AIC7XXX 6.2.5 driver
When will we see v. 6.2.5 included in the kernel (2.5 and 2.4, I think)?
The patch to kernel 2.4.17 is available since January. It should fix a kernel panic at boot that has been in the kernel since November at least. The bug is in AIC7XXX V. 6.2.4. Thanks
Quoted 4 lines · Carlo Scarfoglio
When will we see v. 6.2.5 included in the kernel (2.5 and 2.4, I think)? The patch to kernel 2.4.17 is available since January. It should fix a kernel panic at boot that has been in the kernel since November at least. The bug is in AIC7XXX V. 6.2.4.
You should ask Justin whether he has submitted it for inclusion or not. I offered to port to 2.5 at least, but heard nothing.
Signature · 1 line
Jens Axboe
Quoted 2 lines
You should ask Justin whether he has submitted it for inclusion or not. I offered to port to 2.5 at least, but heard nothing.
I was sent a patch for it which included some scsi changes, broke support for the CMD ide controllers and didn't apply in the aic7xxx area. So I threw it in /dev/null
Quoted 6 lines · Alan Cox
Quoted 2 lines
You should ask Justin whether he has submitted it for inclusion or not. I offered to port to 2.5 at least, but heard nothing.
I was sent a patch for it which included some scsi changes, broke support for the CMD ide controllers and didn't apply in the aic7xxx area. So I threw it in
/dev/null
Right, it was an incredibly strange mix. I just disregarded the cmd stuff, the scsi stuff looked sane. I trust Justin knows the aic7xxx parts were sane :-)
Signature · 1 line
Jens Axboe
Quoted 6 lines
Quoted 2 lines
You should ask Justin whether he has submitted it for inclusion or not. I offered to port to 2.5 at least, but heard nothing.
I was sent a patch for it which included some scsi changes, broke support for the CMD ide controllers and didn't apply in the aic7xxx area. So I threw it in
/dev/null
All I got from you was an email with the text "Thanks!" in it. I'll take that to indicate a toss to /dev/null in the future. 8-)
Seriously, if you let me know exactly which version you want my patches relative to, I'll regenerate them. From a message Jens sent me, it sounds like his change to fix a bug I fixed differently had already been included in certain 2.4.18-pre kernels which was the cause of the patch not working for him. I can only assume you ran into the same problem. My patch was relative to 2.4.17, not a more recent, yet unblessed, kernel.
As to the CMD640 patch. Can you let me know why you believe it breaks the CMD640? The current scheme leaks transactions on the bus and *will* upset certain controllers that don't expect their register spaces to be uncerimoniously poked. Is your complaint that it only handles the attachment of a single controller? The old code is no different in this regard - Scott just had the courtesy to document that this was still not corrected by this change. If you need us to fix this other bug too in order to get the other changes into the tree, just let us know and we'll do the extra work.
Information about the SCSI mid-layer changes were posted to the SCSI list and I believe CC'd to you. If you need that information again, I'd be happy to resend it.
--
Justin
Quoted 3 lines
included in certain 2.4.18-pre kernels which was the cause of the patch not working for him. I can only assume you ran into the same problem. My patch was relative to 2.4.17, not a more recent, yet unblessed, kernel.
Relative to 2.4.18-rc would be good
Quoted 2 lines
As to the CMD640 patch. Can you let me know why you believe it breaks the CMD640? The current scheme leaks transactions on the bus and *will*
The CMD640 can't do 32bit config space access stuff - thats why the code is there to hack around it. That code needs to know or check what pci config mode it should have used.
Quoted 3 lines
Information about the SCSI mid-layer changes were posted to the SCSI list and I believe CC'd to you. If you need that information again, I'd be happy to resend it.
The midlayer stuff Im more than happy with
Quoted 5 lines
Quoted 3 lines
included in certain 2.4.18-pre kernels which was the cause of the patch not working for him. I can only assume you ran into the same problem. My patch was relative to 2.4.17, not a more recent, yet unblessed, kernel.
Relative to 2.4.18-rc would be good
Okay. I will do this sometime this weekend. I'm currently in the middle of a business trip which makes testing a new patch difficult.
Quoted 6 lines
Quoted 2 lines
As to the CMD640 patch. Can you let me know why you believe it breaks the CMD640? The current scheme leaks transactions on the bus and *will*
The CMD640 can't do 32bit config space access stuff - thats why the code is there to hack around it. That code needs to know or check what pci config mode it should have used.
Right. If you look at the comments added to the code, the PCI APIs always prefer direct config space access over using the PCI BIOS. The Linux direct access methods only perform 32 bit cycles for 32bit API calls. The ony time Linux will use the PCIBIOS (which may do everything as 32bit accesses) is if the direct access method fails. So, the machines where direct access doesn't work can't be supported by the old, manual, CMD640 code anyway. On machines that have working direct access, using the PCI APIs is equivalent to the old method with the added bonus that you don't blindly attempt to perform an access type that the chipset does not support. This was the root cause of the "CMD640 driver performs drive by shootings of other devices" issue.
Quoted 5 lines
Quoted 3 lines
Information about the SCSI mid-layer changes were posted to the SCSI list and I believe CC'd to you. If you need that information again, I'd be happy to resend it.
The midlayer stuff Im more than happy with
Okay.
--
Justin