sort.c +++ b/lib/list_sort.c @@ -165,56 +165,67 @@ static int cmp(void *priv, struct list_head *a, struct list_head *b) static int __init list_sort_test(void) { - int i, count; - struct list_head *head = kmalloc(sizeof(*head), GFP_KERNEL); - struct list_head *cur; + int i, count = 1, err = -EINVAL; + struct debug_el *el; + struct list_head *cur, *tmp; + LIST_HEAD(head); printk(KERN_DEBUG "testing list_sort()\n"); - cur = head; for (i = 0; i < TEST_LIST_LEN; i++) { - struct debug_el *el = kmalloc(sizeof(*el), GFP_KERNEL); - BUG_ON(!el); + el = kmalloc(sizeof(*el), GFP_KERNEL); + if (!el) { + printk(KERN_ERR "cancel list_sort() testing - cannot " + "allocate memory\n"); + goto exit; + } /* force some equivalencies */ el->value = random32() % (TEST_LIST_LEN/3); el->serial = i; - - el->list.prev = cur; - cur->next = &el->list; - cur = cur->next; + list_add_tail(&el->list, &head); } - head->prev = cur; - list_sort(NULL, head, cmp); + list_sort(NULL, &head, cmp); + + for (cur = head.next; cur->next != &head; cur = cur->next) { + struct debug_el *el1; + int cmp_result; - count = 1; - for (cur = head->next; cur->next != head; cur = cur->next) { - struct debug_el *el = container_of(cur, struct debug_el, list); - int cmp_result = cmp(NULL, cur, cur->next); if (cur->next->prev != cur) { printk(KERN_ERR "list_sort() returned " "a corrupted list!\n"); - return 1; - } else if (cmp_result > 0) { + goto exit; + } + + cmp_result = cmp(NULL, cur, cur->next); + if (cmp_result > 0) { printk(KERN_ERR "list_sort() failed to sort!\n"); - return 1; - } else if (cmp_result == 0 && - el->serial >= container_of(cur->next, - struct debug_el, list)->serial) { + goto exit; + } + + el = container_of(cur, struct debug_el, list); + el1 = container_of(cur->next, struct debug_el, list); + if (cmp_result == 0 && el->serial >= el1->serial) { printk(KERN_ERR "list_sort() failed to preserve order " "of equivalent elements!\n"); - return 1; + goto exit; } - kfree(cur->prev); count++; } - kfree(cur); + if (count != TEST_LIST_LEN) { printk(KERN_ERR "list_sort() returned list of " "different length!\n"); - return 1; + goto exit; + } + + err = 0; +exit: + list_for_each_safe(cur, tmp, &head) { + list_del(cur); + kfree(container_of(cur, struct debug_el, list)); } - return 0; + return err; } module_init(list_sort_test); #endif /* CONFIG_TEST_LIST_SORT */ -- cgit v1.2.2 From 014afa943d44f0df8e65bc4bd071c67772277d93 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 26 Oct 2010 14:23:08 -0700 Subject: lib/list_sort: test: unify test messages This patch unifies 'list_sort_test()' messages a bit and makes sure all of them start with the "list_sort_test:" prefix to make it obvious for users where the messages come from. Signed-off-by: Artem Bityutskiy Cc: Don Mullis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/list_sort.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) (limited to 'lib') diff --git a/lib/list_sort.c b/lib/list_sort.c index 2b99ff80f4be..01aff9e80821 100644 --- a/lib/list_sort.c +++ b/lib/list_sort.c @@ -170,12 +170,12 @@ static int __init list_sort_test(void) struct list_head *cur, *tmp; LIST_HEAD(head); - printk(KERN_DEBUG "testing list_sort()\n"); + printk(KERN_DEBUG "list_sort_test: start testing list_sort()\n"); for (i = 0; i < TEST_LIST_LEN; i++) { el = kmalloc(sizeof(*el), GFP_KERNEL); if (!el) { - printk(KERN_ERR "cancel list_sort() testing - cannot " + printk(KERN_ERR "list_sort_test: error: cannot " "allocate memory\n"); goto exit; } @@ -192,30 +192,31 @@ static int __init list_sort_test(void) int cmp_result; if (cur->next->prev != cur) { - printk(KERN_ERR "list_sort() returned " - "a corrupted list!\n"); + printk(KERN_ERR "list_sort_test: error: list is " + "corrupted\n"); goto exit; } cmp_result = cmp(NULL, cur, cur->next); if (cmp_result > 0) { - printk(KERN_ERR "list_sort() failed to sort!\n"); + printk(KERN_ERR "list_sort_test: error: list is not " + "sorted\n"); goto exit; } el = container_of(cur, struct debug_el, list); el1 = container_of(cur->next, struct debug_el, list); if (cmp_result == 0 && el->serial >= el1->serial) { - printk(KERN_ERR "list_sort() failed to preserve order " - "of equivalent elements!\n"); + printk(KERN_ERR "list_sort_test: error: order of " + "equivalent elements not preserved\n"); goto exit; } count++; } if (count != TEST_LIST_LEN) { - printk(KERN_ERR "list_sort() returned list of " - "different length!\n"); + printk(KERN_ERR "list_sort_test: error: bad list length %d", + count); goto exit; } -- cgit v1.2.2 From 041b78f232bb87b2de8ca3fed50384bc7dc9c2de Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 26 Oct 2010 14:23:08 -0700 Subject: lib/list_sort: test: check element addresses Improve 'lib_sort()' test and check that: o 'cmp()' is called only for elements which were present in the original list, i.e., the 'a' and 'b' parameters are valid o the resulted (sorted) list consists onlly of the original elements o intdoruce "poison" fields to make sure data around 'struc list_head' field are not corrupted. Signed-off-by: Artem Bityutskiy Cc: Don Mullis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/list_sort.c | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++------ 1 file changed, 67 insertions(+), 8 deletions(-) (limited to 'lib') diff --git a/lib/list_sort.c b/lib/list_sort.c index 01aff9e80821..d7325c6b103f 100644 --- a/lib/list_sort.c +++ b/lib/list_sort.c @@ -145,23 +145,65 @@ EXPORT_SYMBOL(list_sort); #include +/* + * The pattern of set bits in the list length determines which cases + * are hit in list_sort(). + */ +#define TEST_LIST_LEN (512+128+2) /* not including head */ + +#define TEST_POISON1 0xDEADBEEF +#define TEST_POISON2 0xA324354C + struct debug_el { + unsigned int poison1; struct list_head list; + unsigned int poison2; int value; unsigned serial; }; -static int cmp(void *priv, struct list_head *a, struct list_head *b) +/* Array, containing pointers to all elements in the test list */ +static struct debug_el **elts __initdata; + +static int __init check(struct debug_el *ela, struct debug_el *elb) { - return container_of(a, struct debug_el, list)->value - - container_of(b, struct debug_el, list)->value; + if (ela->serial >= TEST_LIST_LEN) { + printk(KERN_ERR "list_sort_test: error: incorrect serial %d\n", + ela->serial); + return -EINVAL; + } + if (elb->serial >= TEST_LIST_LEN) { + printk(KERN_ERR "list_sort_test: error: incorrect serial %d\n", + elb->serial); + return -EINVAL; + } + if (elts[ela->serial] != ela || elts[elb->serial] != elb) { + printk(KERN_ERR "list_sort_test: error: phantom element\n"); + return -EINVAL; + } + if (ela->poison1 != TEST_POISON1 || ela->poison2 != TEST_POISON2) { + printk(KERN_ERR "list_sort_test: error: bad poison: %#x/%#x\n", + ela->poison1, ela->poison2); + return -EINVAL; + } + if (elb->poison1 != TEST_POISON1 || elb->poison2 != TEST_POISON2) { + printk(KERN_ERR "list_sort_test: error: bad poison: %#x/%#x\n", + elb->poison1, elb->poison2); + return -EINVAL; + } + return 0; } -/* - * The pattern of set bits in the list length determines which cases - * are hit in list_sort(). - */ -#define TEST_LIST_LEN (512+128+2) /* not including head */ +static int __init cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct debug_el *ela, *elb; + + ela = container_of(a, struct debug_el, list); + elb = container_of(b, struct debug_el, list); + + check(ela, elb); + return ela->value - elb->value; +} static int __init list_sort_test(void) { @@ -172,6 +214,13 @@ static int __init list_sort_test(void) printk(KERN_DEBUG "list_sort_test: start testing list_sort()\n"); + elts = kmalloc(sizeof(void *) * TEST_LIST_LEN, GFP_KERNEL); + if (!elts) { + printk(KERN_ERR "list_sort_test: error: cannot allocate " + "memory\n"); + goto exit; + } + for (i = 0; i < TEST_LIST_LEN; i++) { el = kmalloc(sizeof(*el), GFP_KERNEL); if (!el) { @@ -182,6 +231,9 @@ static int __init list_sort_test(void) /* force some equivalencies */ el->value = random32() % (TEST_LIST_LEN/3); el->serial = i; + el->poison1 = TEST_POISON1; + el->poison2 = TEST_POISON2; + elts[i] = el; list_add_tail(&el->list, &head); } @@ -211,6 +263,12 @@ static int __init list_sort_test(void) "equivalent elements not preserved\n"); goto exit; } + + if (check(el, el1)) { + printk(KERN_ERR "list_sort_test: error: element check " + "failed\n"); + goto exit; + } count++; } @@ -222,6 +280,7 @@ static int __init list_sort_test(void) err = 0; exit: + kfree(elts); list_for_each_safe(cur, tmp, &head) { list_del(cur); kfree(container_of(cur, struct debug_el, list)); -- cgit v1.2.2 From ea00c30b5b31baa91be29bee966204eccc15e9d3 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Tue, 26 Oct 2010 14:23:09 -0700 Subject: percpu_counter: use this_cpu_ptr() instead of per_cpu_ptr() this_cpu_ptr() avoids an array lookup and can use the percpu offset of the local cpu directly. Signed-off-by: Christoph Lameter Cc: Eric Dumazet Cc: Tejun Heo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/percpu_counter.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/percpu_counter.c b/lib/percpu_counter.c index 1d954ea72331..604678d7d06d 100644 --- a/lib/percpu_counter.c +++ b/lib/percpu_counter.c @@ -73,9 +73,9 @@ void __percpu_counter_add(struct percpu_counter *fbc, s64 amount, s32 batch) { s64 count; s32 *pcount; - int cpu = get_cpu(); - pcount = per_cpu_ptr(fbc->counters, cpu); + preempt_disable(); + pcount = this_cpu_ptr(fbc->counters); count = *pcount + amount; if (count >= batch || count <= -batch) { spin_lock(&fbc->lock); @@ -85,7 +85,7 @@ void __percpu_counter_add(struct percpu_counter *fbc, s64 amount, s32 batch) } else { *pcount = count; } - put_cpu(); + preempt_enable(); } EXPORT_SYMBOL(__percpu_counter_add); -- cgit v1.2.2 From 5d051decfc27cdf33fbbd2bfca958d0d2c903569 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Tue, 26 Oct 2010 14:23:09 -0700 Subject: lib/parser: cleanup match_number() Use new variable 'len' to make code more readable. Signed-off-by: Namhyung Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/parser.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/parser.c b/lib/parser.c index fb34977246bb..6e89eca5cca0 100644 --- a/lib/parser.c +++ b/lib/parser.c @@ -128,12 +128,13 @@ static int match_number(substring_t *s, int *result, int base) char *endp; char *buf; int ret; + size_t len = s->to - s->from; - buf = kmalloc(s->to - s->from + 1, GFP_KERNEL); + buf = kmalloc(len + 1, GFP_KERNEL); if (!buf) return -ENOMEM; - memcpy(buf, s->from, s->to - s->from); - buf[s->to - s->from] = '\0'; + memcpy(buf, s->from, len); + buf[len] = '\0'; *result = simple_strtol(buf, &endp, base); ret = 0; if (endp == buf) -- cgit v1.2.2 From 658716d19f8f155c67d4677ba68034b8e492dfbe Mon Sep 17 00:00:00 2001 From: Brian Behlendorf Date: Tue, 26 Oct 2010 14:23:10 -0700 Subject: div64_u64(): improve precision on 32bit platforms The current implementation of div64_u64 for 32bit systems returns an approximately correct result when the divisor exceeds 32bits. Since doing 64bit division using 32bit hardware is a long since solved problem we just use one of the existing proven methods. Additionally, add a div64_s64 function to correctly handle doing signed 64bit division. Addresses https://bugzilla.redhat.com/show_bug.cgi?id=616105 Signed-off-by: Brian Behlendorf Signed-off-by: Oleg Nesterov Cc: Ben Woodard Cc: Jeremy Fitzhardinge Cc: Mark Grondona Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/div64.c | 52 ++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 42 insertions(+), 10 deletions(-) (limited to 'lib') diff --git a/lib/div64.c b/lib/div64.c index a111eb8de9cf..5b4919191778 100644 --- a/lib/div64.c +++ b/lib/div64.c @@ -77,26 +77,58 @@ s64 div_s64_rem(s64 dividend, s32 divisor, s32 *remainder) EXPORT_SYMBOL(div_s64_rem); #endif -/* 64bit divisor, dividend and result. dynamic precision */ +/** + * div64_u64 - unsigned 64bit divide with 64bit divisor + * @dividend: 64bit dividend + * @divisor: 64bit divisor + * + * This implementation is a modified version of the algorithm proposed + * by the book 'Hacker's Delight'. The original source and full proof + * can be found here and is available for use without restriction. + * + * 'http://www.hackersdelight.org/HDcode/newCode/divDouble.c' + */ #ifndef div64_u64 u64 div64_u64(u64 dividend, u64 divisor) { - u32 high, d; + u32 high = divisor >> 32; + u64 quot; - high = divisor >> 32; - if (high) { - unsigned int shift = fls(high); + if (high == 0) { + quot = div_u64(dividend, divisor); + } else { + int n = 1 + fls(high); + quot = div_u64(dividend >> n, divisor >> n); - d = divisor >> shift; - dividend >>= shift; - } else - d = divisor; + if (quot != 0) + quot--; + if ((dividend - quot * divisor) >= divisor) + quot++; + } - return div_u64(dividend, d); + return quot; } EXPORT_SYMBOL(div64_u64); #endif +/** + * div64_s64 - signed 64bit divide with 64bit divisor + * @dividend: 64bit dividend + * @divisor: 64bit divisor + */ +#ifndef div64_s64 +s64 div64_s64(s64 dividend, s64 divisor) +{ + s64 quot, t; + + quot = div64_u64(abs64(dividend), abs64(divisor)); + t = (dividend ^ divisor) >> 63; + + return (quot ^ t) - t; +} +EXPORT_SYMBOL(div64_s64); +#endif + #endif /* BITS_PER_LONG == 32 */ /* -- cgit v1.2.2 From 56083ab17e0075e538270823c374b59cc97e73b9 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 26 Oct 2010 14:19:08 -0700 Subject: docbook: add idr/ida to kernel-api docbook Add idr/ida to kernel-api docbook. Fix typos and kernel-doc notation. Signed-off-by: Randy Dunlap Acked-by: Tejun Heo Cc: Naohiro Aota Cc: Jiri Kosina Signed-off-by: Linus Torvalds --- lib/idr.c | 49 +++++++++++++++++++++++++------------------------ 1 file changed, 25 insertions(+), 24 deletions(-) (limited to 'lib') diff --git a/lib/idr.c b/lib/idr.c index e35850d3004a..e15502e8b21e 100644 --- a/lib/idr.c +++ b/lib/idr.c @@ -106,7 +106,7 @@ static void idr_mark_full(struct idr_layer **pa, int id) } /** - * idr_pre_get - reserver resources for idr allocation + * idr_pre_get - reserve resources for idr allocation * @idp: idr handle * @gfp_mask: memory allocation flags * @@ -115,8 +115,8 @@ static void idr_mark_full(struct idr_layer **pa, int id) * caller should pass in GFP_KERNEL if possible. This of course requires that * no spinning locks be held. * - * If the system is REALLY out of memory this function returns 0, - * otherwise 1. + * If the system is REALLY out of memory this function returns %0, + * otherwise %1. */ int idr_pre_get(struct idr *idp, gfp_t gfp_mask) { @@ -292,12 +292,12 @@ static int idr_get_new_above_int(struct idr *idp, void *ptr, int starting_id) * required locks. * * If allocation from IDR's private freelist fails, idr_get_new_above() will - * return -EAGAIN. The caller should retry the idr_pre_get() call to refill + * return %-EAGAIN. The caller should retry the idr_pre_get() call to refill * IDR's preallocation and then retry the idr_get_new_above() call. * - * If the idr is full idr_get_new_above() will return -ENOSPC. + * If the idr is full idr_get_new_above() will return %-ENOSPC. * - * @id returns a value in the range @starting_id ... 0x7fffffff + * @id returns a value in the range @starting_id ... %0x7fffffff */ int idr_get_new_above(struct idr *idp, void *ptr, int starting_id, int *id) { @@ -322,12 +322,12 @@ EXPORT_SYMBOL(idr_get_new_above); * @id: pointer to the allocated handle * * If allocation from IDR's private freelist fails, idr_get_new_above() will - * return -EAGAIN. The caller should retry the idr_pre_get() call to refill + * return %-EAGAIN. The caller should retry the idr_pre_get() call to refill * IDR's preallocation and then retry the idr_get_new_above() call. * - * If the idr is full idr_get_new_above() will return -ENOSPC. + * If the idr is full idr_get_new_above() will return %-ENOSPC. * - * @id returns a value in the range 0 ... 0x7fffffff + * @id returns a value in the range %0 ... %0x7fffffff */ int idr_get_new(struct idr *idp, void *ptr, int *id) { @@ -390,7 +390,7 @@ static void sub_remove(struct idr *idp, int shift, int id) } /** - * idr_remove - remove the given id and free it's slot + * idr_remove - remove the given id and free its slot * @idp: idr handle * @id: unique key */ @@ -439,7 +439,7 @@ EXPORT_SYMBOL(idr_remove); * function will remove all id mappings and leave all idp_layers * unused. * - * A typical clean-up sequence for objects stored in an idr tree, will + * A typical clean-up sequence for objects stored in an idr tree will * use idr_for_each() to free all objects, if necessay, then * idr_remove_all() to remove all ids, and idr_destroy() to free * up the cached idr_layers. @@ -544,7 +544,7 @@ EXPORT_SYMBOL(idr_find); * not allowed. * * We check the return of @fn each time. If it returns anything other - * than 0, we break out and return that value. + * than %0, we break out and return that value. * * The caller must serialize idr_for_each() vs idr_get_new() and idr_remove(). */ @@ -639,8 +639,8 @@ EXPORT_SYMBOL(idr_get_next); * @id: lookup key * * Replace the pointer registered with an id and return the old value. - * A -ENOENT return indicates that @id was not found. - * A -EINVAL return indicates that @id was not within valid constraints. + * A %-ENOENT return indicates that @id was not found. + * A %-EINVAL return indicates that @id was not within valid constraints. * * The caller must serialize with writers. */ @@ -698,10 +698,11 @@ void idr_init(struct idr *idp) EXPORT_SYMBOL(idr_init); -/* +/** + * DOC: IDA description * IDA - IDR based ID allocator * - * this is id allocator without id -> pointer translation. Memory + * This is id allocator without id -> pointer translation. Memory * usage is much lower than full blown idr because each id only * occupies a bit. ida uses a custom leaf node which contains * IDA_BITMAP_BITS slots. @@ -734,8 +735,8 @@ static void free_bitmap(struct ida *ida, struct ida_bitmap *bitmap) * following function. It preallocates enough memory to satisfy the * worst possible allocation. * - * If the system is REALLY out of memory this function returns 0, - * otherwise 1. + * If the system is REALLY out of memory this function returns %0, + * otherwise %1. */ int ida_pre_get(struct ida *ida, gfp_t gfp_mask) { @@ -767,11 +768,11 @@ EXPORT_SYMBOL(ida_pre_get); * Allocate new ID above or equal to @ida. It should be called with * any required locks. * - * If memory is required, it will return -EAGAIN, you should unlock + * If memory is required, it will return %-EAGAIN, you should unlock * and go back to the ida_pre_get() call. If the ida is full, it will - * return -ENOSPC. + * return %-ENOSPC. * - * @p_id returns a value in the range @starting_id ... 0x7fffffff. + * @p_id returns a value in the range @starting_id ... %0x7fffffff. */ int ida_get_new_above(struct ida *ida, int starting_id, int *p_id) { @@ -853,11 +854,11 @@ EXPORT_SYMBOL(ida_get_new_above); * * Allocate new ID. It should be called with any required locks. * - * If memory is required, it will return -EAGAIN, you should unlock + * If memory is required, it will return %-EAGAIN, you should unlock * and go back to the idr_pre_get() call. If the idr is full, it will - * return -ENOSPC. + * return %-ENOSPC. * - * @id returns a value in the range 0 ... 0x7fffffff. + * @id returns a value in the range %0 ... %0x7fffffff. */ int ida_get_new(struct ida *ida, int *p_id) { -- cgit v1.2.2