Re: [PATCH 8/9] scsi: remove the use_clustering flag
From: Christoph Hellwig <hch@lst.de>
Date: 2018-12-08 17:51:43
On Thu, Dec 06, 2018 at 09:13:24AM -0800, Bart Van Assche wrote:
On Thu, 2018-12-06 at 07:52 -0800, Christoph Hellwig wrote:quoted
The same effects can be archived by setting the dma_boundary to^^^^^^^^ achieved?quoted
PAGE_SIZE - 1 and the max_segment_size to PAGE_SIZE, so shift those settings into the drivers. Note that in many cases the setting might be bogus, but this keeps the status quo.[ ... ]quoted
Documentation/scsi/scsi_mid_low_api.txt | 2 -- arch/ia64/hp/sim/simscsi.c | 3 ++- drivers/scsi/a2091.c | 3 ++- drivers/scsi/advansys.c | 4 ++-- drivers/scsi/aha152x.c | 3 ++- drivers/scsi/arm/acornscsi.c | 3 ++- drivers/scsi/arm/arxescsi.c | 3 ++- drivers/scsi/arm/cumana_1.c | 3 ++- drivers/scsi/arm/cumana_2.c | 3 ++- drivers/scsi/arm/eesox.c | 3 ++- drivers/scsi/arm/oak.c | 3 ++- drivers/scsi/atari_scsi.c | 3 ++- drivers/scsi/cxgbi/cxgb3i/cxgb3i.c | 3 ++- drivers/scsi/cxgbi/cxgb4i/cxgb4i.c | 3 ++- drivers/scsi/dc395x.c | 3 ++- drivers/scsi/dmx3191d.c | 3 ++- drivers/scsi/g_NCR5380.c | 3 ++- drivers/scsi/gvp11.c | 3 ++- drivers/scsi/hosts.c | 1 - drivers/scsi/ips.c | 1 - drivers/scsi/iscsi_tcp.c | 3 ++- drivers/scsi/mac_esp.c | 3 ++- drivers/scsi/mac_scsi.c | 3 ++- drivers/scsi/mvumi.c | 3 ++- drivers/scsi/nsp32.c | 3 ++- drivers/scsi/pcmcia/nsp_cs.c | 3 ++- drivers/scsi/pcmcia/qlogic_stub.c | 3 ++- drivers/scsi/qlogicfas.c | 3 ++- drivers/scsi/scsi_debug.c | 6 ++++-- drivers/scsi/scsi_lib.c | 3 --- drivers/scsi/sgiwd93.c | 3 ++- drivers/scsi/stex.c | 3 ++- drivers/scsi/sun3_scsi.c | 3 ++- drivers/scsi/ufs/ufshcd.c | 3 ++- drivers/target/loopback/tcm_loop.c | 3 ++- drivers/usb/storage/uas.c | 3 ++- include/scsi/scsi_host.h | 13 ------------- 37 files changed, 66 insertions(+), 54 deletions(-)Please split this patch into two patches: one that affects SCSI LLD drivers and another patch that affects the SCSI core. That will make both reviewing and bisecting easier.
Just splitting it in two would not really help bisecting. We could split it into one per driver, but we rarely need to bisect which driver is affected, as that usually is pretty obvious.