Skip to contentSkip to conversation

QLogic drivers in 2.4

DISCUSSION Bjorn Helgaas 7 participants 14 messages Sep 19, 2003 – Sep 25, 2003

lore
Share & tools
lore
#0 lore raw

These drivers, and the ql2xxx stuff in particular, comprise one of the big pieces of the 2.4 ia64 patch that hasn't gone upstream yet:

 drivers/scsi/ql12160_fw.h                    | 3125 +++---
 drivers/scsi/ql1280_fw.h                     | 3774 ++++---
 drivers/scsi/ql2100_fw.h                     | 4740 +++++++++
 drivers/scsi/ql2200_fw.h                     | 5179 ++++++++++
 drivers/scsi/ql2300_fw.h                     | 5486 +++++++++++
 drivers/scsi/qla1280.c                       | 1399 +-
 drivers/scsi/qla1280.h                       |  297 
 drivers/scsi/qla2x00.c                       |12823 +++++++++++++++++++++++++++
 drivers/scsi/qla2x00.h                       | 2492 +++++
 drivers/scsi/qla2x00exioct.h                 |  823 +
 drivers/scsi/qla2x00exioctln.h               |  131 
 drivers/scsi/qla2x00ioctl.c                  | 1179 ++
 drivers/scsi/qlavendor.c                     |  325 

It looks like at least the ql2xxx stuff is (a) not in the 2.5 tree, and (b) not in Red Hat kernels.

So who cares about this stuff? If I dropped it out of the ia64 patch, would anybody be upset enough to help sort this out?

Bjorn

#1 lore raw
Quoted 2 lines

It looks like at least the ql2xxx stuff is (a) not in the 2.5 tree, and (b) not in Red Hat kernels.

Redhat AS2.1 and EL3 have at least qla2xxx and qla1280.

Quoted 1 line

So who cares about this stuff?

I do, :-)

#2 lore raw
In reply to Bjorn Helgaas
Quoted 4 lines · Kenneth W
Quoted 2 lines

It looks like at least the ql2xxx stuff is (a) not in the 2.5 tree, and (b) not in Red Hat kernels.

Redhat AS2.1 and EL3 have at least qla2xxx and qla1280.

$ l redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla*
redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla1280.c
redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla1280.h
$ l redhat/2.4.18-e.37/src/drivers/scsi/qla*
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.c
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.h

I see they have qla1280, but I don't see qla2xxx. Am I looking in the wrong place?

Quoted 3 lines
Quoted 1 line

So who cares about this stuff?

I do, :-)

Do you care enough to help get these changes upstream? I'm tired of carrying them in the ia64 patch.

Bjorn

#3 lore raw
In reply to Bjorn Helgaas
Quoted 6 lines · Bjorn Helgaas

It looks like at least the ql2xxx stuff is (a) not in the 2.5 tree, and (b) not in Red Hat kernels.

So who cares about this stuff? If I dropped it out of the ia64 patch, would anybody be upset enough to help sort this out?

The qla2x00 in your tree is utter crap, so please drop it. For qla1280 I still hope Jes pushes his latest to Marcelo..

#4 lore raw
In reply to Bjorn Helgaas
Quoted 11 lines
Quoted 1 line

Redhat AS2.1 and EL3 have at least qla2xxx and qla1280.

$ l redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla*
redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla1280.c
redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla1280.h
$ l redhat/2.4.18-e.37/src/drivers/scsi/qla*
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.c
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.h

I see they have qla1280, but I don't see qla2xxx. Am I looking in the wrong place?

drivers/addon

Bill

#5 lore raw
In reply to Bjorn Helgaas
Quoted 11 lines · Bill Nottingham
Quoted 9 lines
$ l redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla*
redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla1280.c
redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla1280.h
$ l redhat/2.4.18-e.37/src/drivers/scsi/qla*
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.c
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.h

I see they have qla1280, but I don't see qla2xxx. Am I looking in the wrong place?

drivers/addon

Note that for RHEL3 (and probably RHEL2.1, I'll need to check) we aren't using the qla2xxx out of the ia64 patch, but a separate driver from qlogic.

Bill

#6 lore raw
In reply to Bjorn Helgaas
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Bjorn" = Bjorn Helgaas <[email protected]> writes:

Jes actively maintains the qla1280 driver. If he hasn't pushed the latest bits to Marcelo yet, he should (hint!).

Bjorn> It looks like at least the ql2xxx stuff is (a) not in the 2.5 Bjorn> tree, and (b) not in Red Hat kernels.

The qla2xxx driver hasn't been submitted to Marcelo. For this reason Red Hat puts it in drivers/addon and not drivers/scsi.

I think the qla1280 is important enough (core SCSI on several boxes) that you should keep it around until it's synced with Marcelo.

But in general I agree with the principle of not having generic drivers in the IA-64 patch.

Signature · 2 lines
Martin K. Petersen	Wild Open Source, Inc.
[email protected]	http://www.wildopensource.com/
#7 lore raw
In reply to Bjorn Helgaas
Quoted 3 lines
Quoted 1 line

Redhat AS2.1 and EL3 have at least qla2xxx and qla1280.

$ l

redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla*

Quoted 1 line

redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla12 80.c

Quoted 1 line

redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla12 80.h

Quoted 6 lines
$ l redhat/2.4.18-e.37/src/drivers/scsi/qla*
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.c
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.h

I see they have qla1280, but I don't see qla2xxx. Am I looking in the wrong place?

Redhat stuff the qla2xxx under drivers/addon/qla2200

Quoted 2 lines

Do you care enough to help get these changes upstream? I'm tired of carrying them in the ia64 patch.

Probably not that far.

#8 lore raw
In reply to Bjorn Helgaas
Quoted 1 line · Bjorn Helgaas

These drivers, and the ql2xxx stuff in particular,

Yeah - they don't belong in the ia64-linux specific patch and it's probably (partially) my fault they are there.

Quoted 10 lines

comprise one of the big pieces of the 2.4 ia64 patch that hasn't gone upstream yet:

 drivers/scsi/ql12160_fw.h                    | 3125 +++---
 drivers/scsi/ql1280_fw.h                     | 3774 ++++---
 drivers/scsi/ql2100_fw.h                     | 4740 +++++++++
 drivers/scsi/ql2200_fw.h                     | 5179 ++++++++++
 drivers/scsi/ql2300_fw.h                     | 5486 +++++++++++
 drivers/scsi/qla1280.c                       | 1399 +-
 drivers/scsi/qla1280.h                       |  297 

I'd think Jes could(had?) submit(ted) qla1280 bits himself upstream.

Quoted 6 lines
 drivers/scsi/qla2x00.c                       |12823 +++++++++++++++++++++++++++
 drivers/scsi/qla2x00.h                       | 2492 +++++
 drivers/scsi/qla2x00exioct.h                 |  823 +
 drivers/scsi/qla2x00exioctln.h               |  131 
 drivers/scsi/qla2x00ioctl.c                  | 1179 ++
 drivers/scsi/qlavendor.c                     |  325 

aren't these the "old" isp2x00 FC drivers?
I see the "new" qla2xxx drivers in their own directory:
drivers/scsi/qla2xxx

in linux-ia64-2.4 bk tree.

Quoted 2 lines

It looks like at least the ql2xxx stuff is (a) not in the 2.5 tree,

I thought for 2.5.x someone was maintaining an 8.x qla FC driver which was in parallel with qla2xxx 6.x FC drivers maintained by qlogic. I'm not sure what happened here since I don't see it in David's linux-ia64-2.5 bk tree.

Quoted 1 line

and (b) not in Red Hat kernels.

Red Hat Advanced Server kernels have qla2xxx FC support someplace. I'll try to unpack a recent ia64 RHAS kernel and see where it's hiding if no one else knows off hand.

Quoted 1 line

So who cares about this stuff?

AFAIK, HP doesn't care about "old" FC qla drivers on ia64.
IIRC, i2000 support has qla1280 and HP should support that.
HP cares more about new qlogic FC drivers. e.g.:
scsi6 : QLogic QLA2312 PCI to Fibre Channel Host Adapter: bus 128 device 1 irq 62

	Firmware version:  3.01.18, Driver version 6.04.00

(HP will actually need a post release of 6.05)

Quoted 2 lines

If I dropped it out of the ia64 patch, would anybody be upset enough to help sort this out?

If necessary, I can organize help in-house.

thanks,
grant

#9 lore raw
In reply to Bjorn Helgaas

On a second thought, the qla2x00 driver in ia64 tree is so old that it doesn't add much value for the effort that people put in. Dropping it isn't going to doom the earth or something.

-----Original Message-----
From: Chen, Kenneth W
Sent: Friday, September 19, 2003 11:24 AM
To: 'Bjorn Helgaas'; [email protected]
Subject: RE: QLogic drivers in 2.4

Quoted 3 lines
Quoted 1 line

Redhat AS2.1 and EL3 have at least qla2xxx and qla1280.

$ l

redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla*

Quoted 1 line

redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla12 80.c

Quoted 1 line

redhat/rhel3-b2/usr/src/linux-2.4.21-1.1931.2.399.ent/drivers/scsi/qla12 80.h

Quoted 6 lines
$ l redhat/2.4.18-e.37/src/drivers/scsi/qla*
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.c
redhat/2.4.18-e.37/src/drivers/scsi/qla1280.h

I see they have qla1280, but I don't see qla2xxx. Am I looking in the wrong place?

Redhat stuff the qla2xxx under drivers/addon/qla2200

Quoted 2 lines

Do you care enough to help get these changes upstream? I'm tired of carrying them in the ia64 patch.

Probably not that far.

#10 lore raw
In reply to Bjorn Helgaas
Quoted 1 line · Grant Grundler

I'd think Jes could(had?) submit(ted) qla1280 bits himself upstream.

Last time I asked him (ok, that's a looong time ago) he didn't want to submit it for 2.4. His driver is in 2.5 though but is 2.4-compatible.

It would be really nice (TM) to see it in Marcelo's tree.

Quoted 4 lines

I thought for 2.5.x someone was maintaining an 8.x qla FC driver which was in parallel with qla2xxx 6.x FC drivers maintained by qlogic. I'm not sure what happened here since I don't see it in David's linux-ia64-2.5 bk tree.

Well, that someone is qlogic. But I don's see why it should go into the ia64 tree. Once qlogic got the driver in shape it and/or the feral driver should go into mainline. This is actually on the 2.6 should-fix list.

#11 lore raw
In reply to Bjorn Helgaas
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Christoph" = Christoph Hellwig <[email protected]> writes:

Christoph> The qla2x00 in your tree is utter crap, so please drop it. Christoph> For qla1280 I still hope Jes pushes his latest to Marcelo..

Thats my intention, the 1280 driver has gotten enough exposure in the 2.5 tree now that I think it's about ready for 2.4.

Jes

#12 lore raw
In reply to Bjorn Helgaas
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Christoph" = Christoph Hellwig <[email protected]> writes:

Christoph> Well, that someone is qlogic. But I don's see why it Christoph> should go into the ia64 tree. Once qlogic got the driver Christoph> in shape it and/or the feral driver should go into Christoph> mainline. This is actually on the 2.6 should-fix list.

Hi,

I am a little wary of pushing the Feral driver to mainstraim in it's current state (well at least the state it was in a few months ago when I last looked). It's register access macros suffers from being multi-OS wrappers and it got memory ordering and posted writes completely wrong. It's certainly fixable, but if we push it for mainline, I'd really like to see someone swing the club over it a bit and convince it to be a little more Linuxy.

I understand the heritage of the Feral driver, so please don't take this as a rant, it's merely an issue of what the target for the driver is.

Cheers,
Jes

#13 lore raw
In reply to Bjorn Helgaas
Quoted 1 line · Christoph Hellwig

Well, that someone is qlogic.

ah - even better.

Quoted 1 line

But I don's see why it should go into the ia64 tree.

It shouldn't - should be mainline submittal.

Quoted 2 lines

Once qlogic got the driver in shape it and/or the feral driver should go into mainline.

I exchanged email with Matthew Jacob <[email protected]> in Jan 2003 and data corruption occurred in < 10 minutes of functional testing. (Sorry Jes, I remembered wrong - it wasn't a panic).
I could add retesting this driver to my TODO list if someone thinks it's worthwhile.

feral.com driver stopped being interesting to me becuase the qlogic drivers have usually passed functional testing (found very few issues). I find stylistic "bugs" easier to fix than correctness issues...

hth,
grant