Date: Wed, 22 Jan 2020 00:43:51 +0000 (UTC) From: Alexander Motin <mav@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-12@freebsd.org Subject: svn commit: r356955 - stable/12/sys/dev/nvme Message-ID: <202001220043.00M0hp2B028975@repo.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: mav Date: Wed Jan 22 00:43:51 2020 New Revision: 356955 URL: https://svnweb.freebsd.org/changeset/base/356955 Log: MFC r355631 (by imp): Move reset to the interrutp processing stage This trims the boot time a bit more for AWS and other platforms that have nvme drives. There's no reason too do this inline. This has been in my tree a while, but IIRC I talked to Jim Harris about this at one of our face to face meetings. Modified: stable/12/sys/dev/nvme/nvme.c stable/12/sys/dev/nvme/nvme_ctrlr.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sys/dev/nvme/nvme.c ============================================================================== --- stable/12/sys/dev/nvme/nvme.c Wed Jan 22 00:41:18 2020 (r356954) +++ stable/12/sys/dev/nvme/nvme.c Wed Jan 22 00:43:51 2020 (r356955) @@ -130,25 +130,6 @@ nvme_attach(device_t dev) int status; status = nvme_ctrlr_construct(ctrlr, dev); - - if (status != 0) { - nvme_ctrlr_destruct(ctrlr, dev); - return (status); - } - - /* - * Reset controller twice to ensure we do a transition from cc.en==1 to - * cc.en==0. This is because we don't really know what status the - * controller was left in when boot handed off to OS. Linux doesn't do - * this, however. If we adopt that policy, see also nvme_ctrlr_resume(). - */ - status = nvme_ctrlr_hw_reset(ctrlr); - if (status != 0) { - nvme_ctrlr_destruct(ctrlr, dev); - return (status); - } - - status = nvme_ctrlr_hw_reset(ctrlr); if (status != 0) { nvme_ctrlr_destruct(ctrlr, dev); return (status); Modified: stable/12/sys/dev/nvme/nvme_ctrlr.c ============================================================================== --- stable/12/sys/dev/nvme/nvme_ctrlr.c Wed Jan 22 00:41:18 2020 (r356954) +++ stable/12/sys/dev/nvme/nvme_ctrlr.c Wed Jan 22 00:43:51 2020 (r356955) @@ -909,6 +909,25 @@ void nvme_ctrlr_start_config_hook(void *arg) { struct nvme_controller *ctrlr = arg; + int status; + + /* + * Reset controller twice to ensure we do a transition from cc.en==1 to + * cc.en==0. This is because we don't really know what status the + * controller was left in when boot handed off to OS. Linux doesn't do + * this, however. If we adopt that policy, see also nvme_ctrlr_resume(). + */ + status = nvme_ctrlr_hw_reset(ctrlr); + if (status != 0) { + nvme_ctrlr_fail(ctrlr); + return; + } + + status = nvme_ctrlr_hw_reset(ctrlr); + if (status != 0) { + nvme_ctrlr_fail(ctrlr); + return; + } nvme_qpair_reset(&ctrlr->adminq); nvme_admin_qpair_enable(&ctrlr->adminq);
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?202001220043.00M0hp2B028975>