Re: [PATCH v3] selftests/seccomp: build and pass on arm64

On 10/15/2015 05:00 PM, Shuah Khan wrote:

On 10/15/2015 04:07 PM, Kees Cook wrote:

quoted

On Thu, Oct 15, 2015 at 1:06 PM, Shuah Khan [off-list ref] wrote:

quoted

On 10/15/2015 12:42 PM, Kees Cook wrote:

quoted

On Thu, Oct 15, 2015 at 7:07 AM, Shuah Khan [off-list ref] wrote:

quoted

On 10/06/2015 01:30 PM, Kees Cook wrote:

quoted

Changing arm64 syscalls is done via a specific register set, more like s390
than like arm (specific ptrace call) and x86 (part of general registers).
Since (restarting) poll doesn't exist on arm64, switch to using nanosleep
for testing restart_syscall. And since it looks like the syscall ABI is
inconsistent on arm-compat, so we must work around it (and document it) in
the test.

Signed-off-by: Kees Cook <redacted>
---
v3:
 - correctly set syscall number on native arm64.
v2:
 - switch to nanosleep from a bad mix of poll and ppoll for testing restart.
---

Is this good to go? Failed to apply to linux-kselftest next.
If you can rebase and resend. I can get this into 4.4-rc1

Yes please. :)

ok. Please rebase to linux-kselftest next and resend the patch.

Am I looking at the right tree? linux-kselftest#next doesn't appear to
have the s390 patch that was included in 4.3.

I see what happened. Your patch is linux-next fixes and that went into
4.3-rc2. I can get linux-next rebase to 4.3-rc2 and get your patch in.
Thanks for clearing this up.

oops rebase linux-kselftest next i.e

-- Shuah


-- 
Shuah Khan
Sr. Linux Kernel Developer
Open Source Innovation Group
Samsung Research America (Silicon Valley)
shuahkh-JPH+aEBZ4P+UEJcrhfAQsw@public.gmane.org | (970) 217-8978