From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ozlabs.org (bilbo.ozlabs.org [203.11.71.1]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 40sbFN5mgtzF1fW for ; Fri, 25 May 2018 16:00:08 +1000 (AEST) Message-ID: Subject: Re: [PATCH 2/2] selftests/powerpc: Add perf breakpoint test From: Michael Neuling To: Michael Ellerman Cc: linuxppc-dev@lists.ozlabs.org Date: Fri, 25 May 2018 16:00:08 +1000 In-Reply-To: <87tvqx4a7q.fsf@concordia.ellerman.id.au> References: <20180522061428.5142-1-mikey@neuling.org> <20180522061428.5142-2-mikey@neuling.org> <87tvqx4a7q.fsf@concordia.ellerman.id.au> Content-Type: text/plain; charset="UTF-8" Mime-Version: 1.0 List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Thu, 2018-05-24 at 20:30 +1000, Michael Ellerman wrote: > Michael Neuling writes: >=20 > > This tests perf hardware breakpoints (ie PERF_TYPE_BREAKPOINT) on > > powerpc. >=20 > This doesn't work for me on a P8 guest: >=20 > test: perf-hwbreak > tags: git_version:bb5602e > !! killing perf-hwbreak > !! child died by signal 15 > failure: perf-hwbreak >=20 >=20 > That means the harness killed it after 2 minutes. >=20 > Sometimes it gets further: >=20 > test: perf-hwbreak > tags: git_version:v4.17-rc3-109-gbb20240fb508 > threads=3D16 loops=3D1048576 scalar test > !! killing perf-hwbreak > !! child died by signal 15 > failure: perf-hwbreak >=20 >=20 > Do we just need to bump the timeout up, or is something going wrong > here? Yeah, sorry my bad. The test was trying to randomly test all the combinati= ons. =20 I've rewritten it now to explicitly test all the combinations. This gives i= t a more consistent runtime. I do randomly vary the loop count but this is only= a small variation (~10%) to avoid using a count from a previous test. Mikey