Re: blk-mq: preparation of the ground for BFQ inclusion
On 12/02/2016 12:38 AM, Paolo Valente wrote: > >> Il giorno 02 dic 2016, alle ore 01:27, Jens Axboe ha >> scritto: >> >> On 12/01/2016 02:03 PM, Jens Axboe wrote: >>> On 11/14/2016 03:42 PM, Jens Axboe wrote: On 11/14/2016 03:13 PM, Paolo Valente wrote: > Hi Jens, > have you had time to look into the first extensions/changes required? > Any time plan? I was busy last week on the writeback and on the polling. I should have something end this week. >>> >>> It's coming together, running some testing and fixing some issues. But I >>> now have unmodified legacy IO schedulers running under blk-mq, for >>> single queue devices. Scheduler queue depth is independent of the device >>> queue depth. >> >> Here is the branch, btw: >> >> http://git.kernel.dk/cgit/linux-block/log/?h=for-4.11/blk-mq-legacy-sched >> >> if you want to look at it or test it. It's not in a clean state, I >> wanted to get this reasonably working before cleaning it up. Should be >> ready for testing, I've successfully run various IO tests on deadline >> and CFQ with null_blk loaded with queue_mode=2 submit_queues=1 >> use_sched=1. >> > > Good news! > > I'm familiarizing with your patch. To avoid useless, too-early report, > tell me if/when I can start doing some test too. Hopefully after today, trying to shake out a bug related to non file system requests that's making SCSI probing broken. Once that's fixed up, stability should be fine for testing. -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: blk-mq: preparation of the ground for BFQ inclusion
> Il giorno 02 dic 2016, alle ore 01:27, Jens Axboe ha > scritto: > > On 12/01/2016 02:03 PM, Jens Axboe wrote: >> On 11/14/2016 03:42 PM, Jens Axboe wrote: >>> On 11/14/2016 03:13 PM, Paolo Valente wrote: Hi Jens, have you had time to look into the first extensions/changes required? Any time plan? >>> >>> I was busy last week on the writeback and on the polling. I should have >>> something end this week. >> >> It's coming together, running some testing and fixing some issues. But I >> now have unmodified legacy IO schedulers running under blk-mq, for >> single queue devices. Scheduler queue depth is independent of the device >> queue depth. > > Here is the branch, btw: > > http://git.kernel.dk/cgit/linux-block/log/?h=for-4.11/blk-mq-legacy-sched > > if you want to look at it or test it. It's not in a clean state, I > wanted to get this reasonably working before cleaning it up. Should be > ready for testing, I've successfully run various IO tests on deadline > and CFQ with null_blk loaded with queue_mode=2 submit_queues=1 > use_sched=1. > Good news! I'm familiarizing with your patch. To avoid useless, too-early report, tell me if/when I can start doing some test too. Thanks, Paolo > -- > Jens Axboe > -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: blk-mq: preparation of the ground for BFQ inclusion
On 12/01/2016 02:03 PM, Jens Axboe wrote: > On 11/14/2016 03:42 PM, Jens Axboe wrote: >> On 11/14/2016 03:13 PM, Paolo Valente wrote: >>> Hi Jens, >>> have you had time to look into the first extensions/changes required? >>> Any time plan? >> >> I was busy last week on the writeback and on the polling. I should have >> something end this week. > > It's coming together, running some testing and fixing some issues. But I > now have unmodified legacy IO schedulers running under blk-mq, for > single queue devices. Scheduler queue depth is independent of the device > queue depth. Here is the branch, btw: http://git.kernel.dk/cgit/linux-block/log/?h=for-4.11/blk-mq-legacy-sched if you want to look at it or test it. It's not in a clean state, I wanted to get this reasonably working before cleaning it up. Should be ready for testing, I've successfully run various IO tests on deadline and CFQ with null_blk loaded with queue_mode=2 submit_queues=1 use_sched=1. -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: blk-mq: preparation of the ground for BFQ inclusion
On 11/14/2016 03:42 PM, Jens Axboe wrote: > On 11/14/2016 03:13 PM, Paolo Valente wrote: >> Hi Jens, >> have you had time to look into the first extensions/changes required? >> Any time plan? > > I was busy last week on the writeback and on the polling. I should have > something end this week. It's coming together, running some testing and fixing some issues. But I now have unmodified legacy IO schedulers running under blk-mq, for single queue devices. Scheduler queue depth is independent of the device queue depth. -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: blk-mq: preparation of the ground for BFQ inclusion
On 11/14/2016 03:13 PM, Paolo Valente wrote: Hi Jens, have you had time to look into the first extensions/changes required? Any time plan? I was busy last week on the writeback and on the polling. I should have something end this week. -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html