I am still seeing dropped entries on a slightly different case: # fio -version fio-2.12-5-g0b2e # fio --name=test_job --ioengine=libaio --direct=1 --rw=randread --iodepth=256 --size=100% --numjobs=2 --bs=4k --filename=/dev/nvme0n1 --group_reporting --write_bw_log=test_job --write_iops_log=test_job --write_lat_log=test_job --log_avg_msec=1000 --disable_lat=0 --disable_clat=0 --disable_slat=0 --runtime=60s --time_based --output=test_job # cat test_job_iops.1.log 1000, 375601, 0, 0 3000, 243539, 0, 0 4000, 462076, 0, 0 ... Regards, Jeff -----Original Message----- From: fio-owner@xxxxxxxxxxxxxxx [mailto:fio-owner@xxxxxxxxxxxxxxx] On Behalf Of Karl Cronburg Sent: Thursday, July 7, 2016 12:47 PM To: Jens Axboe <axboe@xxxxxxxxx> Cc: fio@xxxxxxxxxxxxxxx Subject: Re: what is the sample logger doing when the buffer fills up? Yes, my patch fixed the dropping of entries I was seeing. On Thu, Jul 7, 2016 at 2:36 PM, Jens Axboe <axboe@xxxxxxxxx> wrote: > On Wed, Jul 06 2016, Karl Cronburg wrote: >> Hello all, >> >> When I change add_clat_sample() in stat.c to call add_log_sample() >> seven times instead of once, and I set DEF_LOG_ENTRIES=4096, I get >> the following in the clat log file (note first column is line number in the file): >> >> 4089 143, 9832, 0, 131072 >> 4090 143, 9832, 0, 131072 >> 4091 143, 9832, 0, 131072 >> 4092 143, 9832, 0, 131072 >> 4093 143, 9832, 0, 131072 >> 4094 143, 9832, 0, 131072 >> 4095 143, 9832, 0, 131072 >> 4096 143, 9834, 0, 131072 >> 4097 170, 15982, 0, 131072 >> 4098 170, 15982, 0, 131072 >> 4099 170, 15982, 0, 131072 >> 4100 170, 15982, 0, 131072 >> 4101 170, 15982, 0, 131072 >> 4102 170, 15982, 0, 131072 >> 4103 170, 15982, 0, 131072 >> 4104 170, 15985, 0, 131072 >> 4105 170, 15985, 0, 131072 >> >> Notice the lonely '9834' value in the center, which should have 7 >> exact copies of itself but gets cut off right at the buffer size of >> 4096 entries. >> >> So if I'm interpreting this correctly, samples get silently dropped >> (or not properly printed?) when the buffer fills up. Is this a bug? >> Am I misinterpreting what the doubly linked list is for? I was under >> the impression that a new sample buffer should be added to the linked >> list when the current one fills up. > > Fair to assume this is what your patch fixed, correct? Just wanted to > ensure that it's already fixed. > > -- > Jens Axboe > -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html Western Digital Corporation (and its subsidiaries) E-mail Confidentiality Notice & Disclaimer: This e-mail and any files transmitted with it may contain confidential or legally privileged information of WDC and/or its affiliates, and are intended solely for the use of the individual or entity to which they are addressed. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited. If you have received this e-mail in error, please notify the sender immediately and delete the e-mail in its entirety from your system. ��.n��������+%������w��{.n�������^n�r������&��z�ޗ�zf���h���~����������_��+v���)ߣ�
