Skip to contentSkip to conversation

About the naming of ata_drive_probe_reset()

DISCUSSION Albert Lee 3 participants 3 messages May 18, 2006 libata

lore
Share & tools
lore

Albert Lee

author
#0 lore raw

Hi Tejun,

The ata_drive_probe_reset() is not drive-specific.
Actually what got reset is the port. (I guess the "drive" doesn't mean device here.) Maybe something like ata_port_probe_reset() or ata_do_probe_reset() is less ambiguous?

-- albert

#1 lore raw
Quoted 4 lines · Albert Lee

Hi Tejun,

The ata_drive_probe_reset() is not drive-specific. Actually what got reset is the port. (I guess the "drive" doesn't mean device here.) Maybe something like ata_port_probe_reset() or ata_do_probe_reset() is less ambiguous?

The 'drive' is actually a verb. The function drives (verb) probe_reset
(object) using the given methods. Maybe a bad naming. Anyways, the
whole probe_reset stuff is gonna be removed with the hotplug patchset, so I think it can be left as it is until then.

However, I also used the ata_drive_*() naming convention else where, like ata_bmdma_drive_eh(). We may need to rename them. I still want to signify that the function isn't actually a callback but something that can be used to drive a callback. I like 'drive' but as you have pointed out it can be confused with hard 'drive'. Maybe ata_steer_*()? Any other ideas?

Signature · 1 line
tejun
#2 lore raw
Quoted 5 lines · Tejun Heo

I still want to
signify that the function isn't actually a callback but something that can be used to drive a callback. I like 'drive' but as you have pointed out it can be confused with hard 'drive'. Maybe ata_steer_*()? Any other ideas?

a comment from the peanut gallery - how about something along the lines of:
ata_exec
ata_run
ata_init
ata_start
ata_kick

?

++doug