CRIS architecture update
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 6 lines · Linux Kernel Mailing List
ChangeSet 1.1726.1.144, 2004/06/01 08:52:29-07:00, [email protected]
[PATCH] CRIS architecture update
From: "Mikael Starvik" <[email protected]>
Quoted 8 lines
- Lots of fixes from 2.4.
- Updated for 2.6.6.
- Added IDE driver
Signed-off-by Andrew Morton <[email protected]>Signed-off-by Linus Torvalds <[email protected]>
Who reviewed the ethernet driver?
Who reviewed the IDE driver?
Has Bart seen this new driver?
Why was this committed without first being run by the subsystem maintainers?
In ethernet, the MII phy probe is wrong (don't need at id 0 first), it should be using linux/mii.h bits rather than inventing its own, del_timer versus del_timer_sync is questionable, the newly-added full duplex handling seems to be incorrect (ref drivers/net/mii.c and drivers that use mii_if->full_duplex), and cosmetic issues I won't bother to mention.
In IDE, it uses virt_to_page() when building the scatter/gather list -- something that IMO should not have been allowed in -mm much less mainline -- and other yuckiness. In the same function, it's questionable whether or not it breaks with lba48. etrax_dma_intr doesn't appear to check for some DMA engine conditions that code comments elsewhere in the driver indicate _do_ occur. The ATAPI-specific e100_start_dma code doesn't look like it will work with all current ATAPI drivers (ide-{cdrom,floppy,tape,scsi,...}.c).
I'm happy that the CRIS people resurfaced, too, but that's no reason to just shove an unreviewed patch into mainline.
Jeff