- 22 Nov, 2015 4 commits
-
-
Jesper Dangaard Brouer authored
Initial implementation missed support for kmem cgroup support in kmem_cache_free_bulk() call, add this. If CONFIG_MEMCG_KMEM is not enabled, the compiler should be smart enough to not add any asm code. Incoming bulk free objects can belong to different kmem cgroups, and object free call can happen at a later point outside memcg context. Thus, we need to keep the orig kmem_cache, to correctly verify if a memcg object match against its "root_cache" (s->memcg_params.root_cache). Signed-off-by:
Jesper Dangaard Brouer <brouer@redhat.com> Reviewed-by:
Vladimir Davydov <vdavydov@virtuozzo.com> Cc: Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Jesper Dangaard Brouer authored
The call slab_pre_alloc_hook() interacts with kmemgc and is not allowed to be called several times inside the bulk alloc for loop, due to the call to memcg_kmem_get_cache(). This would result in hitting the VM_BUG_ON in __memcg_kmem_get_cache. As suggested by Vladimir Davydov, change slab_post_alloc_hook() to be able to handle an array of objects. A subtle detail is, loop iterator "i" in slab_post_alloc_hook() must have same type (size_t) as size argument. This helps the compiler to easier realize that it can remove the loop, when all debug statements inside loop evaluates to nothing. Note, this is only an issue because the kernel is compiled with GCC option: -fno-strict-overflow In slab_alloc_node() the compiler inlines and optimizes the invocation of slab_post_alloc_hook(s, flags, 1, &object) by removing the loop and access object directly. Signed-off-by:
-
Jesper Dangaard Brouer authored
This change focus on improving the speed of object freeing in the "slowpath" of kmem_cache_free_bulk. The calls slab_free (fastpath) and __slab_free (slowpath) have been extended with support for bulk free, which amortize the overhead of the (locked) cmpxchg_double. To use the new bulking feature, we build what I call a detached freelist. The detached freelist takes advantage of three properties: 1) the free function call owns the object that is about to be freed, thus writing into this memory is synchronization-free. 2) many freelist's can co-exist side-by-side in the same slab-page each with a separate head pointer. 3) it is the visibility of the head pointer that needs synchronization. Given these properties, the brilliant part is that the detached freelist can be constructed without any need for synchronization. The freelist is constructed directly in the page objects, without any synchronization needed. The detached freelist is allocated on the stack of the function call kmem_cache_free_bulk. Thus, the freelist head pointer is not visible to other CPUs. All objects in a SLUB freelist must belong to the same slab-page. Thus, constructing the detached freelist is about matching objects that belong to the same slab-page. The bulk free array is scanned is a progressive manor with a limited look-ahead facility. Kmem debug support is handled in call of slab_free(). Notice kmem_cache_free_bulk no longer need to disable IRQs. This only slowed down single free bulk with approx 3 cycles. Performance data: Benchmarked[1] obj size 256 bytes on CPU i7-4790K @ 4.00GHz SLUB fastpath single object quick reuse: 47 cycles(tsc) 11.931 ns To get stable and comparable numbers, the kernel have been booted with "slab_merge" (this also improve performance for larger bulk sizes). Performance data, compared against fallback bulking: bulk - fallback bulk - improvement with this patch 1 - 62 cycles(tsc) 15.662 ns - 49 cycles(tsc) 12.407 ns- improved 21.0% 2 - 55 cycles(tsc) 13.935 ns - 30 cycles(tsc) 7.506 ns - improved 45.5% 3 - 53 cycles(tsc) 13.341 ns - 23 cycles(tsc) 5.865 ns - improved 56.6% 4 - 52 cycles(tsc) 13.081 ns - 20 cycles(tsc) 5.048 ns - improved 61.5% 8 - 50 cycles(tsc) 12.627 ns - 18 cycles(tsc) 4.659 ns - improved 64.0% 16 - 49 cycles(tsc) 12.412 ns - 17 cycles(tsc) 4.495 ns - improved 65.3% 30 - 49 cycles(tsc) 12.484 ns - 18 cycles(tsc) 4.533 ns - improved 63.3% 32 - 50 cycles(tsc) 12.627 ns - 18 cycles(tsc) 4.707 ns - improved 64.0% 34 - 96 cycles(tsc) 24.243 ns - 23 cycles(tsc) 5.976 ns - improved 76.0% 48 - 83 cycles(tsc) 20.818 ns - 21 cycles(tsc) 5.329 ns - improved 74.7% 64 - 74 cycles(tsc) 18.700 ns - 20 cycles(tsc) 5.127 ns - improved 73.0% 128 - 90 cycles(tsc) 22.734 ns - 27 cycles(tsc) 6.833 ns - improved 70.0% 158 - 99 cycles(tsc) 24.776 ns - 30 cycles(tsc) 7.583 ns - improved 69.7% 250 - 104 cycles(tsc) 26.089 ns - 37 cycles(tsc) 9.280 ns - improved 64.4% Performance data, compared current in-kernel bulking: bulk - curr in-kernel - improvement with this patch 1 - 46 cycles(tsc) - 49 cycles(tsc) - improved (cycles:-3) -6.5% 2 - 27 cycles(tsc) - 30 cycles(tsc) - improved (cycles:-3) -11.1% 3 - 21 cycles(tsc) - 23 cycles(tsc) - improved (cycles:-2) -9.5% 4 - 18 cycles(tsc) - 20 cycles(tsc) - improved (cycles:-2) -11.1% 8 - 17 cycles(tsc) - 18 cycles(tsc) - improved (cycles:-1) -5.9% 16 - 18 cycles(tsc) - 17 cycles(tsc) - improved (cycles: 1) 5.6% 30 - 18 cycles(tsc) - 18 cycles(tsc) - improved (cycles: 0) 0.0% 32 - 18 cycles(tsc) - 18 cycles(tsc) - improved (cycles: 0) 0.0% 34 - 78 cycles(tsc) - 23 cycles(tsc) - improved (cycles:55) 70.5% 48 - 60 cycles(tsc) - 21 cycles(tsc) - improved (cycles:39) 65.0% 64 - 49 cycles(tsc) - 20 cycles(tsc) - improved (cycles:29) 59.2% 128 - 69 cycles(tsc) - 27 cycles(tsc) - improved (cycles:42) 60.9% 158 - 79 cycles(tsc) - 30 cycles(tsc) - improved (cycles:49) 62.0% 250 - 86 cycles(tsc) - 37 cycles(tsc) - improved (cycles:49) 57.0% Performance with normal SLUB merging is significantly slower for larger bulking. This is believed to (primarily) be an effect of not having to share the per-CPU data-structures, as tuning per-CPU size can achieve similar performance. bulk - slab_nomerge - normal SLUB merge 1 - 49 cycles(tsc) - 49 cycles(tsc) - merge slower with cycles:0 2 - 30 cycles(tsc) - 30 cycles(tsc) - merge slower with cycles:0 3 - 23 cycles(tsc) - 23 cycles(tsc) - merge slower with cycles:0 4 - 20 cycles(tsc) - 20 cycles(tsc) - merge slower with cycles:0 8 - 18 cycles(tsc) - 18 cycles(tsc) - merge slower with cycles:0 16 - 17 cycles(tsc) - 17 cycles(tsc) - merge slower with cycles:0 30 - 18 cycles(tsc) - 23 cycles(tsc) - merge slower with cycles:5 32 - 18 cycles(tsc) - 22 cycles(tsc) - merge slower with cycles:4 34 - 23 cycles(tsc) - 22 cycles(tsc) - merge slower with cycles:-1 48 - 21 cycles(tsc) - 22 cycles(tsc) - merge slower with cycles:1 64 - 20 cycles(tsc) - 48 cycles(tsc) - merge slower with cycles:28 128 - 27 cycles(tsc) - 57 cycles(tsc) - merge slower with cycles:30 158 - 30 cycles(tsc) - 59 cycles(tsc) - merge slower with cycles:29 250 - 37 cycles(tsc) - 56 cycles(tsc) - merge slower with cycles:19 Joint work with Alexander Duyck. [1] https://github.com/netoptimizer/prototype-kernel/blob/master/kernel/mm/slab_bulk_test01.c [akpm@linux-foundation.org: BUG_ON -> WARN_ON;return] Signed-off-by:Alexander Duyck <alexander.h.duyck@redhat.com> Acked-by:
Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Signed-off-by:
-
Jesper Dangaard Brouer authored
Make it possible to free a freelist with several objects by adjusting API of slab_free() and __slab_free() to have head, tail and an objects counter (cnt). Tail being NULL indicate single object free of head object. This allow compiler inline constant propagation in slab_free() and slab_free_freelist_hook() to avoid adding any overhead in case of single object free. This allows a freelist with several objects (all within the same slab-page) to be free'ed using a single locked cmpxchg_double in __slab_free() and with an unlocked cmpxchg_double in slab_free(). Object debugging on the free path is also extended to handle these freelists. When CONFIG_SLUB_DEBUG is enabled it will also detect if objects don't belong to the same slab-page. These changes are needed for the next patch to bulk free the detached freelists it introduces and constructs. Micro benchmarking showed no performance reduction due to this change, when debugging is turned off (compiled with CONFIG_SLUB_DEBUG). Signed-off-by:
-
- 21 Nov, 2015 20 commits
-
-
Jesper Dangaard Brouer authored
The #ifdef of CONFIG_SLUB_DEBUG is located very far from the associated #else. For readability mark it with a comment. Signed-off-by:
-
Christoph Lameter authored
Use the new function that can do allocation while interrupts are disabled. Avoids irq on/off sequences. Signed-off-by:
-
Christoph Lameter authored
Bulk alloc needs a function like that because it enables interrupts before calling __slab_alloc which promptly disables them again using the expensive local_irq_save(). Signed-off-by:
-
Jeff Layton authored
sparse says: include/linux/gfp.h:274:26: warning: incorrect type in return expression (different base types) include/linux/gfp.h:274:26: expected bool include/linux/gfp.h:274:26: got restricted gfp_t ...add a forced cast to silence the warning. Signed-off-by:Jeff Layton <jeff.layton@primarydata.com> Cc: Mel Gorman <mgorman@techsingularity.net> Signed-off-by:
-
Junxiao Bi authored
New created file's mode is not masked with umask, and this makes umask not work for ocfs2 volume. Fixes: 702e5bc6 ("ocfs2: use generic posix ACL infrastructure") Signed-off-by:
Junxiao Bi <junxiao.bi@oracle.com> Cc: Gang He <ghe@suse.com> Cc: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Viresh Kumar authored
Add entry for operating performance points into MAINTAINERS file. This will also allow get_maintainers to list OPP stakeholders properly. Signed-off-by:
Viresh Kumar <viresh.kumar@linaro.org> Acked-by:
Stephen Boyd <sboyd@codeaurora.org> Cc: Rafael Wysocki <rjw@rjwysocki.net> Signed-off-by:
-
Vitaly Kuznetsov authored
Commit 08d78658 ("panic: release stale console lock to always get the logbuf printed out") introduced an unwanted bad unlock balance report when panic() is called directly and not from OOPS (e.g. from out_of_memory()). The difference is that in case of OOPS we disable locks debug in oops_enter() and on direct panic call nobody does that. Fixes: 08d78658 ("panic: release stale console lock to always get the logbuf printed out") Reported-by:
kernel test robot <ying.huang@linux.intel.com> Signed-off-by:
Vitaly Kuznetsov <vkuznets@redhat.com> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Baoquan He <bhe@redhat.com> Cc: Prarit Bhargava <prarit@redhat.com> Cc: Xie XiuQi <xiexiuqi@huawei.com> Cc: Seth Jennings <sjenning@redhat.com> Cc: "K. Y. Srinivasan" <kys@microsoft.com> Cc: Jan Kara <jack@suse.cz> Cc: Petr Mladek <pmladek@suse.cz> Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com> Signed-off-by:
-
Richard Weinberger authored
sigsuspend() is nowhere used except in signal.c itself, so we can mark it static do not pollute the global namespace. But this patch is more than a boring cleanup patch, it fixes a real issue on UserModeLinux. UML has a special console driver to display ttys using xterm, or other terminal emulators, on the host side. Vegard reported that sometimes UML is unable to spawn a xterm and he's facing the following warning: WARNING: CPU: 0 PID: 908 at include/linux/thread_info.h:128 sigsuspend+0xab/0xc0() It turned out that this warning makes absolutely no sense as the UML xterm code calls sigsuspend() on the host side, at least it tries. But as the kernel itself offers a sigsuspend() symbol the linker choose this one instead of the glibc wrapper. Interestingly this code used to work since ever but always blocked signals on the wrong side. Some recent kernel change made the WARN_ON() trigger and uncovered the bug. It is a wonderful example of how much works by chance on computers. :-) Fixes: 68f3f16d ("new helper: sigsuspend()") Signed-off-by:
Richard Weinberger <richard@nod.at> Reported-by:
Vegard Nossum <vegard.nossum@oracle.com> Tested-by:
Oleg Nesterov <oleg@redhat.com> Cc: <stable@vger.kernel.org> [3.5+] Signed-off-by:
-
Andrey Ryabinin authored
Kmemleak reports the following leak: unreferenced object 0xfffffbfff41ea000 (size 20480): comm "modprobe", pid 65199, jiffies 4298875551 (age 542.568s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [<ffffffff82354f5e>] kmemleak_alloc+0x4e/0xc0 [<ffffffff8152e718>] __vmalloc_node_range+0x4b8/0x740 [<ffffffff81574072>] kasan_module_alloc+0x72/0xc0 [<ffffffff810efe68>] module_alloc+0x78/0xb0 [<ffffffff812f6a24>] module_alloc_update_bounds+0x14/0x70 [<ffffffff812f8184>] layout_and_allocate+0x16f4/0x3c90 [<ffffffff812faa1f>] load_module+0x2ff/0x6690 [<ffffffff813010b6>] SyS_finit_module+0x136/0x170 [<ffffffff8239bbc9>] system_call_fastpath+0x16/0x1b [<ffffffffffffffff>] 0xffffffffffffffff kasan_module_alloc() allocates shadow memory for module and frees it on module unloading. It doesn't store the pointer to allocated shadow memory because it could be calculated from the shadowed address, i.e. kasan_mem_to_shadow(addr). Since kmemleak cannot find pointer to allocated shadow, it thinks that memory leaked. Use kmemleak_ignore() to tell kmemleak that this is not a leak and shadow memory doesn't contain any pointers. Signed-off-by:
Andrey Ryabinin <aryabinin@virtuozzo.com> Acked-by:
Catalin Marinas <catalin.marinas@arm.com> Signed-off-by:
-
OGAWA Hirofumi authored
For the root directory, . and .. are faked (using dir_emit_dots()) and ctx->pos is reset from 2 to 0. A corrupted root directory could cause fat_get_entry() to fail, but ->iterate() (fat_readdir()) reports progress to the VFS (with ctx->pos rewound to 0), so any following calls to ->iterate() continue to return the same entries again and again. The result is that userspace will never see the end of the directory, causing e.g. 'ls' to hang in a getdents() loop. [hirofumi@mail.parknet.co.jp: cleanup and make sure to correct fake_offset] Reported-by:
Richard Weinberger <richard.weinberger@gmail.com> Signed-off-by:
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Mike Kravetz authored
Hugh Dickins pointed out problems with the new hugetlbfs fallocate hole punch code. These problems are in the routine remove_inode_hugepages and mostly occur in the case where there are holes in the range of pages to be removed. These holes could be the result of a previous hole punch or simply sparse allocation. The current code could access pages outside the specified range. remove_inode_hugepages handles both hole punch and truncate operations. Page index handling was fixed/cleaned up so that the loop index always matches the page being processed. The code now only makes a single pass through the range of pages as it was determined page faults could not race with truncate. A cond_resched() was added after removing up to PAGEVEC_SIZE pages. Some totally unnecessary code in hugetlbfs_fallocate() that remained from early development was also removed. Tested with fallocate tests submitted here: http://librelist.com/browser//libhugetlbfs/2015/6/25/patch-tests-add-tests-for-fallocate-system-call/ And, some ftruncate tests under development Fixes: b5cec28d ("hugetlbfs: truncate_hugepages() takes a range of pages") Signed-off-by:
Mike Kravetz <mike.kravetz@oracle.com> Acked-by:
Hugh Dickins <hughd@google.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: "Hillf Danton" <hillf.zj@alibaba-inc.com> Cc: <stable@vger.kernel.org> [4.3] Signed-off-by:
-
Yang Shi authored
When building kernel with gcc 5.2, the below warning is raised: mm/page-writeback.c: In function 'balance_dirty_pages.isra.10': mm/page-writeback.c:1545:17: warning: 'm_dirty' may be used uninitialized in this function [-Wmaybe-uninitialized] unsigned long m_dirty, m_thresh, m_bg_thresh; The m_dirty{thresh, bg_thresh} are initialized in the block of "if (mdtc)", so if mdts is null, they won't be initialized before being used. Initialize m_dirty to zero, also initialize m_thresh and m_bg_thresh to keep consistency. They are used later by if condition: !mdtc || m_dirty <= dirty_freerun_ceiling(m_thresh, m_bg_thresh) If mdtc is null, dirty_freerun_ceiling will not be called at all, so the initialization will not change any behavior other than just ceasing the compile warning. (akpm: the patch actually reduces .text size by ~20 bytes on gcc-4.x.y) [akpm@linux-foundation.org: add comment] Signed-off-by:Yang Shi <yang.shi@linaro.org> Cc: Tejun Heo <tj@kernel.org> Signed-off-by:
-
Christoph Hellwig authored
pci_set_dma_mask returns a negative errno value, not a bool like pci_dma_supported. This of course was just a giant test for attention :) Signed-off-by:
Christoph Hellwig <hch@lst.de> Reported-by:
Jongman Heo <jongman.heo@samsung.com> Tested-by: Jongman Heo <jongman.heo@samsung.com> [pcnet32] Acked-by:
Mauro Carvalho Chehab <mchehab@osg.samsung.com> Cc: Hans Verkuil <hverkuil@xs4all.nl> Cc: Antti Palosaari <crope@iki.fi> Signed-off-by:
-
Jason J. Herne authored
MADV_NOHUGEPAGE processing is too restrictive. kvm already disables hugepage but hugepage_madvise() takes the error path when we ask to turn on the MADV_NOHUGEPAGE bit and the bit is already on. This causes Qemu's new postcopy migration feature to fail on s390 because its first action is to madvise the guest address space as NOHUGEPAGE. This patch modifies the code so that the operation succeeds without error now. For consistency reasons do the same for MADV_HUGEPAGE. Signed-off-by:
Jason J. Herne <jjherne@linux.vnet.ibm.com> Reviewed-by:
Andrea Arcangeli <aarcange@redhat.com> Acked-by:
Christian Borntraeger <borntraeger@de.ibm.com> Acked-by:
David Rientjes <rientjes@google.com> Signed-off-by:
-
Jerome Marchand authored
Commit 71394fe5 ("mm: vmalloc: add flag preventing guard hole allocation") missed a spot. Currently remove_vm_area() decreases vm->size to "remove" the guard hole page, even when it isn't present. All but one users just free the vm_struct rigth away and never access vm->size anyway. Don't touch the size in remove_vm_area() and have __vunmap() use the proper get_vm_area_size() helper. Signed-off-by:
Jerome Marchand <jmarchan@redhat.com> Acked-by:
-
Naoya Horiguchi authored
PageIdle is exported in include/uapi/linux/kernel-page-flags.h, so let's make page-types.c tool handle it. Signed-off-by:
Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Reviewed-by:
-
Dan Carpenter authored
This code causes a static checker warning because it's a user controlled variable where we cap the upper bound but not the lower bound. Let's return an -EINVAL for negative timeouts. [akpm@linux-foundation.org: remove unneeded `else'] Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by:
Jan Kara <jack@suse.com> Cc: Petr Vandrovec <petr@vandrovec.name> Cc: David Howells <dhowells@redhat.com> Cc: Jan Kara <jack@suse.cz> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by:
-
Daniel Baluta authored
This patchset introduces IIO software triggers, offers a way of configuring them via configfs and adds the IIO hrtimer based interrupt source to be used with software triggers. The architecture is now split in 3 parts, to remove all IIO trigger specific parts from IIO configfs core: (1) IIO configfs - creates the root of the IIO configfs subsys. (2) IIO software triggers - software trigger implementation, dynamically creating /config/iio/triggers group. (3) IIO hrtimer trigger - is the first interrupt source for software triggers (with syfs to follow). Each trigger type can implement its own set of attributes. Lockdep seems to be happy with the locking in configfs patch. This patch (of 5): We don't want to hardcode default groups at subsystem creation time. We export: * configfs_register_group * configfs_unregister_group to allow drivers to programatically create/destroy groups later, after module init time. This is needed for IIO configfs support. (akpm: the other 4 patches to be merged via the IIO tree) Signed-off-by:Daniel Baluta <daniel.baluta@intel.com> Suggested-by:
Lars-Peter Clausen <lars@metafoo.de> Reviewed-by:
Joel Becker <jlbec@evilplan.org> Cc: Hartmut Knaack <knaack.h@gmx.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Paul Bolle <pebolle@tiscali.nl> Cc: Adriana Reus <adriana.reus@intel.com> Cc: Cristina Opriceana <cristina.opriceana@gmail.com> Cc: Peter Meerwald <pmeerw@pmeerw.net> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Signed-off-by:
-
Moritz Fischer authored
Nominate myself as Reviewer. Signed-off-by:
Moritz Fischer <moritz.fischer@ettus.com> Acked-by:
Alan Tull <atull@opensource.altera.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
-
Rasmus Villemoes authored
The various allocators return aligned memory. Telling the compiler that allows it to generate better code in many cases, for example when the return value is immediately passed to memset(). Some code does become larger, but at least we win twice as much as we lose: $ scripts/bloat-o-meter /tmp/vmlinux vmlinux add/remove: 0/0 grow/shrink: 13/52 up/down: 995/-2140 (-1145) An example of the different (and smaller) code can be seen in mm_alloc(). Before: : 48 8d 78 08 lea 0x8(%rax),%rdi : 48 89 c1 mov %rax,%rcx : 48 89 c2 mov %rax,%rdx : 48 c7 00 00 00 00 00 movq $0x0,(%rax) : 48 c7 80 48 03 00 00 movq $0x0,0x348(%rax) : 00 00 00 00 : 31 c0 xor %eax,%eax : 48 83 e7 f8 and $0xfffffffffffffff8,%rdi : 48 29 f9 sub %rdi,%rcx : 81 c1 50 03 00 00 add $0x350,%ecx : c1 e9 03 shr $0x3,%ecx : f3 48 ab rep stos %rax,%es:(%rdi) After: : 48 89 c2 mov %rax,%rdx : b9 6a 00 00 00 mov $0x6a,%ecx : 31 c0 xor %eax,%eax : 48 89 d7 mov %rdx,%rdi : f3 48 ab rep stos %rax,%es:(%rdi) So gcc's strategy is to do two possibly (but not really, of course) unaligned stores to the first and last word, then do an aligned rep stos covering the middle part with a little overlap. Maybe arches which do not allow unaligned stores gain even more. I don't know if gcc can actually make use of alignments greater than 8 for anything, so one could probably drop the __assume_xyz_alignment macros and just use __assume_aligned(8). The increases in code size are mostly caused by gcc deciding to opencode strlen() using the check-four-bytes-at-a-time trick when it knows the buffer is sufficiently aligned (one function grew by 200 bytes). Now it turns out that many of these strlen() calls showing up were in fact redundant, and they're gone from -next. Applying the two patches to next-20151001 bloat-o-meter instead says add/remove: 0/0 grow/shrink: 6/52 up/down: 244/-2140 (-1896) Signed-off-by:
Rasmus Villemoes <linux@rasmusvillemoes.dk> Acked-by:
-
- 20 Nov, 2015 10 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pmLinus Torvalds authored
Pull more power management and ACPI updates from Rafael Wysocki: "These are mostly fixes and cleanups (ACPI core, PM core, cpufreq, ACPI EC driver, device properties) including three reverts of recent intel_pstate driver commits due to a regression introduced by one of them plus support for Atom Airmont cores in intel_pstate (which really boils down to adding new frequency tables for Airmont) and additional turbostat updates. Specifics: - Revert three recent intel_pstate driver commits one of which introduced a regression and the remaining two depend on the problematic one (Rafael Wysocki). - Fix breakage related to the recently introduced ACPI _CCA object support in the PCI DMA setup code (Suravee Suthikulpanit). - Fix up the recently introduced ACPI CPPC support to only use the hardware-reduced version of the PCCT structure as the only architecture to support it (ARM64) will only use hardware-reduced ACPI anyway (Ashwin Chaugule). - Fix a cpufreq mediatek driver build problem (Arnd Bergmann). - Fix the SMBus transaction handling implementation in the ACPI core to avoid re-entrant calls to wait_event_timeout() which makes intermittent boot stalls related to the Smart Battery Subsystem initialization go away and revert a workaround of another problem with the same underlying root cause (Chris Bainbridge). - Fix the generic wakeup interrupts framework to avoid using invalid IRQ numbers (Dmitry Torokhov). - Remove a redundant check from the ACPI EC driver (Markus Elfring). - Modify the intel_pstate driver so it can support more Atom flavors than just one (Baytrail) and add support for Atom Airmont cores (which require new freqnency tables) to it (Philippe Longepe). - Clean up MSR-related symbols in turbostat (Len Brown)" * tag 'pm+acpi-4.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: PCI: Fix OF logic in pci_dma_configure() Revert "Documentation: kernel_parameters for Intel P state driver" cpufreq: mediatek: fix build error cpufreq: intel_pstate: Add separate support for Airmont cores cpufreq: intel_pstate: Replace BYT with ATOM Revert "cpufreq: intel_pstate: Use ACPI perf configuration" Revert "cpufreq: intel_pstate: Avoid calculation for max/min" ACPI-EC: Drop unnecessary check made before calling acpi_ec_delete_query() Revert "ACPI / SBS: Add 5 us delay to fix SBS hangs on MacBook" ACPI / SMBus: Fix boot stalls / high CPU caused by reentrant code PM / wakeirq: check that wake IRQ is valid before accepting it ACPI / CPPC: Use h/w reduced version of the PCCT structure x86: remove unused definition of MSR_NHM_PLATFORM_INFO tools/power turbostat: use new name for MSR_PLATFORM_INFO -
git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linuxLinus Torvalds authored
Pull powerpc fixlet from Michael Ellerman: "Wire up sys_mlock2()" * tag 'powerpc-4.4-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: powerpc: Wire up sys_mlock2()
-
git://git.infradead.org/users/vkoul/slave-dmaLinus Torvalds authored
Pull dmaengine fixes from Vinod Koul: "This has odd fixes spreadout drivers, not major here - usbdmac fixes for pm - edma build and logic fixes - build warn fixes for few drivers" * tag 'dmaengine-fix-4.4-rc2' of git://git.infradead.org/users/vkoul/slave-dma: dmaengine: at_hdmac: use %pad format string for dma_addr_t dmaengine: at_xdmac: use %pad format string for dma_addr_t dmaengine: imx-sdma: remove __init annotation on sdma_event_remap dmaengine: edma: predecence bug in GET_NUM_QDMACH() dmaengine: edma: fix build without CONFIG_OF dmaengine: of_dma: Correct return code for of_dma_request_slave_channel in case !CONFIG_OF dmaengine: sh: usb-dmac: Fix pm_runtime_{enable,disable}() imbalance dmaengine: sh: usb-dmac: Fix crash on runtime suspend
-
git://people.freedesktop.org/~airlied/linuxLinus Torvalds authored
Pull drm fixes from Dave Airlie: "A varied bunch of fixes, the radeon pull is probably a bit larger than I'd like, but it contains 2 weeks of stuff, and the Fiji fixes are a bit large, but they are Fiji specific. Otherwise: - mgag200: One cursor regression oops fix. - vc4: A few small fixes and cleanups. - core: Atomic fixes and Atomic helper fixes - i915: Revert for the backlight regression along with a bunch of fixes" * 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (58 commits) drm/atomic-helper: Check encoder/crtc constraints Revert "drm/i915: skip modeset if compatible for everyone." drm/mgag200: fix kernel hang in cursor code. drm/amdgpu: reserve/unreserve objects out of map/unmap operations drm/amdgpu: move bo_reserve out of amdgpu_vm_clear_bo drm/amdgpu: add lock for interval tree in vm drm/amdgpu: keep the owner for VMIDs drm/amdgpu: move VM manager clean into the VM code again drm/amdgpu: cleanup VM coding style drm/amdgpu: remove unused VM manager field drm/amdgpu: cleanup scheduler command submission drm/amdgpu: fix typo in firmware name drm/i915: Consider SPLL as another shared pll, v2. drm/i915: Fix gpu frequency change tracing drm/vc4: Make sure that planes aren't scaled. drm/vc4: Fix some failure to track __iomem decorations on pointers. drm/vc4: checking for NULL instead of IS_ERR drm/vc4: fix itnull.cocci warnings drm/vc4: fix platform_no_drv_owner.cocci warnings drm/vc4: vc4_plane_duplicate_state() can be static ... -
git://git.code.sf.net/p/openipmi/linux-ipmiLinus Torvalds authored
Pull IPMI updates from Corey Minyard: "Some fixes for small IPMI problems. The most significant is that the driver wasn't starting the timer for some messages, which would result in problems if that message failed for some reason. The others are small optimizations or making things a little neater" * tag 'for-linus-4.4' of git://git.code.sf.net/p/openipmi/linux-ipmi: ipmi watchdog : add panic_wdt_timeout parameter char: ipmi: Move MODULE_DEVICE_TABLE() to follow struct ipmi: Stop the timer immediately if idle ipmi: Start the timer and thread on internal msgs
-
Linus Torvalds authored
Merge tag 'renesas-sh-drivers-for-v4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas Pull SH driver fixlet from Simon Horman: "I am sending this change after v4.4-rc1 has been released as it depends on SoC changes which are present in that rc: = Remove now unnecessary reference to CONFIG_ARCH_SHMOBILE_MULTI" * tag 'renesas-sh-drivers-for-v4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas: drivers: sh: Get rid of CONFIG_ARCH_SHMOBILE_MULTI
-
Rafael J. Wysocki authored
* acpi-smbus: Revert "ACPI / SBS: Add 5 us delay to fix SBS hangs on MacBook" ACPI / SMBus: Fix boot stalls / high CPU caused by reentrant code * acpi-ec: ACPI-EC: Drop unnecessary check made before calling acpi_ec_delete_query() * acpi-pci: PCI: Fix OF logic in pci_dma_configure()
-
Rafael J. Wysocki authored
* pm-sleep: PM / wakeirq: check that wake IRQ is valid before accepting it
-
Rafael J. Wysocki authored
* pm-cpufreq: Revert "Documentation: kernel_parameters for Intel P state driver" cpufreq: mediatek: fix build error cpufreq: intel_pstate: Add separate support for Airmont cores cpufreq: intel_pstate: Replace BYT with ATOM Revert "cpufreq: intel_pstate: Use ACPI perf configuration" Revert "cpufreq: intel_pstate: Avoid calculation for max/min" * acpi-cppc: ACPI / CPPC: Use h/w reduced version of the PCCT structure
-
Suravee Suthikulpanit authored
This patch fixes a bug introduced by previous commit, which incorrectly checkes the of_node of the end-point device. Instead, it should check the of_node of the host bridge. Fixes: 50230713 ("PCI: OF: Move of_pci_dma_configure() to pci_dma_configure()") Reported-by:
Robin Murphy <robin.murphy@arm.com> Signed-off-by:
Suravee Suthikulpanit <suravee.suthikulpanit@amd.com> Acked-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com>
-
- 19 Nov, 2015 6 commits
-
-
git://anongit.freedesktop.org/drm-intelDave Airlie authored
i915 fixes for 4.4, including the revert for the backlight regression Olof reported. Otherwise fixes all around. * tag 'drm-intel-fixes-2015-11-19' of git://anongit.freedesktop.org/drm-intel: Revert "drm/i915: skip modeset if compatible for everyone." drm/i915: Consider SPLL as another shared pll, v2. drm/i915: Fix gpu frequency change tracing drm/i915: Don't clobber the addfb2 ioctl params drm/i915: Clear intel_crtc->atomic before updating it. drm/i915: get runtime PM reference around GEM set_caching IOCTL drm/i915: Fix GT frequency rounding drm/i915: quirk backlight present on Macbook 4, 1 drm/i915: Fix crtc_y assignment in intel_find_initial_plane_obj()
-
git://anongit.freedesktop.org/drm-intelDave Airlie authored
Here are some drm core fixes for v4.4 that I've picked up. Atomic fixes from Maarten, and atomic helper fixes from Ville and Daniel. Admittedly the topmost commit didn't sit in our tree for very long, but does come with reviews and testing from trustworthy people. * tag 'topic/drm-fixes-2015-11-19' of git://anongit.freedesktop.org/drm-intel: drm/atomic-helper: Check encoder/crtc constraints drm: Fix primary plane size for stereo doubled modes for legacy setcrtc drm/core: Fix old_fb handling in pan_display_atomic. drm/core: Fix old_fb handling in restore_fbdev_mode_atomic. drm/atomic: add a drm_atomic_clean_old_fb helper. drm/core: Fix old_fb handling in drm_mode_atomic_ioctl. drm/core: Set legacy_cursor_update in drm_atomic_helper_disable_plane.
-
git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linuxLinus Torvalds authored
Pull arm64 fixes from Catalin Marinas: - Fix size alignment in __iommu_{alloc,free}_attrs - Kernel memory mapping fix with CONFIG_DEBUG_RODATA for page sizes other than 4KB and a fix of the mark_rodata_ro permissions - dma_get_ops() simplification and behaviour alignment between DT and ACPI - function_graph trace fix for cpu_suspend() (CPUs returning from deep sleep via a different path and confusing the tracer) - Use of non-global mappings for UEFI run-time services to avoid a (potentially theoretical) TLB conflict - Crypto priority reduction of core AES cipher (the accelerated asynchronous implementation is preferred when available) - Reverting an old commit that removed BogoMIPS from /proc/cpuinfo on arm64. Apparently, we had it for a relatively short time and libvirt started checking for its presence - Compiler warnings fixed (ptrace.h inclusion from compat.h, smp_load_acquire with const argument) * tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux: arm64: restore bogomips information in /proc/cpuinfo arm64: barriers: fix smp_load_acquire to work with const arguments arm64: Fix R/O permissions in mark_rodata_ro arm64: crypto: reduce priority of core AES cipher arm64: use non-global mappings for UEFI runtime regions arm64: kernel: pause/unpause function graph tracer in cpu_suspend() arm64: do not include ptrace.h from compat.h arm64: simplify dma_get_ops arm64: mm: use correct mapping granularity under DEBUG_RODATA arm64/dma-mapping: Fix sizes in __iommu_{alloc,free}_attrs -
git://git.kernel.org/pub/scm/linux/kernel/git/jikos/livepatchingLinus Torvalds authored
Pull livepatching fix from Jiri Kosina: "A fix for module handling in case kASLR has been enabled, from Zhou Chengming" * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/livepatching: livepatch: x86: fix relocation computation with kASLR
-
git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hidLinus Torvalds authored
Pull HID fixes from Jiri Kosina: "Two functional fixes for wacom HID driver from Ping Cheng and Jiri Kosina" * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid: HID: wacom: fixup quirks setup for WACOM_DEVICETYPE_PAD HID: wacom: Add outbounding area for DTU1141
-
git://git.linaro.org/people/ulf.hansson/mmcLinus Torvalds authored
Pull MMC fixes from Ulf Hansson: "Here are some mmc fixes intended for v4.4 rc2. It's based on a commit prior rc1 as I wanted to get them a bit more tested in next before sending you the pull request. MMC core: - Improve reliability when selecting HS200 mode - Improve reliability when selecting HS400 mode - mmc: remove bondage between REQ_META and reliable write MMC host: - pxamci: Fix read-only gpio detection polarity - mtk-sd: Preinitialize delay_phase to fix the case when delay is zero - android-goldfish: Fix build dependency by adding HAS_DMA - dw_mmc: Remove Seungwon Jeon from MAINTAINERS" * tag 'mmc-v4.4-rc1' of git://git.linaro.org/people/ulf.hansson/mmc: mmc: remove bondage between REQ_META and reliable write mmc: MMC_GOLDFISH should depend on HAS_DMA mmc: mediatek: Preinitialize delay_phase in get_best_delay() MAINTAINERS: mmc: Remove Seungwon Jeon from dw_mmc mmc: mmc: Improve reliability of mmc_select_hs400() mmc: mmc: Move mmc_switch_status() mmc: mmc: Fix HS setting in mmc_select_hs400() mmc: mmc: Improve reliability of mmc_select_hs200() mmc: pxamci: fix read-only gpio detection polarity
-
