Page MenuHomeFreeBSD

nvme: Remove duplicate completion printing routine
ClosedPublic

Authored by imp on Aug 3 2023, 10:39 PM.
Tags
None
Referenced Files
Unknown Object (File)
Fri, Apr 12, 5:34 AM
Unknown Object (File)
Mon, Apr 8, 9:21 AM
Unknown Object (File)
Jan 12 2024, 6:14 AM
Unknown Object (File)
Nov 7 2023, 11:50 AM
Unknown Object (File)
Oct 15 2023, 9:39 AM
Unknown Object (File)
Oct 15 2023, 9:39 AM
Unknown Object (File)
Sep 28 2023, 8:47 AM
Unknown Object (File)
Sep 15 2023, 2:31 PM
Subscribers

Details

Summary

Both nvme_dump_completion and nvme_qpair_print_completion print
completions. The latter is better. Recode the two instances of
nvme_dump_completion to use nvme_qpair_print_completion and delete the
former. No sense having two nearly identical routines

Sponsored by: Netflix

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Not Applicable
Unit
Tests Not Applicable

Event Timeline

imp requested review of this revision.Aug 3 2023, 10:39 PM

nvme_dump_completion() also printed the phase value, that I think may be useful to read the dump, which prints all completion entries, completed or not.

In D41308#940723, @mav wrote:

nvme_dump_completion() also printed the phase value, that I think may be useful to read the dump, which prints all completion entries, completed or not.

That's a good idea. I'll add it back. I hadn't thought it was important when I first did it, but you make a good point.

nvme_printf() in nvme_qpair_print_completion() is a wrapper around device_printf(). So dump will now include device name and not at the beginning of the line. It will be weird. Plus P makes sense only in case of dump and IMO would be more readable if it would be the first there.

This revision is now accepted and ready to land.Aug 6 2023, 6:22 PM
In D41308#940802, @mav wrote:

nvme_printf() in nvme_qpair_print_completion() is a wrapper around device_printf(). So dump will now include device name and not at the beginning of the line. It will be weird. Plus P makes sense only in case of dump and IMO would be more readable if it would be the first there.

I'm planning in the next round to make this and the other reporting routines sbug versions. Can we defer this point until then? Or should I hold back on this one until I have that series?