Message ID | 20230901114336.31339-1-quic_nitirawa@quicinc.com (mailing list archive) |
---|---|
Headers | show |
Series | scsi: ufs: qcom: Align programming sequence as per HW spec | expand |
On 1.09.2023 13:43, Nitin Rawat wrote: > This patch aligns programming sequence as per Qualcomm UFS > hardware specification. > > changes from v5: > - Addressed Mani comment to FIELD_PREP and FIELD_FIT. > - Optimised ufs_qcom_set_core_clk_ctrl API. > - Updated commit text for few patches to capture more details. > Any reason I received this twice? Konrad
On Fri, Sep 01, 2023 at 05:13:30PM +0530, Nitin Rawat wrote: > This patch aligns programming sequence as per Qualcomm UFS > hardware specification. reading this series, it is difficult for me to understand as a user of the driver if this should have any noticeable effect. Some of the patches mention that there is no functional change, some only say align with the HPG but change programming sequence, frequency, etc if I understand correctly on a quick glance. I think being a bit verbose in some of the patches with respect to explaining the effect of the patch (or lack of a noticeable effect) would be a beneficial improvement to this series if there's another version. I agree that aligning with the HPG instead of doing some undefined sequence is a good idea, I'm just reading some of the changes and thinking "I have no idea if this is going to fix something (no Fixes: tag but it almost sounds like one), will this improve something, or will this just change the programming sequence to a known and recommended sequence?". Thanks for the patches, Andrew
On Fri, Sep 01, 2023 at 09:56:05AM -0500, Andrew Halaney wrote: > On Fri, Sep 01, 2023 at 05:13:30PM +0530, Nitin Rawat wrote: > > This patch aligns programming sequence as per Qualcomm UFS > > hardware specification. > > reading this series, it is difficult for me to understand as a user of > the driver if this should have any noticeable effect. > > Some of the patches mention that there is no functional change, some > only say align with the HPG but change programming sequence, frequency, > etc if I understand correctly on a quick glance. > > I think being a bit verbose in some of the patches with respect to > explaining the effect of the patch (or lack of a noticeable effect) > would be a beneficial improvement to this series if there's another > version. > > I agree that aligning with the HPG instead of doing some undefined > sequence is a good idea, I'm just reading some of the changes and > thinking "I have no idea if this is going to fix something (no Fixes: > tag but it almost sounds like one), will this improve something, or will this > just change the programming sequence to a known and recommended > sequence?". > Very valid feedback, Andrew. Correct or not, there are a fair amount of users out there who runs the current implementation. Changes to that should be described in a way that doesn't depend on inside-knowledge. Regards, Bjorn