From: Adrian Hunter Subject: Re: mmc: block: bonnie++ runs with errors on arc/hsdk board Date: Fri, 16 Mar 2018 09:20:35 +0200 Message-ID: <94e11cab-8c46-4780-1413-f08e8f848342@intel.com> References: <1520942674.10285.8.camel@synopsys.com> <486fe92e-9b53-53a4-f98d-129a97032470@intel.com> <1520956602.10285.10.camel@synopsys.com> <1521030769.10285.16.camel@synopsys.com> <83de0f90-0919-f449-b432-494b9bf524a1@intel.com> <345e9053-9251-ad27-d05b-ec80b69ee235@synopsys.com> <1521134509.6358.1.camel@synopsys.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: "Evgeniy.Didin@synopsys.com" , "ulf.hansson@linaro.org" , "linux-ext4@vger.kernel.org" , "tytso@mit.edu" , "vincent.weaver@maine.edu" , "peterz@infradead.org" , "linus.walleij@linaro.org" , "linux-mmc@vger.kernel.org" , "jh80.chung@samsung.com" , "adilger.kernel@dilger.ca" , "linux-snps-arc@lists.infradead.org" , "Eugeniy.Paltsev@synopsys.com" To: Alexey Brodkin , Vineet Gupta Return-path: In-Reply-To: <1521134509.6358.1.camel@synopsys.com> Content-Language: en-US List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-snps-arc" Errors-To: linux-snps-arc-bounces+gla-linux-snps-arc=m.gmane.org@lists.infradead.org List-Id: linux-ext4.vger.kernel.org On 15/03/18 19:21, Alexey Brodkin wrote: > Hi Vineet, > > On Thu, 2018-03-15 at 09:38 -0700, Vineet Gupta wrote: >> On 03/15/2018 02:27 AM, Adrian Hunter wrote: >>> Please try: >>> >>> perf record -a -e mmc:* -- bonnie++ -u root -r 256 -s 512 -x 1 -d /mnt >>> >>> and share the resulting perf.data file when you are able >>> to reproduce the hung task message. >>> >>> You may want to add this patch first: >>> >>> https://marc.info/?l=linux-mmc&m=152110580324200 >> >> While this SoC has perf counters, Unfortunately the overflow interrupts are not >> wired up. We can do perf stats but not sampling. > > I guess here Adrian was asking for pure software MMC-related events which are > orthogonal to CPU's perf interrupts. > > Adrian? Yes, these are mmc tracepoints which are software events. The file is full of them, so it certainly worked.