summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* init/main.c: obsolete_checksetup can be booleanYaowei Bai2016-01-201-5/+5
| | | | | | | | | | | Make obsolete_checksetup() return bool due to this particular function only using either one or zero as its return value. No functional change. Signed-off-by: Yaowei Bai <baiyaowei@cmss.chinamobile.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* epoll: add EPOLLEXCLUSIVE flagJason Baron2016-01-202-3/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently, epoll file descriptors or epfds (the fd returned from epoll_create[1]()) that are added to a shared wakeup source are always added in a non-exclusive manner. This means that when we have multiple epfds attached to a shared fd source they are all woken up. This creates thundering herd type behavior. Introduce a new 'EPOLLEXCLUSIVE' flag that can be passed as part of the 'event' argument during an epoll_ctl() EPOLL_CTL_ADD operation. This new flag allows for exclusive wakeups when there are multiple epfds attached to a shared fd event source. The implementation walks the list of exclusive waiters, and queues an event to each epfd, until it finds the first waiter that has threads blocked on it via epoll_wait(). The idea is to search for threads which are idle and ready to process the wakeup events. Thus, we queue an event to at least 1 epfd, but may still potentially queue an event to all epfds that are attached to the shared fd source. Performance testing was done by Madars Vitolins using a modified version of Enduro/X. The use of the 'EPOLLEXCLUSIVE' flag reduce the length of this particular workload from 860s down to 24s. Sample epoll_clt text: EPOLLEXCLUSIVE Sets an exclusive wakeup mode for the epfd file descriptor that is being attached to the target file descriptor, fd. Thus, when an event occurs and multiple epfd file descriptors are attached to the same target file using EPOLLEXCLUSIVE, one or more epfds will receive an event with epoll_wait(2). The default in this scenario (when EPOLLEXCLUSIVE is not set) is for all epfds to receive an event. EPOLLEXCLUSIVE may only be specified with the op EPOLL_CTL_ADD. Signed-off-by: Jason Baron <jbaron@akamai.com> Tested-by: Madars Vitolins <m@silodev.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Al Viro <viro@ftp.linux.org.uk> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Eric Wong <normalperson@yhbt.net> Cc: Jonathan Corbet <corbet@lwn.net> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Hagen Paul Pfeifer <hagen@jauu.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* checkpatch: fix a number of COMPLEX_MACRO false positivesVladimir Zapolskiy2016-01-201-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | A simple search over the kernel souce displays a number of correctly defined multiline macro, which generally are used as an array element initializer: % find ../linux -type f | xargs grep -B1 -H '^[:space]*\[.*\\$' However checkpatch.pl unexpectedly complains about all these macro definitions: % ./scripts/checkpatch.pl --types COMPLEX_MACRO -f include/linux/perf/arm_pmu.h ERROR: Macros with complex values should be enclosed in parentheses +#define PERF_MAP_ALL_UNSUPPORTED \ + [0 ... PERF_COUNT_HW_MAX - 1] = HW_OP_UNSUPPORTED The change intends to fix this type of false positives by flattening only array members and skipping array element designators. Signed-off-by: Vladimir Zapolskiy <vz@mleia.com> Acked-by: Joe Perches <joe@perches.com> Cc: Andy Whitcroft <apw@canonical.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* checkpatch: improve macros with flow control testJoe Perches2016-01-201-1/+1
| | | | | | | | | | | | | | | | The current test excludes any macro with ## concatenation from being reported with hidden flow control. Some macros are used with return or goto statements along with ##args or ##__VA_ARGS__. A somewhat common case is a logging macro like pr_info(fmt, ...) then a return or goto statement. Check the concatenated variable for args or __VA_ARGS__ and allow those macros to also be reported when they contain a return or goto. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* checkpatch: warn when casting constants to c90 int or longer typesJoe Perches2016-01-201-0/+42
| | | | | | | | | | | | | | | | | | | | | | Linus Torvalds wrote: > I can't but help to react that this: > #define IOMMU_ERROR_CODE (~(unsigned long) 0) > Not that this *matters*, but it's a bit odd to have to cast constants > to perfectly regular C types. So add a test that looks for constants that are cast to standard C90 int or longer types and suggest using C90 "6.4.4.1 Integer constants" integer-suffixes instead. Miscellanea: o Add a --fix option too Signed-off-by: Joe Perches <joe@perches.com> Suggested-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* lib/clz_tab.c: put in lib-y rather than obj-yChris Metcalf2016-01-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | The clz table (__clz_tab) in lib/clz_tab.c is also provided as part of libgcc.a, and many architectures link against libgcc. To allow the linker to avoid a multiple-definition link failure, clz_tab.o has to be in lib/lib.a rather than lib/builtin.o. The specific issue is that libgcc.a comes before lib/builtin.o on vmlinux.o's link command line, so its _clz.o is pulled to satisfy __clz_tab, and then when the remainder of lib/builtin.o is pulled in to satisfy all the other dependencies, the __clz_tab symbols conflict. By putting clz_tab.o in lib.a, the linker can simply avoid pulling it into vmlinux.o when this situation arises. The definitions of __clz_tab are the same in libgcc.a and in the kernel; arguably we could also simply rename the kernel version, but it's unlikely the libgcc version will ever change to become incompatible, so just using it seems reasonably safe. Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* include/linux/radix-tree.h: fix error in docs about locksAdam Barth2016-01-201-1/+1
| | | | | | | | | | | | This text refers to the "first 7 functions", which was correct when written but became incorrect when Johannes Weiner added another function to the list in 139e561660fe ("lib: radix_tree: tree node interface"). Change the text to correctly refer to the first 8 functions. Signed-off-by: Adam Barth <aurorean@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: print statistics at the endAndy Shevchenko2016-01-201-3/+23
| | | | | | | | | | Like others test are doing print the gathered statistics after test module is finished. Return from the module based on the result. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: test all possible group sizes for overflowAndy Shevchenko2016-01-201-3/+18
| | | | | | | | | | | Currently the only one combination is tested for overflow, i.e. rowsize = 16, groupsize = 1, len = 1. Do various test to go through all possible branches. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: check all bytes in real bufferAndy Shevchenko2016-01-201-21/+17
| | | | | | | | | | | | | | | | | | | | After processing by hex_dump_to_buffer() check all the parts to be expected. Part 1. The actual expected hex dump with or without ASCII part. Part 2. Check if the buffer is dirty beyond needed. Part 3. Return code should be as expected. This is done by using comparison of the return code and memcmp() against the test buffer. We fill the buffer by FILL_CHAR ('#') characters, so, we expect to have a tail of the buffer will be left untouched. The terminating NUL is also checked by memcmp(). Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: switch to memcmp()Andy Shevchenko2016-01-201-3/+3
| | | | | | | | | Better to use memcmp() against entire buffer to check that nothing is happened to the data in the tail. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: replace magic numbers by their meaningAndy Shevchenko2016-01-201-4/+13
| | | | | | | | | | | | | The magic numbers of the length are converted to their actual meaning, such as end of the buffer with and without ASCII part. We don't touch the rest of the magic constants that will be removed in the following commits. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: go through all possible lengths of bufferAndy Shevchenko2016-01-201-14/+13
| | | | | | | | | | When test for overflow do iterate the buffer length in a range 0 .. BUF_SIZE. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: define FILL_CHAR constantAndy Shevchenko2016-01-201-6/+13
| | | | | | | | | | | | | Define a character to fill the test buffers. Though the character should be printable since it's used when errors are reported. It should neither be from hex digit [a-fA-F0-9] dictionary nor space. It is recommended not to use one which is present in ASCII part of the test data. Later on we might switch to unprintable character to make test case more robust. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Suggested-by: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: introduce test_hexdump_prepare_test() helperAndy Shevchenko2016-01-201-7/+19
| | | | | | | | | The function prepares the expected result in the provided buffer. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* test_hexdump: rename to test_hexdumpAndy Shevchenko2016-01-202-1/+1
| | | | | | | | | | | | | | | | | The test suite currently doesn't cover many corner cases when hex_dump_to_buffer() runs into overflow. Refactor and amend test suite to cover most of the cases. This patch (of 9): Just to follow the scheme that most of the test modules are using. There is no fuctional change. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* drivers/firmware/broadcom/bcm47xx_nvram.c: use __ioread32_copy() instead of ↵Stephen Boyd2016-01-201-8/+3
| | | | | | | | | | | | | | | | open-coding Now that we have a generic library function for this, replace the open-coded instance. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Cc: Hauke Mehrtens <hauke@hauke-m.de> Cc: <zajec5@gmail.com> Cc: Paul Walmsley <paul@pwsan.com> Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com> Cc: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* drivers/soc/qcom/smd.c: use __ioread32_copy() instead of open-coding itStephen Boyd2016-01-201-9/+4
| | | | | | | | | | | | | | Now that we have a generic library function for this, replace the open-coded instance. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Reviewed-by: Bjorn Andersson <bjorn.andersson@sonymobile.com> Cc: <zajec5@gmail.com> Cc: David Howells <dhowells@redhat.com> Cc: Hauke Mehrtens <hauke@hauke-m.de> Cc: Paul Walmsley <paul@pwsan.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* lib/iomap_copy.c: add __ioread32_copy()Stephen Boyd2016-01-202-0/+22
| | | | | | | | | | | | | | Some drivers need to read data out of iomem areas 32-bits at a time. Add an API to do this. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com> Cc: <zajec5@gmail.com> Cc: David Howells <dhowells@redhat.com> Cc: Hauke Mehrtens <hauke@hauke-m.de> Cc: Paul Walmsley <paul@pwsan.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* arch/frv/include/asm/io.h: accept const void pointers for read{b,w,l}()Stephen Boyd2016-01-201-3/+14
| | | | | | | | | | | | | | | | | | | | | | | | The SMD driver is reading and writing chunks of data to iomem, and there's an __iowrite32_copy() function for the writing part, but no __ioread32_copy() function for the reading part. This series adds __ioread32_copy() and uses it in two places. This patch (of 4): The frv port uses compiler builtins, __builtin_read*(), for the I/O read routines. Unfortunately, these don't accept const void pointers although the generic ASM implementations do, so generic code passing const pointers to these APIs cause compilers to emit warnings. Add wrapper functions that cast away the const to avoid the warnings. Signed-off-by: Stephen Boyd <sboyd@codeaurora.org> Cc: David Howells <dhowells@redhat.com> Cc: <zajec5@gmail.com> Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com> Cc: Hauke Mehrtens <hauke@hauke-m.de> Cc: Paul Walmsley <paul@pwsan.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* string_helpers: fix precision loss for some inputsJames Bottomley2016-01-201-20/+43
| | | | | | | | | | | | | | | | | | | | | | | It was noticed that we lose precision in the final calculation for some inputs. The most egregious example is size=3000 blk_size=1900 in units of 10 should yield 5.70 MB but in fact yields 3.00 MB (oops). This is because the current algorithm doesn't correctly account for all the remainders in the logarithms. Fix this by doing a correct calculation in the remainders based on napier's algorithm. Additionally, now we have the correct result, we have to account for arithmetic rounding because we're printing 3 digits of precision. This means that if the fourth digit is five or greater, we have to round up, so add a section to ensure correct rounding. Finally account for all possible inputs correctly, including zero for block size. Fixes: b9f28d863594c429e1df35a0474d2663ca28b307 Signed-off-by: James Bottomley <JBottomley@Odin.com> Reported-by: Vitaly Kuznetsov <vkuznets@redhat.com> Cc: <stable@vger.kernel.org> [delay until after 4.4 release] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* ./CREDITS: add credit information for Martin KepplingerMartin Kepplinger2016-01-201-0/+10
| | | | | | Signed-off-by: Martin Kepplinger <martink@posteo.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* scripts/get_maintainer.pl: handle file names beginning with ./Joe Perches2016-01-201-0/+4
| | | | | | | | | | | | | | | The problem is that get_maintainer.pl doesn't work if you have a ./ prefix on the filename. For example, if you type: ./scripts/get_maintainer.pl -f ./drivers/usb/usb-skeleton.c then the current code only includes LKML and people from the git log, it doesn't include Greg or the linux-usb list. Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Cc: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* misc: ibmasm: fix build errorsRandy Dunlap2016-01-201-0/+1
| | | | | | | | | | | | | | | | | Fix build when CONFIG_SERIAL_8250=m and CONFIG_IBM_ASM=y. Fixes these build errors: drivers/built-in.o: In function `ibmasm_remove_one': module.c:(.text+0xf6874): undefined reference to `ibmasm_unregister_uart' drivers/built-in.o: In function `ibmasm_init_one': module.c:(.text+0xf6c37): undefined reference to `ibmasm_register_uart' Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Max Asbock <masbock@us.ibm.com> Cc: Vernon Mauery <vernux@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* zsmalloc: fix migrate_zspage-zs_free race conditionJunil Lee2016-01-201-1/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | record_obj() in migrate_zspage() does not preserve handle's HANDLE_PIN_BIT, set by find_aloced_obj()->trypin_tag(), and implicitly (accidentally) un-pins the handle, while migrate_zspage() still performs an explicit unpin_tag() on the that handle. This additional explicit unpin_tag() introduces a race condition with zs_free(), which can pin that handle by this time, so the handle becomes un-pinned. Schematically, it goes like this: CPU0 CPU1 migrate_zspage find_alloced_obj trypin_tag set HANDLE_PIN_BIT zs_free() pin_tag() obj_malloc() -- new object, no tag record_obj() -- remove HANDLE_PIN_BIT set HANDLE_PIN_BIT unpin_tag() -- remove zs_free's HANDLE_PIN_BIT The race condition may result in a NULL pointer dereference: Unable to handle kernel NULL pointer dereference at virtual address 00000000 CPU: 0 PID: 19001 Comm: CookieMonsterCl Tainted: PC is at get_zspage_mapping+0x0/0x24 LR is at obj_free.isra.22+0x64/0x128 Call trace: get_zspage_mapping+0x0/0x24 zs_free+0x88/0x114 zram_free_page+0x64/0xcc zram_slot_free_notify+0x90/0x108 swap_entry_free+0x278/0x294 free_swap_and_cache+0x38/0x11c unmap_single_vma+0x480/0x5c8 unmap_vmas+0x44/0x60 exit_mmap+0x50/0x110 mmput+0x58/0xe0 do_exit+0x320/0x8dc do_group_exit+0x44/0xa8 get_signal+0x538/0x580 do_signal+0x98/0x4b8 do_notify_resume+0x14/0x5c This patch keeps the lock bit in migration path and update value atomically. Signed-off-by: Junil Lee <junil0814.lee@lge.com> Signed-off-by: Minchan Kim <minchan@kernel.org> Acked-by: Vlastimil Babka <vbabka@suse.cz> Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com> Cc: <stable@vger.kernel.org> [4.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* mm: arch: remove duplicate definitions of MADV_FREEGuenter Roeck2016-01-204-4/+0
| | | | | | | | | | | | | | | | | | | | | Commits 21f55b018ba5 ("arch/*/include/uapi/asm/mman.h: : let MADV_FREE have same value for all architectures") and ef58978f1eaa ("mm: define MADV_FREE for some arches") both defined MADV_FREE, but did not use the same values. This results in build errors such as ./arch/alpha/include/uapi/asm/mman.h:53:0: error: "MADV_FREE" redefined ./arch/alpha/include/uapi/asm/mman.h:50:0: note: this is the location of the previous definition for the affected architectures. Fixes: 21f55b018ba5 ("arch/*/include/uapi/asm/mman.h: : let MADV_FREE have same value for all architectures") Fixes: ef58978f1eaa ("mm: define MADV_FREE for some arches") Signed-off-by: Guenter Roeck <linux@roeck-us.net> Cc: Chen Gang <gang.chen.5i5j@gmail.com> Cc: Minchan Kim <minchan@kernel.org> Acked-by: Helge Deller <deller@gmx.de> [parisc] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* fs/proc/task_mmu.c: add workaround for old compilersKirill A. Shutemov2016-01-201-1/+1
| | | | | | | | | | | | | | | | | | For THP=n, HPAGE_PMD_NR in smaps_account() expands to BUILD_BUG(). That's fine since this codepath is eliminated by modern compilers. But older compilers have not that efficient dead code elimination. It causes problem at least with gcc 4.1.2 on m68k: fs/built-in.o: In function `smaps_account': task_mmu.c:(.text+0x4f8fa): undefined reference to `__compiletime_assert_471' Let's replace HPAGE_PMD_NR with 1 << compound_order(page). Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reported-by: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* thp: fix interrupt unsafe locking in split_huge_page()Kirill A. Shutemov2016-01-201-4/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | split_queue_lock can be taken from interrupt context in some cases, but I forgot to convert locking in split_huge_page() to interrupt-safe primitives. Let's fix this. lockdep output: ====================================================== [ INFO: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected ] 4.4.0+ #259 Tainted: G W ------------------------------------------------------ syz-executor/18183 [HC0[0]:SC0[2]:HE0:SE0] is trying to acquire: (split_queue_lock){+.+...}, at: free_transhuge_page+0x24/0x90 mm/huge_memory.c:3436 and this task is already holding: (slock-AF_INET){+.-...}, at: spin_lock_bh include/linux/spinlock.h:307 (slock-AF_INET){+.-...}, at: lock_sock_fast+0x45/0x120 net/core/sock.c:2462 which would create a new lock dependency: (slock-AF_INET){+.-...} -> (split_queue_lock){+.+...} but this new dependency connects a SOFTIRQ-irq-safe lock: (slock-AF_INET){+.-...} ... which became SOFTIRQ-irq-safe at: mark_irqflags kernel/locking/lockdep.c:2799 __lock_acquire+0xfd8/0x4700 kernel/locking/lockdep.c:3162 lock_acquire+0x1dc/0x430 kernel/locking/lockdep.c:3585 __raw_spin_lock include/linux/spinlock_api_smp.h:144 _raw_spin_lock+0x33/0x50 kernel/locking/spinlock.c:151 spin_lock include/linux/spinlock.h:302 udp_queue_rcv_skb+0x781/0x1550 net/ipv4/udp.c:1680 flush_stack+0x50/0x330 net/ipv6/udp.c:799 __udp4_lib_mcast_deliver+0x694/0x7f0 net/ipv4/udp.c:1798 __udp4_lib_rcv+0x17dc/0x23e0 net/ipv4/udp.c:1888 udp_rcv+0x21/0x30 net/ipv4/udp.c:2108 ip_local_deliver_finish+0x2b3/0xa50 net/ipv4/ip_input.c:216 NF_HOOK_THRESH include/linux/netfilter.h:226 NF_HOOK include/linux/netfilter.h:249 ip_local_deliver+0x1c4/0x2f0 net/ipv4/ip_input.c:257 dst_input include/net/dst.h:498 ip_rcv_finish+0x5ec/0x1730 net/ipv4/ip_input.c:365 NF_HOOK_THRESH include/linux/netfilter.h:226 NF_HOOK include/linux/netfilter.h:249 ip_rcv+0x963/0x1080 net/ipv4/ip_input.c:455 __netif_receive_skb_core+0x1620/0x2f80 net/core/dev.c:4154 __netif_receive_skb+0x2a/0x160 net/core/dev.c:4189 netif_receive_skb_internal+0x1b5/0x390 net/core/dev.c:4217 napi_skb_finish net/core/dev.c:4542 napi_gro_receive+0x2bd/0x3c0 net/core/dev.c:4572 e1000_clean_rx_irq+0x4e2/0x1100 drivers/net/ethernet/intel/e1000e/netdev.c:1038 e1000_clean+0xa08/0x24a0 drivers/net/ethernet/intel/e1000/e1000_main.c:3819 napi_poll net/core/dev.c:5074 net_rx_action+0x7eb/0xdf0 net/core/dev.c:5139 __do_softirq+0x26a/0x920 kernel/softirq.c:273 invoke_softirq kernel/softirq.c:350 irq_exit+0x18f/0x1d0 kernel/softirq.c:391 exiting_irq ./arch/x86/include/asm/apic.h:659 do_IRQ+0x86/0x1a0 arch/x86/kernel/irq.c:252 ret_from_intr+0x0/0x20 arch/x86/entry/entry_64.S:520 arch_safe_halt ./arch/x86/include/asm/paravirt.h:117 default_idle+0x52/0x2e0 arch/x86/kernel/process.c:304 arch_cpu_idle+0xa/0x10 arch/x86/kernel/process.c:295 default_idle_call+0x48/0xa0 kernel/sched/idle.c:92 cpuidle_idle_call kernel/sched/idle.c:156 cpu_idle_loop kernel/sched/idle.c:252 cpu_startup_entry+0x554/0x710 kernel/sched/idle.c:300 rest_init+0x192/0x1a0 init/main.c:412 start_kernel+0x678/0x69e init/main.c:683 x86_64_start_reservations+0x2a/0x2c arch/x86/kernel/head64.c:195 x86_64_start_kernel+0x158/0x167 arch/x86/kernel/head64.c:184 to a SOFTIRQ-irq-unsafe lock: (split_queue_lock){+.+...} which became SOFTIRQ-irq-unsafe at: mark_irqflags kernel/locking/lockdep.c:2817 __lock_acquire+0x146e/0x4700 kernel/locking/lockdep.c:3162 lock_acquire+0x1dc/0x430 kernel/locking/lockdep.c:3585 __raw_spin_lock include/linux/spinlock_api_smp.h:144 _raw_spin_lock+0x33/0x50 kernel/locking/spinlock.c:151 spin_lock include/linux/spinlock.h:302 split_huge_page_to_list+0xcc0/0x1c50 mm/huge_memory.c:3399 split_huge_page include/linux/huge_mm.h:99 queue_pages_pte_range+0xa38/0xef0 mm/mempolicy.c:507 walk_pmd_range mm/pagewalk.c:50 walk_pud_range mm/pagewalk.c:90 walk_pgd_range mm/pagewalk.c:116 __walk_page_range+0x653/0xcd0 mm/pagewalk.c:204 walk_page_range+0xfe/0x2b0 mm/pagewalk.c:281 queue_pages_range+0xfb/0x130 mm/mempolicy.c:687 migrate_to_node mm/mempolicy.c:1004 do_migrate_pages+0x370/0x4e0 mm/mempolicy.c:1109 SYSC_migrate_pages mm/mempolicy.c:1453 SyS_migrate_pages+0x640/0x730 mm/mempolicy.c:1374 entry_SYSCALL_64_fastpath+0x16/0x7a arch/x86/entry/entry_64.S:185 other info that might help us debug this: Possible interrupt unsafe locking scenario: CPU0 CPU1 ---- ---- lock(split_queue_lock); local_irq_disable(); lock(slock-AF_INET); lock(split_queue_lock); <Interrupt> lock(slock-AF_INET); Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reported-by: Dmitry Vyukov <dvyukov@google.com> Acked-by: David Rientjes <rientjes@google.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* lib/libcrc32c.c: fix build warningJean Delvare2016-01-201-0/+1
| | | | | | | | | | | | Fix the following build warning: lib/libcrc32c.c:42:5: warning: no previous prototype for "crc32c" [-Wmissing-prototypes] u32 crc32c(u32 crc, const void *address, unsigned int length) ^ Signed-off-by: Jean Delvare <jdelvare@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Merge tag 'dmaengine-fix-4.5-rc1' of ↵Linus Torvalds2016-01-204-32/+30
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.infradead.org/users/vkoul/slave-dma Pull dmaengine fixes from Vinod Koul: "Here is my second pull request for this window: A few driver fixes have piled up and one missed rcar bindings patch which got somehow lost in for-linus branch so cherry-picked that one. Fixes are for dw, at_hdmac, edma" * tag 'dmaengine-fix-4.5-rc1' of git://git.infradead.org/users/vkoul/slave-dma: dmaengine: rcar-dmac: Document SoC specific bindings dmaengine: at_xdmac: fix resume for cyclic transfers dmaengine: dw: fix cyclic transfer callbacks dmaengine: dw: fix cyclic transfer setup dmaengine: edma: Fix paRAM slot allocation for entry channel 0
| * dmaengine: rcar-dmac: Document SoC specific bindingsSimon Horman2016-01-181-3/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In general Renesas hardware is not documented to the extent where the relationship between IP blocks on different SoCs can be assumed although they may appear to operate the same way. Furthermore the documentation typically does not specify a version for individual IP blocks. For these reasons a convention of using the SoC name in place of a version and providing SoC-specific compat strings has been adopted. Although not universally liked this convention is used in the bindings for most drivers for Renesas hardware. The purpose of this patch is to update the Renesas R-Car DMA Controller driver to follow this convention. Cc: devicetree@vger.kernel.org Cc: Magnus Damm <magnus.damm@gmail.com> Cc: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Signed-off-by: Simon Horman <horms+renesas@verge.net.au> Acked-by: Geert Uytterhoeven <geert+renesas@glider.be> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Vinod Koul <vinod.koul@intel.com>
| * dmaengine: at_xdmac: fix resume for cyclic transfersSongjun Wu2016-01-181-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | When having cyclic transfers, the channel was paused when performing suspend but was not correctly resumed. Signed-off-by: Songjun Wu <songjun.wu@atmel.com> Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com> Fixes: e1f7c9eee707 ("dmaengine: at_xdmac: creation of the atmel eXtended DMA Controller driver") Cc: <stable@vger.kernel.org> # 4.1 and later Signed-off-by: Vinod Koul <vinod.koul@intel.com>
| * dmaengine: dw: fix cyclic transfer callbacksMans Rullgard2016-01-141-6/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Cyclic transfer callbacks rely on block completion interrupts which were disabled in commit ff7b05f29fd4 ("dmaengine/dw_dmac: Don't handle block interrupts"). This re-enables block interrupts so the cyclic callbacks can work. Other transfer types are not affected as they set the INT_EN bit only on the last block. Fixes: ff7b05f29fd4 ("dmaengine/dw_dmac: Don't handle block interrupts") Signed-off-by: Mans Rullgard <mans@mansr.com> Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Vinod Koul <vinod.koul@intel.com> Cc: <stable@vger.kernel.org>
| * dmaengine: dw: fix cyclic transfer setupMans Rullgard2016-01-141-22/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 61e183f83069 ("dmaengine/dw_dmac: Reconfigure interrupt and chan_cfg register on resume") moved some channel initialisation to a new function which must be called before starting a transfer. This updates dw_dma_cyclic_start() to use dwc_dostart() like the other modes, thus ensuring dwc_initialize() gets called and removing some code duplication. Fixes: 61e183f83069 ("dmaengine/dw_dmac: Reconfigure interrupt and chan_cfg register on resume") Signed-off-by: Mans Rullgard <mans@mansr.com> Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Vinod Koul <vinod.koul@intel.com> Cc: <stable@vger.kernel.org>
| * dmaengine: edma: Fix paRAM slot allocation for entry channel 0Peter Ujfalusi2016-01-131-1/+1
| | | | | | | | | | | | | | | | edma_alloc_slot was not checking the channel mapping support existence when slot 0 has been requested (used as entry slot for channel/event 0). Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com> Signed-off-by: Vinod Koul <vinod.koul@intel.com>
* | Merge branch 'misc' of ↵Linus Torvalds2016-01-208-124/+188
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild Pull misc kbuild updates from Michal Marek: - Fix for make O=... perf-tar* - make tags revamp and fix for the fallout. Patch for warnings about line breaks inside DEFINE_PER_CPU macros is pending - New coccinelle test * 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild: coccinelle: tests: unsigned value cannot be lesser than zero tags: Unify emacs and exuberant rules tags: Drop the _PE rule tags: Do not try to index defconfigs tags: Process Kconfig files in a single pass tags: Fix erroneous pattern match in a comment aic7xxx: Avoid name collision with <linux/list.h> tags: Treat header files as C code package Makefile: fix perf-tar targets when outdir is set scripts/tags.sh: Teach tags about more powerpc macros
| * | coccinelle: tests: unsigned value cannot be lesser than zeroAndrzej Hajda2016-01-121-0/+75
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Unsigned expressions cannot be lesser than zero. Presence of comparisons 'unsigned (<|<=|>|>=) 0' often indicates a bug, usually wrong type of variable. The patch beside finding such comparisons tries to eliminate false positives, mainly by bypassing range checks. gcc can detect such comparisons also using -Wtype-limits switch, but it warns also in correct cases, making too much noise. Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Acked-by: Julia Lawall <julia.lawall@lip6.fr> Signed-off-by: Michal Marek <mmarek@suse.com>
| * | tags: Unify emacs and exuberant rulesMichal Marek2016-01-051-93/+109
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The emacs rules were constantly lagging behind the exuberant ones. Use a single set of rules for both, to make the script easier to maintain. The language understood by both tools is basic regular expression with some limitations, which are documented in a comment. To be able to store the rules in an array and easily iterate over it, the script requires bash now. In the exuberant case, the change fixes some false matches in <linux/page-flags.h> and also some too greedy matches in the arguments of the DECLARE_*/DEFINE_* macros. In the emacs case, several previously not working rules are matching now. Tested with these versions of the tools: Exuberant Ctags 5.8, Copyright (C) 1996-2009 Darren Hiebert etags (GNU Emacs 24.5) Signed-off-by: Michal Marek <mmarek@suse.com>
| * | tags: Drop the _PE ruleMichal Marek2016-01-051-2/+0
| | | | | | | | | | | | | | | | | | | | | We are not indexing the userspace tools, so the rules only match some false positives in the kernel code. Signed-off-by: Michal Marek <mmarek@suse.com>
| * | tags: Do not try to index defconfigsMichal Marek2016-01-051-12/+0
| | | | | | | | | | | | | | | | | | | | | | | | The defconfig files are in predictable locations, so there is no need to index them. Plus, the script was only looking for files named 'defconfig', which only works on a few architectures nowadays. Signed-off-by: Michal Marek <mmarek@suse.com>
| * | tags: Process Kconfig files in a single passMichal Marek2016-01-051-7/+2
| | | | | | | | | | | | Signed-off-by: Michal Marek <mmarek@suse.com>
| * | tags: Fix erroneous pattern match in a commentMichal Marek2016-01-051-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Apparently, ctags applies the rules before deleting comments: ctags: Warning: include/linux/completion.h:22: null expansion of name pattern "\2" Work around this particular case by requiring the group to contain at least one character. Leave the other patters as they are, until a better solution is found. Signed-off-by: Michal Marek <mmarek@suse.com>
| * | aic7xxx: Avoid name collision with <linux/list.h>Michal Marek2016-01-055-14/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rename the local definition of LIST_HEAD to BSD_LIST_HEAD. This fixes a ctags error if we apply the C rules to header files as well: ctags: Warning: drivers/scsi/aic7xxx/aic79xx.h:1072: null expansion of name pattern "\3" ctags: Warning: drivers/scsi/aic7xxx/aic7xxx.h:919: null expansion of name pattern "\3" Signed-off-by: Michal Marek <mmarek@suse.com>
| * | tags: Treat header files as C codeMichal Marek2016-01-051-33/+33
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows to apply the same patters to both source and header files. The effect is mostly visible in the case of DECLARE_BITMAP, but there are small gains all over the place. There is also lots of random changes in the diff, I believe this is simply because there are still lots of unexpanded macros in the code and the C and C++ parsers fail and recover at different points. Also, qconf.h is parsed as C, but that's a negligible regression. Signed-off-by: Michal Marek <mmarek@suse.com>
| * | package Makefile: fix perf-tar targets when outdir is setRiku Voipio2015-11-241-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | building with $srctree != $objtree, perf-tar-* targets fail to read the MANIFEST file and add the PERF-VERSION-FILE needed by out-of-tree builds. The build errors and an incorrect tar is created: $ make O=build-x86 perf-targz-src-pkg TAR cat: ../tools/perf/MANIFEST: No such file or directory tar: perf-4.1.0-rc8/PERF-VERSION-FILE: Cannot stat: No such file or dir.. tar: Exiting with failure status due to previous errors Kbuild sets objtree to "." and srctree to ".." The command to output MANIFEST becomes: $(cd ..; echo $(cat ../tools/perf/MANIFEST)) Without MANIFEST, the entire kernel source tree is added to the perf source tarball. The *correct* fix is to keep the cd and remove srctree from cat command line since MANIFEST has wildcards that fail to expand working directory isn't srctree. Second, PERF-VERSION-FILE gets not added, because in-tree build path is hardcoded to Makefile: util/PERF-VERSION-GEN ../../$(perf-tar)/ 2>/dev/null) The PERF-VERSION-GEN needs to be run from tools/perf directory, and the output directory needs to be changed from relative to to absolute. This can be achieved using the $(CURDIR) variable. Also remove the error redirect to /dev/null which hid the error. Signed-off-by: Riku Voipio <riku.voipio@linaro.org> Signed-off-by: Michal Marek <mmarek@suse.com>
| * | scripts/tags.sh: Teach tags about more powerpc macrosMichael Ellerman2015-11-201-0/+1
| | | | | | | | | | | | | | | | | | | | | Teach tags.sh about the powerpc PCI macros, eg. readl/writel etc. Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Michal Marek <mmarek@suse.com>
* | | Merge branch 'kconfig' of ↵Linus Torvalds2016-01-203-4/+7
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild Pull kconfig updates from Michal Marek: - Fix for make xconfig segfault - Handle long strings in config symbol values - Fix for mixing boolean and kconfig ternary type * 'kconfig' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild: kconfig: fix qconf segfault by deleting heap objects kconfig: return 'false' instead of 'no' in bool function kconfig: allow kconfig to handle longer path names
| * | | kconfig: fix qconf segfault by deleting heap objectsChris Bainbridge2016-01-111-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On Debian stable (qt-4.8.6) 'make xconfig' intermittently fails due to qconf segfaulting at exit time in QXcbEventReader. The cause of this is destructors on the heap objects never being called, so fix this by properly deleting the heap objects before exit. Signed-off-by: Chris Bainbridge <chris.bainbridge@gmail.com> Signed-off-by: Michal Marek <mmarek@suse.com>
| * | | kconfig: return 'false' instead of 'no' in bool functionVegard Nossum2016-01-051-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | menu_is_visible() is a bool function and should use boolean return values. "no" is a tristate value which happens to also have a value of 0, but we should nevertheless use the right symbol for it. This is a very minor cleanup with no semantic change. Fixes: 86e187ff9 ("kconfig: add an option to determine a menu's visibility") Cc: Arnaud Lacombe <lacombar@gmail.com> Cc: Mauro Carvalho Chehab <mchehab@redhat.com> Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com> Signed-off-by: Michal Marek <mmarek@suse.com>
| * | | kconfig: allow kconfig to handle longer path namesMarkus Mayer2015-12-101-3/+4
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The current (arbitrary) limit of 128 characters for path names has proven too short for Android builds, as longer path names are used there. Change conf.c, so it can handle path lengths up to PATH_MAX characters. Signed-off-by: Markus Mayer <mmayer@broadcom.com> Signed-off-by: Michal Marek <mmarek@suse.com>
OpenPOWER on IntegriCloud