kernel.git/arch/arm/mm, branch linux-3.18.y

ARM: 8814/1: mm: improve/fix ARM v7_dma_inv_range() unaligned address handling

2018-12-21T13:08:49Z

[ Upstream commit a1208f6a822ac29933e772ef1f637c5d67838da9 ] This patch addresses possible memory corruption when v7_dma_inv_range(start_address, end_address) address parameters are not aligned to whole cache lines. This function issues "invalidate" cache management operations to all cache lines from start_address (inclusive) to end_address (exclusive). When start_address and/or end_address are not aligned, the start and/or end cache lines are first issued "clean & invalidate" operation. The assumption is this is done to ensure that any dirty data addresses outside the address range (but part of the first or last cache lines) are cleaned/flushed so that data is not lost, which could happen if just an invalidate is issued. The problem is that these first/last partial cache lines are issued "clean & invalidate" and then "invalidate". This second "invalidate" is not required and worse can cause "lost" writes to addresses outside the address range but part of the cache line. If another component writes to its part of the cache line between the "clean & invalidate" and "invalidate" operations, the write can get lost. This fix is to remove the extra "invalidate" operation when unaligned addressed are used. A kernel module is available that has a stress test to reproduce the issue and a unit test of the updated v7_dma_inv_range(). It can be downloaded from http://ftp.sageembedded.com/outgoing/linux/cache-test-20181107.tgz. v7_dma_inv_range() is call by dmac_[un]map_area(addr, len, direction) when the direction is DMA_FROM_DEVICE. One can (I believe) successfully argue that DMA from a device to main memory should use buffers aligned to cache line size, because the "clean & invalidate" might overwrite data that the device just wrote using DMA. But if a driver does use unaligned buffers, at least this fix will prevent memory corruption outside the buffer. Signed-off-by: Chris Cole Signed-off-by: Russell King Signed-off-by: Sasha Levin

ARM: 8799/1: mm: fix pci_ioremap_io() offset check

2018-11-10T15:39:12Z

[ Upstream commit 3a58ac65e2d7969bcdf1b6acb70fa4d12a88e53e ] IO_SPACE_LIMIT is the ending address of the PCI IO space, i.e something like 0xfffff (and not 0x100000). Therefore, when offset = 0xf0000 is passed as argument, this function fails even though the offset + SZ_64K fits below the IO_SPACE_LIMIT. This makes the last chunk of 64 KB of the I/O space not usable as it cannot be mapped. This patch fixes that by substracing 1 to offset + SZ_64K, so that we compare the addrss of the last byte of the I/O space against IO_SPACE_LIMIT instead of the address of the first byte of what is after the I/O space. Fixes: c2794437091a4 ("ARM: Add fixed PCI i/o mapping") Signed-off-by: Thomas Petazzoni Acked-by: Nicolas Pitre Signed-off-by: Russell King Signed-off-by: Sasha Levin

ARM: dma-mapping: disallow dma_get_sgtable() for non-kernel managed memory

2017-12-25T13:20:08Z

[ Upstream commit 916a008b4b8ecc02fbd035cfb133773dba1ff3d7 ] dma_get_sgtable() tries to create a scatterlist table containing valid struct page pointers for the coherent memory allocation passed in to it. However, memory can be declared via dma_declare_coherent_memory(), or via other reservation schemes which means that coherent memory is not guaranteed to be backed by struct pages. In such cases, the resulting scatterlist table contains pointers to invalid pages, which causes kernel oops later. This patch adds detection of such memory, and refuses to create a scatterlist table for such memory. Reported-by: Shuah Khan Signed-off-by: Russell King Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman

ARM: 8721/1: mm: dump: check hardware RO bit for LPAE

2017-11-30T08:35:49Z

commit 3b0c0c922ff4be275a8beb87ce5657d16f355b54 upstream. When CONFIG_ARM_LPAE is set, the PMD dump relies on the software read-only bit to determine whether a page is writable. This concealed a bug which left the kernel text section writable (AP2=0) while marked read-only in the software bit. In a kernel with the AP2 bug, the dump looks like this: ---[ Kernel Mapping ]--- 0xc0000000-0xc0200000 2M RW NX SHD 0xc0200000-0xc0600000 4M ro x SHD 0xc0600000-0xc0800000 2M ro NX SHD 0xc0800000-0xc4800000 64M RW NX SHD The fix is to check that the software and hardware bits are both set before displaying "ro". The dump then shows the true perms: ---[ Kernel Mapping ]--- 0xc0000000-0xc0200000 2M RW NX SHD 0xc0200000-0xc0600000 4M RW x SHD 0xc0600000-0xc0800000 2M RW NX SHD 0xc0800000-0xc4800000 64M RW NX SHD Fixes: ded947798469 ("ARM: 8109/1: mm: Modify pte_write and pmd_write logic for LPAE") Signed-off-by: Philip Derrin Tested-by: Neil Dick Reviewed-by: Kees Cook Signed-off-by: Russell King Signed-off-by: Greg Kroah-Hartman

ARM: 8692/1: mm: abort uaccess retries upon fatal signal

2017-09-13T21:03:47Z

commit 746a272e44141af24a02f6c9b0f65f4c4598ed42 upstream. When there's a fatal signal pending, arm's do_page_fault() implementation returns 0. The intent is that we'll return to the faulting userspace instruction, delivering the signal on the way. However, if we take a fatal signal during fixing up a uaccess, this results in a return to the faulting kernel instruction, which will be instantly retried, resulting in the same fault being taken forever. As the task never reaches userspace, the signal is not delivered, and the task is left unkillable. While the task is stuck in this state, it can inhibit the forward progress of the system. To avoid this, we must ensure that when a fatal signal is pending, we apply any necessary fixup for a faulting kernel instruction. Thus we will return to an error path, and it is up to that code to make forward progress towards delivering the fatal signal. Signed-off-by: Mark Rutland Reviewed-by: Steve Capper Signed-off-by: Russell King Signed-off-by: Greg Kroah-Hartman

ARM: 8685/1: ensure memblock-limit is pmd-aligned

2017-07-05T12:35:14Z

commit 9e25ebfe56ece7541cd10a20d715cbdd148a2e06 upstream. The pmd containing memblock_limit is cleared by prepare_page_table() which creates the opportunity for early_alloc() to allocate unmapped memory if memblock_limit is not pmd aligned causing a boot-time hang. Commit 965278dcb8ab ("ARM: 8356/1: mm: handle non-pmd-aligned end of RAM") attempted to resolve this problem, but there is a path through the adjust_lowmem_bounds() routine where if all memory regions start and end on pmd-aligned addresses the memblock_limit will be set to arm_lowmem_limit. Since arm_lowmem_limit can be affected by the vmalloc early parameter, the value of arm_lowmem_limit may not be pmd-aligned. This commit corrects this oversight such that memblock_limit is always rounded down to pmd-alignment. Fixes: 965278dcb8ab ("ARM: 8356/1: mm: handle non-pmd-aligned end of RAM") Signed-off-by: Doug Berger Suggested-by: Mark Rutland Signed-off-by: Russell King Signed-off-by: Greg Kroah-Hartman

mm: larger stack guard gap, between vmas

2017-06-26T05:10:54Z

commit 1be7107fbe18eed3e319a6c3e83c78254b693acb upstream. 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 [wt: backport to 4.11: adjust context] [wt: backport to 4.9: adjust context ; kernel doc was not in admin-guide] [wt: backport to 4.4: adjust context ; drop ppc hugetlb_radix changes] [wt: backport to 3.18: adjust context ; no FOLL_POPULATE ; s390 uses generic arch_get_unmapped_area()] Signed-off-by: Willy Tarreau [gkh: minor build fixes for 3.18] Signed-off-by: Greg Kroah-Hartman

ARM: 8296/1: cache-l2x0: clean up aurora cache handling

2017-04-30T03:49:17Z

commit 20e783e39e55c2615fb61d1b3d139ee9edcf6772 upstream. The aurora cache controller is the only remaining user of a couple of functions in this file and are completely unused when that is disabled, leading to build warnings: arch/arm/mm/cache-l2x0.c:167:13: warning: 'l2x0_cache_sync' defined but not used [-Wunused-function] arch/arm/mm/cache-l2x0.c:184:13: warning: 'l2x0_flush_all' defined but not used [-Wunused-function] arch/arm/mm/cache-l2x0.c:194:13: warning: 'l2x0_disable' defined but not used [-Wunused-function] With the knowledge that the code is now aurora-specific, we can simplify it noticeably: - The pl310 errata workarounds are not needed on aurora and can be removed - As confirmed by Thomas Petazzoni from the data sheet, the cache_wait() macro is never needed. - No need to hold the lock across atomic cache sync - We can load the l2x0_base into a local variable across operations There should be no functional change in this patch, but readability and the generated object code improves, along with avoiding the warnings. (on Armada 370 RD and Armada XP GP, boot tested, plus a little bit of DMA traffic by reading data from a SD card) Acked-by: Thomas Petazzoni Tested-by: Thomas Petazzoni Signed-off-by: Arnd Bergmann Signed-off-by: Russell King Signed-off-by: Greg Kroah-Hartman

ARM: 8356/1: mm: handle non-pmd-aligned end of RAM

2015-06-10T17:42:31Z

[ Upstream commit 965278dcb8ab0b1f666cc47937933c4be4aea48d ] At boot time we round the memblock limit down to section size in an attempt to ensure that we will have mapped this RAM with section mappings prior to allocating from it. When mapping RAM we iterate over PMD-sized chunks, creating these section mappings. Section mappings are only created when the end of a chunk is aligned to section size. Unfortunately, with classic page tables (where PMD_SIZE is 2 * SECTION_SIZE) this means that if a chunk is between 1M and 2M in size the first 1M will not be mapped despite having been accounted for in the memblock limit. This has been observed to result in page tables being allocated from unmapped memory, causing boot-time hangs. This patch modifies the memblock limit rounding to always round down to PMD_SIZE instead of SECTION_SIZE. For classic MMU this means that we will round the memblock limit down to a 2M boundary, matching the limits on section mappings, and preventing allocations from unmapped memory. For LPAE there should be no change as PMD_SIZE == SECTION_SIZE. Signed-off-by: Mark Rutland Reported-by: Stefan Agner Tested-by: Stefan Agner Acked-by: Laura Abbott Tested-by: Hans de Goede Cc: Catalin Marinas Cc: Steve Capper Cc: stable@vger.kernel.org Signed-off-by: Russell King Signed-off-by: Sasha Levin

mm/hugetlb: use pmd_page() in follow_huge_pmd()

2015-05-17T23:12:31Z

[ Upstream commit 97534127012f0e396eddea4691f4c9b170aed74b ] Commit 61f77eda9bbf ("mm/hugetlb: reduce arch dependent code around follow_huge_*") broke follow_huge_pmd() on s390, where pmd and pte layout differ and using pte_page() on a huge pmd will return wrong results. Using pmd_page() instead fixes this. All architectures that were touched by that commit have pmd_page() defined, so this should not break anything on other architectures. Fixes: 61f77eda "mm/hugetlb: reduce arch dependent code around follow_huge_*" Signed-off-by: Gerald Schaefer Acked-by: Naoya Horiguchi Cc: Hugh Dickins Cc: Michal Hocko , Andrea Arcangeli Cc: Martin Schwidefsky Acked-by: David Rientjes Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Sasha Levin