kernel.git/arch/arc/mm, branch linux-4.12.y

ARCv2: PAE40: set MSB even if !CONFIG_ARC_HAS_PAE40 but PAE exists in SoC

2017-08-30T08:26:32Z

commit b5ddb6d54729d814356937572d6c9b599f10c29f upstream. PAE40 confiuration in hardware extends some of the address registers for TLB/cache ops to 2 words. So far kernel was NOT setting the higher word if feature was not enabled in software which is wrong. Those need to be set to 0 in such case. Normally this would be done in the cache flush / tlb ops, however since these registers only exist conditionally, this would have to be conditional to a flag being set on boot which is expensive/ugly - specially for the more common case of PAE exists but not in use. Optimize that by zero'ing them once at boot - nobody will write to them afterwards Signed-off-by: Vineet Gupta Signed-off-by: Greg Kroah-Hartman

ARCv2: PAE40: Explicitly set MSB counterpart of SLC region ops addresses

2017-08-30T08:26:32Z

commit 7d79cee2c6540ea64dd917a14e2fd63d4ac3d3c0 upstream. It is necessary to explicitly set both SLC_AUX_RGN_START1 and SLC_AUX_RGN_END1 which hold MSB bits of the physical address correspondingly of region start and end otherwise SLC region operation is executed in unpredictable manner Without this patch, SLC flushes on HSDK (IOC disabled) were taking seconds. Reported-by: Vladimir Kondratiev Signed-off-by: Alexey Brodkin Signed-off-by: Vineet Gupta [vgupta: PAR40 regs only written if PAE40 exist] Signed-off-by: Greg Kroah-Hartman

ARCv2: SLC: Make sure busy bit is set properly for region ops

2017-08-30T08:26:32Z

commit b37174d95b0251611a80ef60abf03752e9d66d67 upstream. c70c473396cb "ARCv2: SLC: Make sure busy bit is set properly on SLC flushing" fixes problem for entire SLC operation where the problem was initially caught. But given a nature of the issue it is perfectly possible for busy bit to be read incorrectly even when region operation was started. So extending initial fix for regional operation as well. Signed-off-by: Alexey Brodkin Signed-off-by: Vineet Gupta Signed-off-by: Greg Kroah-Hartman

mm: larger stack guard gap, between vmas

2017-06-19T13:50:20Z

Stack guard page is a useful feature to reduce a risk of stack smashing into a different mapping. We have been using a single page gap which is sufficient to prevent having stack adjacent to a different mapping. But this seems to be insufficient in the light of the stack usage in userspace. E.g. glibc uses as large as 64kB alloca() in many commonly used functions. Others use constructs liks gid_t buffer[NGROUPS_MAX] which is 256kB or stack strings with MAX_ARG_STRLEN. This will become especially dangerous for suid binaries and the default no limit for the stack size limit because those applications can be tricked to consume a large portion of the stack and a single glibc call could jump over the guard page. These attacks are not theoretical, unfortunatelly. Make those attacks less probable by increasing the stack guard gap to 1MB (on systems with 4k pages; but make it depend on the page size because systems with larger base pages might cap stack allocations in the PAGE_SIZE units) which should cover larger alloca() and VLA stack allocations. It is obviously not a full fix because the problem is somehow inherent, but it should reduce attack space a lot. One could argue that the gap size should be configurable from userspace, but that can be done later when somebody finds that the new 1MB is wrong for some special case applications. For now, add a kernel command line option (stack_guard_gap) to specify the stack gap size (in page units). Implementation wise, first delete all the old code for stack guard page: because although we could get away with accounting one extra page in a stack vma, accounting a larger gap can break userspace - case in point, a program run with "ulimit -S -v 20000" failed when the 1MB gap was counted for RLIMIT_AS; similar problems could come with RLIMIT_MLOCK and strict non-overcommit mode. Instead of keeping gap inside the stack vma, maintain the stack guard gap as a gap between vmas: using vm_start_gap() in place of vm_start (or vm_end_gap() in place of vm_end if VM_GROWSUP) in just those few places which need to respect the gap - mainly arch_get_unmapped_area(), and and the vma tree's subtree_gap support for that. Original-patch-by: Oleg Nesterov Original-patch-by: Michal Hocko Signed-off-by: Hugh Dickins Acked-by: Michal Hocko Tested-by: Helge Deller # parisc Signed-off-by: Linus Torvalds

Merge tag 'arc-4.12-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc

2017-05-09T17:10:15Z

Pull ARC updates from Vineet Gupta: - AXS10x platform clk updates for I2S, PGU - add region based cache flush operation for ARCv2 cores - enforce PAE40 dependency on HIGHMEM - ptrace support for additional regs in ARCv2 cores - fix build failure in linux-next dut to a header include ordering change * tag 'arc-4.12-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc: Revert "ARCv2: Allow enabling PAE40 w/o HIGHMEM" ARC: mm: fix build failure in linux-next for UP builds ARCv2: ptrace: provide regset for accumulator/r30 regs elf: Add ARCv2 specific core note section ARCv2: mm: micro-optimize region flush generated code ARCv2: mm: Merge 2 updates to DC_CTRL for region flush ARCv2: mm: Implement cache region flush operations ARC: mm: Move full_page computation into cache version agnostic wrapper arc: axs10x: Fix ARC PGU default clock frequency arc: axs10x: Add DT bindings for I2S audio playback

ARCv2: mm: Merge 2 updates to DC_CTRL for region flush

2017-05-02T23:16:07Z

ARCv2: mm: Implement cache region flush operations

2017-05-02T22:57:22Z

These are more efficient than the per-line ops Signed-off-by: Vineet Gupta

ARC: mm: Move full_page computation into cache version agnostic wrapper

2017-05-02T22:19:34Z

This reduces code duplication in each of cache version specific handlers Signed-off-by: Vineet Gupta

Merge branch 'work.uaccess' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

2017-05-01T21:41:04Z

Pull uaccess unification updates from Al Viro: "This is the uaccess unification pile. It's _not_ the end of uaccess work, but the next batch of that will go into the next cycle. This one mostly takes copy_from_user() and friends out of arch/* and gets the zero-padding behaviour in sync for all architectures. Dealing with the nocache/writethrough mess is for the next cycle; fortunately, that's x86-only. Same for cleanups in iov_iter.c (I am sold on access_ok() in there, BTW; just not in this pile), same for reducing __copy_... callsites, strn*... stuff, etc. - there will be a pile about as large as this one in the next merge window. This one sat in -next for weeks. -3KLoC" * 'work.uaccess' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (96 commits) HAVE_ARCH_HARDENED_USERCOPY is unconditional now CONFIG_ARCH_HAS_RAW_COPY_USER is unconditional now m32r: switch to RAW_COPY_USER hexagon: switch to RAW_COPY_USER microblaze: switch to RAW_COPY_USER get rid of padding, switch to RAW_COPY_USER ia64: get rid of copy_in_user() ia64: sanitize __access_ok() ia64: get rid of 'segment' argument of __do_{get,put}_user() ia64: get rid of 'segment' argument of __{get,put}_user_check() ia64: add extable.h powerpc: get rid of zeroing, switch to RAW_COPY_USER esas2r: don't open-code memdup_user() alpha: fix stack smashing in old_adjtimex(2) don't open-code kernel_setsockopt() mips: switch to RAW_COPY_USER mips: get rid of tail-zeroing in primitives mips: make copy_from_user() zero tail explicitly mips: clean and reorder the forest of macros... mips: consolidate __invoke_... wrappers ...

ARCv2: SLC: Make sure busy bit is set properly on SLC flushing

2017-03-31T00:32:33Z

As reported in STAR 9001165532, an SLC control reg read (for checking busy state) right after SLC invalidate command may incorrectly return NOT busy causing software to NOT spin-wait while operation is underway. (and for some reason this only happens if L1 cache is also disabled - as required by IOC programming model) Suggested workaround is to do an additional Control Reg read, which ensures the 2nd read gets the right status. Cc: stable@vger.kernel.org #4.10 Signed-off-by: Alexey Brodkin [vgupta: reworte changelog a bit] Signed-off-by: Vineet Gupta