Ben Tebulin reported:
"Since v3.7.2 on two independent machines a very specific Git
repository fails in 9/10 cases on git-fsck due to an SHA1/memory
failures. This only occurs on a very specific repository and can be
reproduced stably on two independent laptops. Git mailing list ran
out of ideas and for me this looks like some very exotic kernel issue"
and bisected the failure to the backport of commit 53a59fc67f ("mm:
limit mmu_gather batching to fix soft lockups on !CONFIG_PREEMPT").
That commit itself is not actually buggy, but what it does is to make it
much more likely to hit the partial TLB invalidation case, since it
introduces a new case in tlb_next_batch() that previously only ever
happened when running out of memory.
The real bug is that the TLB gather virtual memory range setup is subtly
buggered. It was introduced in commit 597e1c3580 ("mm/mmu_gather:
enable tlb flush range in generic mmu_gather"), and the range handling
was already fixed at least once in commit e6c495a96c ("mm: fix the TLB
range flushed when __tlb_remove_page() runs out of slots"), but that fix
was not complete.
The problem with the TLB gather virtual address range is that it isn't
set up by the initial tlb_gather_mmu() initialization (which didn't get
the TLB range information), but it is set up ad-hoc later by the
functions that actually flush the TLB. And so any such case that forgot
to update the TLB range entries would potentially miss TLB invalidates.
Rather than try to figure out exactly which particular ad-hoc range
setup was missing (I personally suspect it's the hugetlb case in
zap_huge_pmd(), which didn't have the same logic as zap_pte_range()
did), this patch just gets rid of the problem at the source: make the
TLB range information available to tlb_gather_mmu(), and initialize it
when initializing all the other tlb gather fields.
This makes the patch larger, but conceptually much simpler. And the end
result is much more understandable; even if you want to play games with
partial ranges when invalidating the TLB contents in chunks, now the
range information is always there, and anybody who doesn't want to
bother with it won't introduce subtle bugs.
Ben verified that this fixes his problem.
Reported-bisected-and-tested-by: Ben Tebulin <tebulin@googlemail.com>
Build-testing-by: Stephen Rothwell <sfr@canb.auug.org.au>
Build-testing-by: Richard Weinberger <richard.weinberger@gmail.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since the introduction of preemptible mmu_gather TLB fast mode has been
broken. TLB fast mode relies on there being absolutely no concurrency;
it frees pages first and invalidates TLBs later.
However now we can get concurrency and stuff goes *bang*.
This patch removes all tlb_fast_mode() code; it was found the better
option vs trying to patch the hole by entangling tlb invalidation with
the scheduler.
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tony Luck <tony.luck@intel.com>
Reported-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The patch adds support for THP (transparent huge pages) to LPAE
systems. When this feature is enabled, the kernel tries to map
anonymous pages as 2MB sections where possible.
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
[steve.capper@linaro.org: symbolic constants used, value of
PMD_SECT_SPLITTING adjusted, tlbflush.h included in pgtable.h,
added PROT_NONE support.]
Signed-off-by: Steve Capper <steve.capper@linaro.org>
Reviewed-by: Will Deacon <will.deacon@arm.com>
LPAE does not use two pmd entries for a pte, so the additional tlb
flushing is not required.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Linux uses two PMD entries for a PTE with the classic page table format,
covering 2MB range. However, the __pte_free_tlb() function only adds a
single TLB flush corresponding to 1MB range covering 'addr'. On
Cortex-A15, level 1 entries can be cached by the TLB independently of
the level 2 entries and without additional flushing a PMD entry would be
left pointing at the wrong PTE. The patch limits the TLB flushing range
to two 4KB pages around the 1MB boundary within PMD.
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Similar to the PTE freeing, this patch introduced __pmd_free_tlb() which
invalidates the TLB before freeing a PMD page. This is needed because on
newer processors the entry in the upper page table may be cached by the
TLB and point to random data after the PMD has been freed.
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Nick Piggin noted upon introducing 4level-fixup.h:
| Add a temporary "fallback" header so architectures can run with
| the 4level pagetables patch without modification. All architectures
| should be converted to use the folding headers (include/asm-generic/
| pgtable-nop?d.h) as soon as possible, and the fallback header removed.
This makes ARM compliant with this statement.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Fix up the arm mmu_gather code to conform to the new API.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Miller <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Tony Luck <tony.luck@intel.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Nick Piggin <npiggin@kernel.dk>
Cc: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There's no need to noMMU to put tlb_flush() in asm/tlbflush.h - it's
part of the tlb shootdown interface. Move it to asm/tlb.h instead, as
per x86.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
We need to delay freeing any mapped page on SMP and ARMv7 systems to
ensure that the data is not accessed by other CPUs, or is used for
speculative prefetch with ARMv7. This includes not only mapped pages
but also pages used for the page tables themselves.
This avoids races with the MMU/other CPUs accessing pages after they've
been freed but before we've invalidated the TLB.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
mm: Pass virtual address to [__]p{te,ud,md}_free_tlb()
Upcoming paches to support the new 64-bit "BookE" powerpc architecture
will need to have the virtual address corresponding to PTE page when
freeing it, due to the way the HW table walker works.
Basically, the TLB can be loaded with "large" pages that cover the whole
virtual space (well, sort-of, half of it actually) represented by a PTE
page, and which contain an "indirect" bit indicating that this TLB entry
RPN points to an array of PTEs from which the TLB can then create direct
entries. Thus, in order to invalidate those when PTE pages are deleted,
we need the virtual address to pass to tlbilx or tlbivax instructions.
The old trick of sticking it somewhere in the PTE page struct page sucks
too much, the address is almost readily available in all call sites and
almost everybody implemets these as macros, so we may as well add the
argument everywhere. I added it to the pmd and pud variants for consistency.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: David Howells <dhowells@redhat.com> [MN10300 & FRV]
Acked-by: Nick Piggin <npiggin@suse.de>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com> [s390]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When unmapping N pages (e.g. shared memory) the amount of TLB flushes
done can be (N*PAGE_SIZE/ZAP_BLOCK_SIZE)*N although it should be N at
maximum. With PREEMPT kernel ZAP_BLOCK_SIZE is 8 pages, so there is a
noticeable performance penalty when unmapping a large VMA and the system
is spending its time in flush_tlb_range().
The problem is that tlb_end_vma() is always flushing the full VMA
range. The subrange that needs to be flushed can be calculated by
tlb_remove_tlb_entry(). This approach was suggested by Hugh Dickins,
and is also used by other arches.
The speed increase is roughly 3x for 8M mappings and for larger mappings
even more.
Signed-off-by: Aaro Koskinen <Aaro.Koskinen@nokia.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Move platform independent header files to arch/arm/include/asm, leaving
those in asm/arch* and asm/plat* alone.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Linus Torvalds
Russell King
Peter Zijlstra
Catalin Marinas
Will Deacon
Peter Zijlstra
Benjamin Herrenschmidt
Aaro Koskinen
Russell King