linux-kernel - Re: arm64: perf test 26 rpi4 oops

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <b26b34cf503ad570a3b40c72f312482921d7104f.camel@gmx.de>
Date:   Mon, 31 Jul 2023 13:37:04 +0200
From:   Mike Galbraith <efault@....de>
To:     Will Deacon <will@...nel.org>
Cc:     lkml <linux-kernel@...r.kernel.org>,
        Mark Rutland <mark.rutland@....com>, lstoakes@...il.com,
        wangkefeng.wang@...wei.com, catalin.marinas@....com,
        ardb@...nel.org
Subject: Re: arm64: perf test 26 rpi4 oops

On Mon, 2023-07-31 at 11:43 +0100, Will Deacon wrote:


> The easiest way to trigger the problem is simply:
>
> # objdump -d /proc/kcore
>
> Looking at the history, I wonder whether this is because of a combination
> of:
>
> e025ab842ec3 ("mm: remove kern_addr_valid() completely")
>
> which removed the kern_addr_valid() check on the basis that kcore used
> copy_from_kernel_nofault() anyway, and:
>
> 2e1c0170771e ("fs/proc/kcore: avoid bounce buffer for ktext data")
>
> which replaced the copy_from_kernel_nofault() with _copy_to_user().
>
> So with both of those applied, we're missing the address check on arm64.

Hm, I see trace_printk() in the future of self and x86_64 desktop box.

While my attempt to bisect the two arm64 allocation failures was doomed
by an apparent inability to reliably type sudo perf test 26, I still
suspect they might be a part of this fossilized bug.. see last two with
funny looking addr/len.

dmesg|tail -2
[ 1506.732999] __vm_enough_memory: pid: 3824, comm: objdump, not enough memory for the allocation of 64913409 pages
[ 1506.733050] __vm_enough_memory: pid: 3824, comm: objdump, not enough memory for the allocation of 64913441 pages
cat trace|grep 3824
...
objdump-3824    [000] .......  1506.708497: vm_unmapped_area: addr=0x7fa9ac8000 err=0 total_vm=0x52f flags=0x1 len=0x2e7000 lo=0x1000 hi=0x7faa24e000 mask=0x0 ofs=0x0
objdump-3824    [000] .......  1506.708657: vm_unmapped_area: addr=0x7faa213000 err=0 total_vm=0x816 flags=0x1 len=0x7000 lo=0x1000 hi=0x7faa24e000 mask=0x0 ofs=0x0
objdump-3824    [000] .......  1506.709740: vm_unmapped_area: addr=0x7fa86f7000 err=0 total_vm=0x81d flags=0x1 len=0x13d1000 lo=0x1000 hi=0x7faa24e000 mask=0x0 ofs=0x0
objdump-3824    [000] .......  1506.732991: vm_unmapped_area: addr=0x17995df000 err=0 total_vm=0x831 flags=0x1 len=0x3de8001000 lo=0x1000 hi=0x7faa24e000 mask=0x0 ofs=0x0
objdump-3824    [000] .......  1506.733049: vm_unmapped_area: addr=0x17995bf000 err=0 total_vm=0x831 flags=0x1 len=0x3de8021000 lo=0x1000 hi=0x7faa24e000 mask=0x0 ofs=0x0
objdump-3824    [000] .......  1506.733846: exit_mmap: mt_mod ffffff8149991980, DESTROY