| Message ID | 1549614167-1128-2-git-send-email-shivasharan.srikanteshwara@broadcom.com (mailing list archive) |
|---|---|
| State | Mainlined |
| Commit | a3742d68484083ce55de8fec492be2db2fff2cfa |
| Headers | show |
| Series | megaraid_sas: Update structures for HOST_DEVICE_LIST DCMD | expand |
On 2/8/19 9:22 AM, Shivasharan S wrote: > Add padding to make the structure variables in MR_HOST_DEVICE_LIST_ENTRY > 64-bit aligned. > Also, add reserved fields to MR_HOST_DEVICE_LIST for future firmware usage. > > Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com> > --- > drivers/scsi/megaraid/megaraid_sas.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/drivers/scsi/megaraid/megaraid_sas.h b/drivers/scsi/megaraid/megaraid_sas.h > index fe1173f02c54..eb48ca17dca0 100644 > --- a/drivers/scsi/megaraid/megaraid_sas.h > +++ b/drivers/scsi/megaraid/megaraid_sas.h > @@ -807,13 +807,14 @@ struct MR_HOST_DEVICE_LIST_ENTRY { > } flags; > u8 scsi_type; > __le16 target_id; > - u8 reserved[2]; > + u8 reserved[4]; > __le64 sas_addr[2]; > } __packed; > > struct MR_HOST_DEVICE_LIST { > __le32 size; > __le32 count; > + __le32 reserved[2]; > struct MR_HOST_DEVICE_LIST_ENTRY host_device_list[1]; > } __packed; > > Ho-hum. Move fields in a structure used for communication with the HBA? How is that supposed to work? Do we need a firmware update for the HBA to handle this correctly? Or did it never work, and we need the padding to retrieve the information correctly? Please clarify. Cheers, Hannes
> Move fields in a structure used for communication with the HBA? > How is that supposed to work? > Do we need a firmware update for the HBA to handle this correctly? > Or did it never work, and we need the padding to retrieve the > information correctly? > Please clarify. I just merged the support for the new operation a few days ago. I assume that this is a fix-up and that none of these have shipped yet?
> > Move fields in a structure used for communication with the HBA? > > How is that supposed to work? > > Do we need a firmware update for the HBA to handle this correctly? > > Or did it never work, and we need the padding to retrieve the > > information correctly? > > > Please clarify. > > I just merged the support for the new operation a few days ago. I assume that > this is a fix-up and that none of these have shipped yet? > Yes, that is correct. Firmware with support for this new DCMD has not yet shipped. The driver changes were tested earlier, but due to additional requirements the structure needed changes later. And this introduced the alignment mismatch and a fix to the earlier submission. Thanks, Shivasharan
diff --git a/drivers/scsi/megaraid/megaraid_sas.h b/drivers/scsi/megaraid/megaraid_sas.h index fe1173f02c54..eb48ca17dca0 100644 --- a/drivers/scsi/megaraid/megaraid_sas.h +++ b/drivers/scsi/megaraid/megaraid_sas.h @@ -807,13 +807,14 @@ struct MR_HOST_DEVICE_LIST_ENTRY { } flags; u8 scsi_type; __le16 target_id; - u8 reserved[2]; + u8 reserved[4]; __le64 sas_addr[2]; } __packed; struct MR_HOST_DEVICE_LIST { __le32 size; __le32 count; + __le32 reserved[2]; struct MR_HOST_DEVICE_LIST_ENTRY host_device_list[1]; } __packed;
Add padding to make the structure variables in MR_HOST_DEVICE_LIST_ENTRY 64-bit aligned. Also, add reserved fields to MR_HOST_DEVICE_LIST for future firmware usage. Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com> --- drivers/scsi/megaraid/megaraid_sas.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)