Messages in this thread | | | Subject | Re: [PATCH v4 10/12] arm64: vdso: replace gettimeofday.S with global vgettimeofday.C | From | Mark Salyzyn <> | Date | Tue, 21 Nov 2017 13:28:13 -0800 |
| |
On 11/21/2017 10:03 AM, Jeffrey Bastian wrote: > On Thu, Nov 02, 2017 at 12:22:36AM -0400, Jon Masters wrote: >> On Nov 2, 2017, at 06:14, Mark Salyzyn <salyzyn@android.com> wrote: >>> Holding off respin or followup patch until after we get a response >>> from Jon Masters <jcm@readhat.com> on their QE tests on this patch >>> series. >> Copying Jeff Bastian who owns QE. Jeff, can you recall what test was >> failing when we tried the gettimeofday C rewrite VDSO patches? > The LTP clock_getres01 test was failing when we tried the initial > implementation of gettimeofday-in-C. This would be a good place to > start when testing the new version of the patches. > > https://github.com/linux-test-project/ltp/blob/master/testcases/kernel/syscalls/clock_getres/clock_getres01.c > > Thanks.
I ran the test (32 and 64 bit) on my latest work (but back-ported to 4.9) and it passes. I did not change __vdso_clock_getres or the fallback (syscall) handler in the patch series though, so the failure has me wondering. Could you re-run the test with latest to be sure, and I will look into setting up to test with a ToT kernel (after the thanksgiving break).
Using bionic libc, with advance modifications to utilize __vdso_clock_getres; required in order to support the test.
NB: in AOSP _today_ bionic does _not_ propagate the errno correctly on the __vdso_clock_gettime vdso path and will understandably fail the similar tests. The public rubber stamp of __vdso_clock_gettime code to support __vdso_clock_getres will not work. errno propagation issue will be fixed in bionic.
-- Mark
| |