On Mon, Apr 13, 2020 at 04:51:25PM +0530, Naresh Kamboju wrote:
On Mon, 13 Apr 2020 at 03:12, Peter Xu peterx@redhat.com wrote:
And since this one is very easy to reproduce, I finally noticed that we have wrongly enabled uffd-wp on x86_32, which is definely not going to work... Because we'll use bit 2 of swap entry assuming that's the uffd-wp bit, while that's part of swp offset on 32bit systems.
Naresh, could you try whether below change fixes the issue for you?
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 8d078642b4be..7ac524d1316e 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -149,7 +149,7 @@ config X86 select HAVE_ARCH_TRACEHOOK select HAVE_ARCH_TRANSPARENT_HUGEPAGE select HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD if X86_64
select HAVE_ARCH_USERFAULTFD_WP if USERFAULTFD
select HAVE_ARCH_USERFAULTFD_WP if X86_64 && USERFAULTFD select HAVE_ARCH_VMAP_STACK if X86_64 select HAVE_ARCH_WITHIN_STACK_FRAMES select HAVE_ASM_MODVERSIONSTested-by: Naresh Kamboju naresh.kamboju@linaro.org
The above patch tested and did not notice the crash.
- cd /opt/kselftests/default-in-kernel/vm/
- ./compaction_test
No of huge pages allocated = 297
- ./compaction_test
[ 46.059785] kauditd_printk_skb: 15 callbacks suppressed [ 46.059786] audit: type=1334 audit(1586776582.927:25): prog-id=17 op=UNLOAD [ 46.071997] audit: type=1334 audit(1586776582.927:26): prog-id=16 op=UNLOAD No of huge pages allocated = 160
Full test log, https://lkft.validation.linaro.org/scheduler/job/1362495#L1308
Build artifacts. https://builds.tuxbuild.com/w0ol7cCsGan0wzPp7bNqkg/
Thanks! I'll post a formal patch soon.