Message ID | 20200904045128.23631-1-njavali@marvell.com |
---|---|
Headers | show |
Series | qla2xxx misc features and bug fixes | expand |
On 2020-09-03 21:51, Nilesh Javali wrote: > This patch tracks number of IOCB resources used in the IO fast path. > If the number of used IOCBs reach a high water limit, driver would > return the IO as busy and let upper layer retry. This prevents over > subscription of IOCB resources where any future error recovery command > is unable to cut through. > Enable IOCB throttling by default. Please use the block layer reserved tag mechanism instead of adding a mechanism that is (a) racy and (b) triggers cache line ping-pong. QT: The Block layer reserve tag does fit resource we're tracking. The reserve tag is per command. The IOCB resource is adapter specific behind all commands. QT: I mean " The Block layer reserve tag does NOT fit the resource we're tracking"
Martin, We do not anticipate further changes to this patch or v3 series. Thanks, Nilesh > -----Original Message----- > From: Quinn Tran <qutran@marvell.com> > Sent: Thursday, September 17, 2020 10:23 PM > To: Bart Van Assche <bvanassche@acm.org>; Nilesh Javali > <njavali@marvell.com>; martin.petersen@oracle.com > Cc: linux-scsi@vger.kernel.org; GR-QLogic-Storage-Upstream <GR-QLogic- > Storage-Upstream@marvell.com> > Subject: RE: [PATCH v3 11/13] qla2xxx: Add IOCB resource tracking > > > On 2020-09-03 21:51, Nilesh Javali wrote: > > This patch tracks number of IOCB resources used in the IO fast path. > > If the number of used IOCBs reach a high water limit, driver would > > return the IO as busy and let upper layer retry. This prevents over > > subscription of IOCB resources where any future error recovery command > > is unable to cut through. > > Enable IOCB throttling by default. > > Please use the block layer reserved tag mechanism instead of adding a > mechanism that is (a) racy and (b) triggers cache line ping-pong. > > QT: The Block layer reserve tag does fit resource we're tracking. The reserve > tag is per command. The IOCB resource is adapter specific behind all > commands. > > QT: I mean " The Block layer reserve tag does NOT fit the resource we're > tracking" >
Nilesh,
> We do not anticipate further changes to this patch or v3 series.
Applied to 5.10/scsi-staging. Next time, please run checkpatch.
On Thu, 3 Sep 2020 21:51:15 -0700, Nilesh Javali wrote: > Martin, > > Please apply the attached miscellaneous qla2xxx features and bug fixes > to the scsi tree at your earliest convenience. > > v1->v2: > Fix compilation error reported by kernel test robot > > [...] Applied to 5.10/scsi-queue, thanks! [01/13] scsi: qla2xxx: Fix I/O failures during remote port toggle testing https://git.kernel.org/mkp/scsi/c/dd8d0bf6fb72 [02/13] scsi: qla2xxx: Setup debugfs entries for remote ports https://git.kernel.org/mkp/scsi/c/1e98fb0f9208 [03/13] scsi: qla2xxx: Allow dev_loss_tmo setting for FC-NVMe devices https://git.kernel.org/mkp/scsi/c/27c8aa5e1b06 [04/13] scsi: qla2xxx: Honor status qualifier in FCP_RSP per spec https://git.kernel.org/mkp/scsi/c/3aac0c0fde17 [05/13] scsi: qla2xxx: Reduce duplicate code in reporting speed https://git.kernel.org/mkp/scsi/c/d68930bae477 [06/13] scsi: qla2xxx: Fix memory size truncation https://git.kernel.org/mkp/scsi/c/d38cb849e17a [07/13] scsi: qla2xxx: Performance tweak https://git.kernel.org/mkp/scsi/c/49db4d4e02aa [08/13] scsi: qla2xxx: Fix I/O errors during LIP reset tests https://git.kernel.org/mkp/scsi/c/a35f87bdcc06 [09/13] scsi: qla2xxx: Make tgt_port_database available in initiator mode https://git.kernel.org/mkp/scsi/c/4e5a05d1ecd9 [10/13] scsi: qla2xxx: Add rport fields in debugfs https://git.kernel.org/mkp/scsi/c/6152d20fa670 [11/13] scsi: qla2xxx: Add IOCB resource tracking https://git.kernel.org/mkp/scsi/c/89c72f4245a8 [12/13] scsi: qla2xxx: Add SLER and PI control support https://git.kernel.org/mkp/scsi/c/cf3c54fb49a4 [13/13] scsi: qla2xxx: Update version to 10.02.00.102-k https://git.kernel.org/mkp/scsi/c/767c8457b729 -- Martin K. Petersen Oracle Linux Engineering