kernel/linux.git/include/linux, branch v4.18.14

mm: migration: fix migration of huge PMD shared pages

2018-10-13T07:33:08+00:00

commit 017b1660df89f5fb4bfe66c34e35f7d2031100c7 upstream. The page migration code employs try_to_unmap() to try and unmap the source page. This is accomplished by using rmap_walk to find all vmas where the page is mapped. This search stops when page mapcount is zero. For shared PMD huge pages, the page map count is always 1 no matter the number of mappings. Shared mappings are tracked via the reference count of the PMD page. Therefore, try_to_unmap stops prematurely and does not completely unmap all mappings of the source page. This problem can result is data corruption as writes to the original source page can happen after contents of the page are copied to the target page. Hence, data is lost. This problem was originally seen as DB corruption of shared global areas after a huge page was soft offlined due to ECC memory errors. DB developers noticed they could reproduce the issue by (hotplug) offlining memory used to back huge pages. A simple testcase can reproduce the problem by creating a shared PMD mapping (note that this must be at least PUD_SIZE in size and PUD_SIZE aligned (1GB on x86)), and using migrate_pages() to migrate process pages between nodes while continually writing to the huge pages being migrated. To fix, have the try_to_unmap_one routine check for huge PMD sharing by calling huge_pmd_unshare for hugetlbfs huge pages. If it is a shared mapping it will be 'unshared' which removes the page table entry and drops the reference on the PMD page. After this, flush caches and TLB. mmu notifiers are called before locking page tables, but we can not be sure of PMD sharing until page tables are locked. Therefore, check for the possibility of PMD sharing before locking so that notifiers can prepare for the worst possible case. Link: http://lkml.kernel.org/r/20180823205917.16297-2-mike.kravetz@oracle.com [mike.kravetz@oracle.com: make _range_in_vma() a static inline] Link: http://lkml.kernel.org/r/6063f215-a5c8-2f0c-465a-2c515ddc952d@oracle.com Fixes: 39dde65c9940 ("shared page table for hugetlb page") Signed-off-by: Mike Kravetz Acked-by: Kirill A. Shutemov Reviewed-by: Naoya Horiguchi Acked-by: Michal Hocko Cc: Vlastimil Babka Cc: Davidlohr Bueso Cc: Jerome Glisse Cc: Mike Kravetz Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

new primitive: discard_new_inode()

2018-10-10T06:56:04+00:00

commit c2b6d621c4ffe9936adf7a55c8b1c769672c306f upstream. We don't want open-by-handle picking half-set-up in-core struct inode from e.g. mkdir() having failed halfway through. In other words, we don't want such inodes returned by iget_locked() on their way to extinction. However, we can't just have them unhashed - otherwise open-by-handle immediately *after* that would've ended up creating a new in-core inode over the on-disk one that is in process of being freed right under us. Solution: new flag (I_CREATING) set by insert_inode_locked() and removed by unlock_new_inode() and a new primitive (discard_new_inode()) to be used by such halfway-through-setup failure exits instead of unlock_new_inode() / iput() combinations. That primitive unlocks new inode, but leaves I_CREATING in place. iget_locked() treats finding an I_CREATING inode as failure (-ESTALE, once we sort out the error propagation). insert_inode_locked() treats the same as instant -EBUSY. ilookup() treats those as icache miss. [Fix by Dan Carpenter folded in] Signed-off-by: Al Viro Cc: Miklos Szeredi Cc: Amir Goldstein Signed-off-by: Greg Kroah-Hartman

Revert "blk-throttle: fix race between blkcg_bio_issue_check() and cgroup_rmdir()"

2018-10-10T06:55:54+00:00

[ Upstream commit 6b06546206868f723f2061d703a3c3c378dcbf4c ] This reverts commit 4c6994806f708559c2812b73501406e21ae5dcd0. Destroying blkgs is tricky because of the nature of the relationship. A blkg should go away when either a blkcg or a request_queue goes away. However, blkg's pin the blkcg to ensure they remain valid. To break this cycle, when a blkcg is offlined, blkgs put back their css ref. This eventually lets css_free() get called which frees the blkcg. The above commit (4c6994806f70) breaks this order of events by trying to destroy blkgs in css_free(). As the blkgs still hold references to the blkcg, css_free() is never called. The race between blkcg_bio_issue_check() and cgroup_rmdir() will be addressed in the following patch by delaying destruction of a blkg until all writeback associated with the blkcg has been finished. Fixes: 4c6994806f70 ("blk-throttle: fix race between blkcg_bio_issue_check() and cgroup_rmdir()") Reviewed-by: Josef Bacik Signed-off-by: Dennis Zhou Cc: Jiufei Xue Cc: Joseph Qi Cc: Tejun Heo Cc: Jens Axboe Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman

arm/arm64: smccc-1.1: Handle function result as parameters

2018-10-03T23:59:27+00:00

[ Upstream commit 755a8bf5579d22eb5636685c516d8dede799e27b ] If someone has the silly idea to write something along those lines: extern u64 foo(void); void bar(struct arm_smccc_res *res) { arm_smccc_1_1_smc(0xbad, foo(), res); } they are in for a surprise, as this gets compiled as: 0000000000000588 : 588: a9be7bfd stp x29, x30, [sp, #-32]! 58c: 910003fd mov x29, sp 590: f9000bf3 str x19, [sp, #16] 594: aa0003f3 mov x19, x0 598: aa1e03e0 mov x0, x30 59c: 94000000 bl 0 <_mcount> 5a0: 94000000 bl 0 5a4: aa0003e1 mov x1, x0 5a8: d4000003 smc #0x0 5ac: b4000073 cbz x19, 5b8 5b0: a9000660 stp x0, x1, [x19] 5b4: a9010e62 stp x2, x3, [x19, #16] 5b8: f9400bf3 ldr x19, [sp, #16] 5bc: a8c27bfd ldp x29, x30, [sp], #32 5c0: d65f03c0 ret 5c4: d503201f nop The call to foo "overwrites" the x0 register for the return value, and we end up calling the wrong secure service. A solution is to evaluate all the parameters before assigning anything to specific registers, leading to the expected result: 0000000000000588 : 588: a9be7bfd stp x29, x30, [sp, #-32]! 58c: 910003fd mov x29, sp 590: f9000bf3 str x19, [sp, #16] 594: aa0003f3 mov x19, x0 598: aa1e03e0 mov x0, x30 59c: 94000000 bl 0 <_mcount> 5a0: 94000000 bl 0 5a4: aa0003e1 mov x1, x0 5a8: d28175a0 mov x0, #0xbad 5ac: d4000003 smc #0x0 5b0: b4000073 cbz x19, 5bc 5b4: a9000660 stp x0, x1, [x19] 5b8: a9010e62 stp x2, x3, [x19, #16] 5bc: f9400bf3 ldr x19, [sp, #16] 5c0: a8c27bfd ldp x29, x30, [sp], #32 5c4: d65f03c0 ret Reported-by: Julien Grall Signed-off-by: Marc Zyngier Signed-off-by: Will Deacon Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman

arm/arm64: smccc-1.1: Make return values unsigned long

2018-10-03T23:59:26+00:00

[ Upstream commit 1d8f574708a3fb6f18c85486d0c5217df893c0cf ] An unfortunate consequence of having a strong typing for the input values to the SMC call is that it also affects the type of the return values, limiting r0 to 32 bits and r{1,2,3} to whatever was passed as an input. Let's turn everything into "unsigned long", which satisfies the requirements of both architectures, and allows for the full range of return values. Reported-by: Julien Grall Signed-off-by: Marc Zyngier Signed-off-by: Will Deacon Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman

hwmon: (ina2xx) fix sysfs shunt resistor read access

2018-10-03T23:59:25+00:00

[ Upstream commit 3ad867001c91657c46dcf6656d52eb6080286fd5 ] fix the sysfs shunt resistor read access: return the shunt resistor value, not the calibration register contents. update email address Signed-off-by: Lothar Felten Signed-off-by: Guenter Roeck Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman

uaccess: Fix is_source param for check_copy_size() in copy_to_iter_mcsafe()

2018-10-03T23:59:19+00:00

commit dfb06cba8c73c0704710b2e3fbe2c35ac66a01b4 upstream. copy_to_iter_mcsafe() is passing in the is_source parameter as "false" to check_copy_size(). This is different than what copy_to_iter() does. Also, the addr parameter passed to check_copy_size() is the source so therefore we should be passing in "true" instead. Fixes: 8780356ef630 ("x86/asm/memcpy_mcsafe: Define copy_to_iter_mcsafe()") Cc: Reported-by: Fan Du Signed-off-by: Dave Jiang Reviewed-by: Vishal Verma Reported-by: Wenwei Tao Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

regulator: Fix 'do-nothing' value for regulators without suspend state

2018-10-03T23:59:19+00:00

commit 3edd79cf5a44b12dbb13bc320f5788aed6562b36 upstream. Some regulators don't have all states defined and in such cases regulator core should not assume anything. However in current implementation of of_get_regulation_constraints() DO_NOTHING_IN_SUSPEND enable value was set only for regulators which had suspend node defined, otherwise the default 0 value was used, what means DISABLE_IN_SUSPEND. This lead to broken system suspend/resume on boards, which had simple regulator constraints definition (without suspend state nodes). To avoid further mismatches between the default and uninitialized values of the suspend enabled/disabled states, change the values of the them, so default '0' means DO_NOTHING_IN_SUSPEND. Fixes: 72069f9957a1: regulator: leave one item to record whether regulator is enabled Signed-off-by: Marek Szyprowski Signed-off-by: Mark Brown Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman

bitfield: fix *_encode_bits()

2018-10-03T23:59:13+00:00

[ Upstream commit e7d4a95da86e0b048702765bbdcdc968aaf312e7 ] There's a bug in *_encode_bits() in using ~field_multiplier() for the check whether or not the constant value fits into the field, this is wrong and clearly ~field_mask() was intended. This was triggering for me for both constant and non-constant values. Additionally, make this case actually into an compile error. Declaring the extern function that will never exist with just a warning is pointless as then later we'll just get a link error. While at it, also fix the indentation in those lines I'm touching. Finally, as suggested by Andy Shevchenko, add some tests and for that introduce also u8 helpers. The tests don't compile without the fix, showing that it's necessary. Fixes: 00b0c9b82663 ("Add primitives for manipulating bitfields both in host- and fixed-endian.") Reviewed-by: Andy Shevchenko Signed-off-by: Johannes Berg Signed-off-by: Kalle Valo Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman

posix-timers: Sanitize overrun handling

2018-10-03T23:59:10+00:00

[ Upstream commit 78c9c4dfbf8c04883941445a195276bb4bb92c76 ] The posix timer overrun handling is broken because the forwarding functions can return a huge number of overruns which does not fit in an int. As a consequence timer_getoverrun(2) and siginfo::si_overrun can turn into random number generators. The k_clock::timer_forward() callbacks return a 64 bit value now. Make k_itimer::ti_overrun[_last] 64bit as well, so the kernel internal accounting is correct. 3Remove the temporary (int) casts. Add a helper function which clamps the overrun value returned to user space via timer_getoverrun(2) or siginfo::si_overrun limited to a positive value between 0 and INT_MAX. INT_MAX is an indicator for user space that the overrun value has been clamped. Reported-by: Team OWL337 Signed-off-by: Thomas Gleixner Acked-by: John Stultz Cc: Peter Zijlstra Cc: Michael Kerrisk Link: https://lkml.kernel.org/r/20180626132705.018623573@linutronix.de Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman