| Message ID | 20210928235442.201875-1-don.brace@microchip.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | smartpqi updates | expand |
Dear Don, Just a small nit regarding most patches in the patch queue. It’d be great if the full text width of 75 characters could be used in the commit message bodies. Currently they are well below that, and therefore take more lines than necessary and are harder to read for me. Kind regards, Paul
-----Original Message-----
From: Paul Menzel [mailto:pmenzel@molgen.mpg.de]
Subject: Re: [smartpqi updates PATCH V2 00/11] smartpqi updates
Dear Don,
Just a small nit regarding most patches in the patch queue.
It’d be great if the full text width of 75 characters could be used in the commit message bodies. Currently they are well below that, and therefore take more lines than necessary and are harder to read for me.
Kind regards,
Paul
---
I can re-word and re-send if you like.
Thanks,
Don Brace
Dear Don, Am 29.09.21 um 16:08 schrieb Don.Brace@microchip.com: > -----Original Message----- > From: Paul Menzel [mailto:pmenzel@molgen.mpg.de] > Subject: Re: [smartpqi updates PATCH V2 00/11] smartpqi updates > > Dear Don, > > > Just a small nit regarding most patches in the patch queue. > > It’d be great if the full text width of 75 characters could be used in the commit message bodies. Currently they are well below that, and therefore take more lines than necessary and are harder to read for me. > > > Kind regards, > > Paul > --- > I can re-word and re-send if you like. If you sent V3 due to other reasons, then yes. Otherwise, please just keep it in mind for the future. Kind regards, Paul
On Tue, 28 Sep 2021 18:54:31 -0500, Don Brace wrote: > These patches are based on Martin Petersen's 5.16/scsi-queue tree > https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git > 5.16/scsi-queue > > This set of changes consist of: > * Aligning device removal with our out of box driver. > * Aligning kdump timing with controller memory dump. > The OS was rebooting before the controller was finished dumping its own > memory. Now the driver will wait for the controller to indicate that its > dump has completed. > * In rare cases where the controller stops responding to the driver, the > driver can set reason codes to aid in debugging. > * Enhance device reset operations. The driver was not obtaining the current > number of outstanding commands during the check for outstanding command > completions. This was causing reset hangs. > * Add in a check for HBA devices undergoing sanitize. This was causing long > boot up delays while the OS waited for sanitize to complete. The fix is to > check for sanitize and keep the HBA disk offline. Note that the SSA spec > states that the disk must be manually re-enabled after sanitize has > completed. The link to the spec is noted in the patch. > * When the OS off-lines a disk, the SCSI command pointers are cleaned up. > The driver was attempting to return some outstanding commands that were > no longer valid. > * Add in more enhanced report physical luns (RPL) command. This is an > internal command that yields more complete WWID information. > * Correct a rare case where a poll for a register status before the > register has been updated. > * When multi-LUN tape devices are added to the OS, the OS does its own > report LUNs and the tape devices were duplicated. A simple fix was to update > slave_alloc/slave_configure to prevent this. > * Add in some new PCI devices. > * Bump the driver version. > > [...] Applied to 5.16/scsi-queue, thanks! [01/11] smartpqi: update device removal management https://git.kernel.org/mkp/scsi/c/819225b03dc7 [02/11] smartpqi: add controller handshake during kdump https://git.kernel.org/mkp/scsi/c/9ee5d6e9ac52 [03/11] smartpqi: capture controller reason codes https://git.kernel.org/mkp/scsi/c/5d1f03e6f49a [04/11] smartpqi: update LUN reset handler https://git.kernel.org/mkp/scsi/c/6ce1ddf53252 [05/11] smartpqi: add tur check for sanitize operation https://git.kernel.org/mkp/scsi/c/be76f90668d8 [06/11] smartpqi: avoid failing ios for offline devices https://git.kernel.org/mkp/scsi/c/4f3cefc3084d [07/11] smartpqi: add extended report physical luns https://git.kernel.org/mkp/scsi/c/28ca6d876c5a [08/11] smartpqi: fix boot failure during lun rebuild https://git.kernel.org/mkp/scsi/c/987d35605b7e [09/11] smartpqi: fix duplicate device nodes for tape changers https://git.kernel.org/mkp/scsi/c/d4dc6aea93cb [10/11] smartpqi: add 3252-8i pci id https://git.kernel.org/mkp/scsi/c/80982656b78e [11/11] smartpqi: update version to 2.1.12-055 https://git.kernel.org/mkp/scsi/c/605ae389ea02
These patches are based on Martin Petersen's 5.16/scsi-queue tree https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git 5.16/scsi-queue This set of changes consist of: * Aligning device removal with our out of box driver. * Aligning kdump timing with controller memory dump. The OS was rebooting before the controller was finished dumping its own memory. Now the driver will wait for the controller to indicate that its dump has completed. * In rare cases where the controller stops responding to the driver, the driver can set reason codes to aid in debugging. * Enhance device reset operations. The driver was not obtaining the current number of outstanding commands during the check for outstanding command completions. This was causing reset hangs. * Add in a check for HBA devices undergoing sanitize. This was causing long boot up delays while the OS waited for sanitize to complete. The fix is to check for sanitize and keep the HBA disk offline. Note that the SSA spec states that the disk must be manually re-enabled after sanitize has completed. The link to the spec is noted in the patch. * When the OS off-lines a disk, the SCSI command pointers are cleaned up. The driver was attempting to return some outstanding commands that were no longer valid. * Add in more enhanced report physical luns (RPL) command. This is an internal command that yields more complete WWID information. * Correct a rare case where a poll for a register status before the register has been updated. * When multi-LUN tape devices are added to the OS, the OS does its own report LUNs and the tape devices were duplicated. A simple fix was to update slave_alloc/slave_configure to prevent this. * Add in some new PCI devices. * Bump the driver version. Changes since V1: * Corrected issues with my e-mail server. Don Brace (3): smartpqi: update device removal management smartpqi: add tur check for sanitize operation smartpqi: update version to 2.1.12-055 Kevin Barnett (2): smartpqi: update LUN reset handler smartpqi: fix duplicate device nodes for tape changers Mahesh Rajashekhara (2): smartpqi: add controller handshake during kdump smartpqi: avoid failing ios for offline devices Mike McGowen (3): smartpqi: add extended report physical luns smartpqi: fix boot failure during lun rebuild smartpqi: add 3252-8i pci id Murthy Bhat (1): smartpqi: capture controller reason codes drivers/scsi/smartpqi/smartpqi.h | 61 +- drivers/scsi/smartpqi/smartpqi_init.c | 540 +++++++++++++----- .../scsi/smartpqi/smartpqi_sas_transport.c | 6 +- drivers/scsi/smartpqi/smartpqi_sis.c | 60 +- drivers/scsi/smartpqi/smartpqi_sis.h | 4 +- 5 files changed, 509 insertions(+), 162 deletions(-)