commit 02b670c1f88e78f42a6c5aee155c7b26960ca054 upstream.
The syzbot-reported stack trace from hell in this discussion thread
actually has three nested page faults:
https://lore.kernel.org/r/000000000000d5f4fc0616e816d4@google.com
... and I think that's actually the important thing here:
- the first page fault is from user space, and triggers the vsyscall
emulation.
- the second page fault is from __do_sys_gettimeofday(), and that should
just have caused the exception that then sets the return value to
-EFAULT
- the third nested page fault is due to _raw_spin_unlock_irqrestore() ->
preempt_schedule() -> trace_sched_switch(), which then causes a BPF
trace program to run, which does that bpf_probe_read_compat(), which
causes that page fault under pagefault_disable().
It's quite the nasty backtrace, and there's a lot going on.
The problem is literally the vsyscall emulation, which sets
current->thread.sig_on_uaccess_err = 1;
and that causes the fixup_exception() code to send the signal *despite* the
exception being caught.
And I think that is in fact completely bogus. It's completely bogus
exactly because it sends that signal even when it *shouldn't* be sent -
like for the BPF user mode trace gathering.
In other words, I think the whole "sig_on_uaccess_err" thing is entirely
broken, because it makes any nested page-faults do all the wrong things.
Now, arguably, I don't think anybody should enable vsyscall emulation any
more, but this test case clearly does.
I think we should just make the "send SIGSEGV" be something that the
vsyscall emulation does on its own, not this broken per-thread state for
something that isn't actually per thread.
The x86 page fault code actually tried to deal with the "incorrect nesting"
by having that:
if (in_interrupt())
return;
which ignores the sig_on_uaccess_err case when it happens in interrupts,
but as shown by this example, these nested page faults do not need to be
about interrupts at all.
IOW, I think the only right thing is to remove that horrendously broken
code.
The attached patch looks like the ObviouslyCorrect(tm) thing to do.
NOTE! This broken code goes back to this commit in 2011:
4fc3490114bb ("x86-64: Set siginfo and context on vsyscall emulation faults")
... and back then the reason was to get all the siginfo details right.
Honestly, I do not for a moment believe that it's worth getting the siginfo
details right here, but part of the commit says:
This fixes issues with UML when vsyscall=emulate.
... and so my patch to remove this garbage will probably break UML in this
situation.
I do not believe that anybody should be running with vsyscall=emulate in
2024 in the first place, much less if you are doing things like UML. But
let's see if somebody screams.
Reported-and-tested-by: syzbot+83e7f982ca045ab4405c@syzkaller.appspotmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Andy Lutomirski <luto@kernel.org>
Link: https://lore.kernel.org/r/CAHk-=wh9D6f7HUkDgZHKmDCHUQmp+Co89GP+b8+z+G56BKeyNg@mail.gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[gpiccoli: Backport the patch due to differences in the trees. The main change
between 5.10.y and 5.15.y is due to renaming the fixup function, by
commit 6456a2a69ee1 ("x86/fault: Rename no_context() to kernelmode_fixup_or_oops()").
Following 2 commits cause divergence in the diffs too (in the removed lines):
cd072dab453a ("x86/fault: Add a helper function to sanitize error code")
d4ffd5df9d18 ("x86/fault: Fix wrong signal when vsyscall fails with pkey")
Finally, there is context adjustment in the processor.h file.]
Signed-off-by: Guilherme G. Piccoli <gpiccoli@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 3fb0fdb3bbe7aed495109b3296b06c2409734023 ]
On 32-bit kernels, the stackprotector canary is quite nasty -- it is
stored at %gs:(20), which is nasty because 32-bit kernels use %fs for
percpu storage. It's even nastier because it means that whether %gs
contains userspace state or kernel state while running kernel code
depends on whether stackprotector is enabled (this is
CONFIG_X86_32_LAZY_GS), and this setting radically changes the way
that segment selectors work. Supporting both variants is a
maintenance and testing mess.
Merely rearranging so that percpu and the stack canary
share the same segment would be messy as the 32-bit percpu address
layout isn't currently compatible with putting a variable at a fixed
offset.
Fortunately, GCC 8.1 added options that allow the stack canary to be
accessed as %fs:__stack_chk_guard, effectively turning it into an ordinary
percpu variable. This lets us get rid of all of the code to manage the
stack canary GDT descriptor and the CONFIG_X86_32_LAZY_GS mess.
(That name is special. We could use any symbol we want for the
%fs-relative mode, but for CONFIG_SMP=n, gcc refuses to let us use any
name other than __stack_chk_guard.)
Forcibly disable stackprotector on older compilers that don't support
the new options and turn the stack canary into a percpu variable. The
"lazy GS" approach is now used for all 32-bit configurations.
Also makes load_gs_index() work on 32-bit kernels. On 64-bit kernels,
it loads the GS selector and updates the user GSBASE accordingly. (This
is unchanged.) On 32-bit kernels, it loads the GS selector and updates
GSBASE, which is now always the user base. This means that the overall
effect is the same on 32-bit and 64-bit, which avoids some ifdeffery.
[ bp: Massage commit message. ]
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/c0ff7dba14041c7e5d1cae5d4df052f03759bef3.1613243844.git.luto@kernel.org
Stable-dep-of: e3f269ed0acc ("x86/pm: Work around false positive kmemleak report in msr_build_context()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
