selftests: ptrace: peeksiginfo failed on x86_64, i386, arm64 and arm. FAILED on stable rc branches 4.19, 4.14, 4.9 and 4.4. PASS on mainline, next and 5.1 stable rc branch.
Test output: ------------------ cd /opt/kselftests/mainline/ptrace ./peeksiginfo Error (peeksiginfo.c:143): Only 0 signals were read
The git bisect show that below commit caused this test to fail.
git bisect bad 5b6b0eac235ef1f915f24eda6d501a754022cbf0 is the first bad commit commit 5b6b0eac235ef1f915f24eda6d501a754022cbf0 Author: Eric W. Biederman ebiederm@xmission.com Date: Tue May 28 18:46:37 2019 -0500
signal/ptrace: Don't leak unitialized kernel memory with PTRACE_PEEK_SIGINFO
commit f6e2aa91a46d2bc79fce9b93a988dbe7655c90c0 upstream.
Recently syzbot in conjunction with KMSAN reported that ptrace_peek_siginfo can copy an uninitialized siginfo to userspace. Inspecting ptrace_peek_siginfo confirms this.
The problem is that off when initialized from args.off can be initialized to a negaive value. At which point the "if (off >= 0)" test to see if off became negative fails because off started off negative.
Prevent the core problem by adding a variable found that is only true if a siginfo is found and copied to a temporary in preparation for being copied to userspace.
Prevent args.off from being truncated when being assigned to off by testing that off is <= the maximum possible value of off. Convert off to an unsigned long so that we should not have to truncate args.off, we have well defined overflow behavior so if we add another check we won't risk fighting undefined compiler behavior, and so that we have a type whose maximum value is easy to test for.
Cc: Andrei Vagin avagin@gmail.com Cc: stable@vger.kernel.org Reported-by: syzbot+0d602a1b0d8c95bdf299@syzkaller.appspotmail.com Fixes: 84c751bd4aeb ("ptrace: add ability to retrieve signals without removing from a queue (v4)") Signed-off-by: "Eric W. Biederman" ebiederm@xmission.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org
:040000 040000 ff9f3109f210274d0b87851d226c35e7305ce44a b36de2c855fe2a0b332f145f0966dc1a0304d4bd M kernel
Test case link, https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/tool...
Test output log link, https://lkft.validation.linaro.org/scheduler/job/777223#L1084
Test results comparison on different branches, https://qa-reports.linaro.org/_/comparetest/?project=22&project=6&pr...
Best regards Naresh Kamboju
Naresh Kamboju naresh.kamboju@linaro.org writes:
selftests: ptrace: peeksiginfo failed on x86_64, i386, arm64 and arm. FAILED on stable rc branches 4.19, 4.14, 4.9 and 4.4. PASS on mainline, next and 5.1 stable rc branch.
Greg.
Looking in my email it appears 4.19, 4.14, 4.9, and 4.4 patches are missing the "found=1" line from the original change. This explains the test failure.
Can you handle this?
Thanks, Eric
Test output:
cd /opt/kselftests/mainline/ptrace ./peeksiginfo Error (peeksiginfo.c:143): Only 0 signals were read
The git bisect show that below commit caused this test to fail.
git bisect bad 5b6b0eac235ef1f915f24eda6d501a754022cbf0 is the first bad commit commit 5b6b0eac235ef1f915f24eda6d501a754022cbf0 Author: Eric W. Biederman ebiederm@xmission.com Date: Tue May 28 18:46:37 2019 -0500
signal/ptrace: Don't leak unitialized kernel memory with PTRACE_PEEK_SIGINFO commit f6e2aa91a46d2bc79fce9b93a988dbe7655c90c0 upstream. Recently syzbot in conjunction with KMSAN reported that ptrace_peek_siginfo can copy an uninitialized siginfo to userspace. Inspecting ptrace_peek_siginfo confirms this. The problem is that off when initialized from args.off can be initialized to a negaive value. At which point the "if (off >= 0)" test to see if off became negative fails because off started off negative. Prevent the core problem by adding a variable found that is only true if a siginfo is found and copied to a temporary in preparation for being copied to userspace. Prevent args.off from being truncated when being assigned to off by testing that off is <= the maximum possible value of off. Convert off to an unsigned long so that we should not have to truncate args.off, we have well defined overflow behavior so if we add another check we won't risk fighting undefined compiler behavior, and so that we have a type whose maximum value is easy to test for. Cc: Andrei Vagin <avagin@gmail.com> Cc: stable@vger.kernel.org Reported-by: syzbot+0d602a1b0d8c95bdf299@syzkaller.appspotmail.com Fixes: 84c751bd4aeb ("ptrace: add ability to retrieve signalswithout removing from a queue (v4)") Signed-off-by: "Eric W. Biederman" ebiederm@xmission.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org
:040000 040000 ff9f3109f210274d0b87851d226c35e7305ce44a b36de2c855fe2a0b332f145f0966dc1a0304d4bd M kernel
Test case link, https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/tool...
Test output log link, https://lkft.validation.linaro.org/scheduler/job/777223#L1084
Test results comparison on different branches, https://qa-reports.linaro.org/_/comparetest/?project=22&project=6&pr...
Best regards Naresh Kamboju
On Mon, Jun 17, 2019 at 08:13:56AM -0500, Eric W. Biederman wrote:
Naresh Kamboju naresh.kamboju@linaro.org writes:
selftests: ptrace: peeksiginfo failed on x86_64, i386, arm64 and arm. FAILED on stable rc branches 4.19, 4.14, 4.9 and 4.4. PASS on mainline, next and 5.1 stable rc branch.
Greg.
Looking in my email it appears 4.19, 4.14, 4.9, and 4.4 patches are missing the "found=1" line from the original change. This explains the test failure.
That's an interesting one.
Thanks for the report Naresh and the quick analysis Eric.
Can you handle this?
I've fixed it up in the queue, thank you!
-- Thanks, Sasha
On Mon, Jun 17, 2019 at 08:13:56AM -0500, Eric W. Biederman wrote:
Naresh Kamboju naresh.kamboju@linaro.org writes:
selftests: ptrace: peeksiginfo failed on x86_64, i386, arm64 and arm. FAILED on stable rc branches 4.19, 4.14, 4.9 and 4.4. PASS on mainline, next and 5.1 stable rc branch.
Greg.
Looking in my email it appears 4.19, 4.14, 4.9, and 4.4 patches are missing the "found=1" line from the original change. This explains the test failure.
Ugh, that was due to me doing the backport by hand, sorry about that. Sasha, thanks for fixing this.
greg k-h
-
ebiederm＠xmission.com -
Greg Kroah-Hartman -
Naresh Kamboju -
Sasha Levin