The recent patchset to use dma_set_mask_and_coherent() introduced a regression where a call to set a 64-bit DMA mask was followed by a call to set a 32-bit DMA mask, leading to I/O errors and data corruption. Patchset is based on a suggestions from Ewan Milne. Hannes Reinecke (8): lpfc: fix calls to dma_set_mask_and_coherent() 3w-9xxx: fix calls to dma_set_mask_and_coherent() 3w-sas: fix calls to dma_set_mask_and_coherent() 3w-sas: fix calls to dma_set_mask_and_coherent() bfa: fix calls to dma_set_mask_and_coherent() csiostor: fix calls to dma_set_mask_and_coherent() hisi_sas: fix calls to dma_set_mask_and_coherent() hptiop: fix calls to dma_set_mask() drivers/scsi/3w-9xxx.c | 14 +++++++++----- drivers/scsi/3w-sas.c | 12 ++++++++---- drivers/scsi/aic94xx/aic94xx_init.c | 8 +++++--- drivers/scsi/bfa/bfad.c | 18 +++++++++++++----- drivers/scsi/csiostor/csio_init.c | 7 +++++-- drivers/scsi/hisi_sas/hisi_sas_main.c | 8 ++++++-- drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 8 +++++--- drivers/scsi/hptiop.c | 10 +++++++--- drivers/scsi/lpfc/lpfc_init.c | 19 ++++++++++++------- 9 files changed, 70 insertions(+), 34 deletions(-) -- 2.16.4 ==== The code changes look fine, I only have 3 commit-msg related comments: 4/8 - Subject line should read "aic94xx:" not "3w-sas:" (from 3/8?) 5/8 - Remove last sentence from patch description 7/8 - Remove last sentence from patch description (The only errors I actually encountered were with the lpfc driver.) Martin, can you fix this up in patchwork? I'll add Reviewed-by: each. -Ewan
