Skip to contentSkip to conversation

anyone maintaining lkcd for ia64?

Linux-ia64 Robert K Gjertsen 4 participants 6 messages Feb 7, 2002 – Feb 8, 2002

lore
Share & tools
lore
#0 lore raw

I've used kdb in the past for ia64, but was taking a look at getting lkcd working on RH 7.2 (2.4.9-18/21) for ia64.
I grabbed the latest stuff from sourceforge.net and I'm working through the patched build now and see that dump_base.c is ia64 unfriendly with its use of irq_stat (using a hacked patch originally from lkcd-2.4.15-1.diff).
I'm slogging through this, but I hate to reinvent the wheel if someone else has a working copy/patch (newer kernel version is fine). Anyone actually using this for newer kernels on ia64 (and has SGI totally dumped working on this)?

Thanks,

--Rob

#1 lore raw

Hi Rob,

No, SGI has not dumped working on an ia64 version of LKCD. Piet Delaney is the engineer primarily working on getting the latest LKCD kernel code integrated and working with the latest Linux kernel rev. I've Cc'd him on this response. If you have any questions, issues, or suggestions regarding this effort, please contact Piet directly.

Thanks,

Tom

----- Original Message -----
From: "Robert K Gjertsen" <[email protected]>
To: <[email protected]>
Sent: Thursday, February 07, 2002 8:51 AM
Subject: [Linux-ia64] anyone maintaining lkcd for ia64?

Quoted 25 lines

I've used kdb in the past for ia64, but was taking a look at getting lkcd working on RH 7.2 (2.4.9-18/21) for ia64.
I grabbed the latest stuff from sourceforge.net and I'm working through the patched build now and see that dump_base.c is ia64 unfriendly with its use of irq_stat (using a hacked patch originally from lkcd-2.4.15-1.diff).
I'm slogging through this, but I hate to reinvent the wheel if someone else has a working copy/patch (newer kernel version is fine). Anyone actually using this for newer kernels on ia64 (and has SGI totally dumped working on this)?

Thanks,

--Rob

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

#2 lore raw
In reply to Robert K Gjertsen

On Thu, Feb 07, 2002 at 10:51:26AM -0600, Robert K Gjertsen wrote:

Hi Rob:

I'm about 90% done on getting a preliminary version of lkcd working for ia64. I hope to update the workspace shortly. I have to clean up some of the changes I made to support the dump page size being smaller that the virtual memory default page size.

I hope to work out a put today with the EFI memory reading by lcrash, how about I send you a copy tomarrow with some of the code cleaned up and I'll check it in after I've made sure I haven't brooken ia32.

-piet

Quoted 24 lines

I've used kdb in the past for ia64, but was taking a look at getting lkcd working on RH 7.2 (2.4.9-18/21) for ia64.
I grabbed the latest stuff from sourceforge.net and I'm working through the patched build now and see that dump_base.c is ia64 unfriendly with its use of irq_stat (using a hacked patch originally from lkcd-2.4.15-1.diff).
I'm slogging through this, but I hate to reinvent the wheel if someone else has a working copy/patch (newer kernel version is fine). Anyone actually using this for newer kernels on ia64 (and has SGI totally dumped working on this)?

Thanks,

--Rob

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

#3 lore raw
In reply to Robert K Gjertsen

Hi piet

Does the lkcd that you are working on support both SGI SN systems and DIG systems.

The main difference is the patch that is required on SGI systems to support discontiguous memory.

Also, we use XSCSI for disk access. Is that an issue??

Are you working on a pure IA64 patch

Quoted 46 lines · Robert K Gjertsen

Hi Rob:

I'm about 90% done on getting a preliminary version of lkcd working for ia64. I hope to update the workspace shortly. I have to clean up some of the changes I made to support the dump page size being smaller that the virtual memory default page size.

I hope to work out a put today with the EFI memory reading by lcrash, how about I send you a copy tomarrow with some of the code cleaned up and I'll check it in after I've made sure I haven't brooken ia32.

-piet

Quoted 17 lines

I've used kdb in the past for ia64, but was taking a look at getting lkcd working on RH 7.2 (2.4.9-18/21) for ia64.
I grabbed the latest stuff from sourceforge.net and I'm working through the patched build now and see that dump_base.c is ia64 unfriendly with its use of irq_stat (using a hacked patch originally from lkcd-2.4.15-1.diff).
I'm slogging through this, but I hate to reinvent the wheel if someone else has a working copy/patch (newer kernel version is fine). Anyone actually using this for newer kernels on ia64 (and has SGI totally dumped working on this)?

Thanks,

Quoted 7 lines

--Rob

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

Signature · 3 lines
Thanks

Jack Steiner    (651-683-5302)   (vnet 233-5302)      [email protected]
#4 lore raw
In reply to Robert K Gjertsen
Quoted 4 lines · Jack Steiner

Hi piet

Does the lkcd that you are working on support both SGI SN systems and DIG systems.

Most likely.

Quoted 3 lines

The main difference is the patch that is required on SGI systems to support discontiguous memory.

What patch?

Quoted 1 line

Also, we use XSCSI for disk access. Is that an issue??

Looks like XSCSI isn't configured in n my 2.4.16 workspace:

	#
	# Alternate SCSI support
	#
	# CONFIG_XSCSI is not set

however it appears to have been turned on in the 2.4.16-product gate. Sounds like it's a lot faster; does the root partition have to change to support XSCSI on the lab machines like monica?

I think I'll delay upgrading my Workspace till after I check in the current lkcd code. If it doesn't handle XSCSI I'll fix it.

BTW, how are things going on fixing kdb and linux so that breakpoints work in 2.4.17? I haven't heard anything on how that's going or the fix for the Machine Checks to non-existing memory that is mapped with the mmu.

Quoted 1 line

Are you working on a pure IA64 patch

I believe it's currently pure, I don't think I'm using anything from the sn1 directory. It's still in development, so we can change things based on what come up.

To get more work done in parallel it likely better to get working code in ASAP to avoid duplication of work.

-piet

Quoted 58 lines
Quoted 46 lines · Robert K Gjertsen

Hi Rob:

I'm about 90% done on getting a preliminary version of lkcd working for ia64. I hope to update the workspace shortly. I have to clean up some of the changes I made to support the dump page size being smaller that the virtual memory default page size.

I hope to work out a put today with the EFI memory reading by lcrash, how about I send you a copy tomarrow with some of the code cleaned up and I'll check it in after I've made sure I haven't brooken ia32.

-piet

Quoted 17 lines

I've used kdb in the past for ia64, but was taking a look at getting lkcd working on RH 7.2 (2.4.9-18/21) for ia64.
I grabbed the latest stuff from sourceforge.net and I'm working through the patched build now and see that dump_base.c is ia64 unfriendly with its use of irq_stat (using a hacked patch originally from lkcd-2.4.15-1.diff).
I'm slogging through this, but I hate to reinvent the wheel if someone else has a working copy/patch (newer kernel version is fine). Anyone actually using this for newer kernels on ia64 (and has SGI totally dumped working on this)?

Thanks,

Quoted 7 lines

--Rob

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

Signature · 9 lines
Thanks

Jack Steiner    (651-683-5302)   (vnet 233-5302)      [email protected]


_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64
#5 lore raw
In reply to Robert K Gjertsen
Quoted 7 lines · Jack Steiner
Quoted 4 lines

Hi piet

Does the lkcd that you are working on support both SGI SN systems and DIG systems.

Quoted 7 lines

Most likely.

Quoted 3 lines

The main difference is the patch that is required on SGI systems to support discontiguous memory.

What patch?

There is a discontig-path on sourceforge. That is the patch that is required to support SN systems.

If you are developing the code using 2.4.16-product, it should support SN systems. It may or may not support DIG systems. It depends on whether whether you have any dependencies on the SN code that is in the tree.

Quoted 5 lines
Quoted 2 lines

Also, we use XSCSI for disk access. Is that an issue??

Looks like XSCSI isn't configured in n my 2.4.16 workspace:

Quoted 8 lines
	#
	# Alternate SCSI support
	#
	# CONFIG_XSCSI is not set

however it appears to have been turned on in the 2.4.16-product gate. Sounds like it's a lot faster; does the root partition have to change to support XSCSI on the lab machines like monica?

It is marginally faster for most workloads. The question that I have is whether the lkcd code that writes the dump image will work with XSCSI. If there are any questions, Jeremy Higdon is the one you need to talk too.

You should probably enable XSCSI in your tree. That way, you will know for sure whether it will work.

Quoted 3 lines

I think I'll delay upgrading my Workspace till after I check in the current lkcd code. If it doesn't handle XSCSI I'll fix it.

When do you expect to do the checkin?? Do you have a TASK for lkcd yet (see linux.engr->1.6rc1)

Quoted 4 lines

BTW, how are things going on fixing kdb and linux so that breakpoints work in 2.4.17? I haven't heard anything on how that's going or the fix for the Machine Checks to non-existing memory that is mapped with the mmu.

Keith Owens should know....

Quoted 79 lines
Quoted 3 lines

Are you working on a pure IA64 patch

I believe it's currently pure, I don't think I'm using anything from the sn1 directory. It's still in development, so we can change things based on what come up.

To get more work done in parallel it likely better to get working code in ASAP to avoid duplication of work.

-piet

Quoted 41 lines
Quoted 41 lines · Robert K Gjertsen

Hi Rob:

I'm about 90% done on getting a preliminary version of lkcd working for ia64. I hope to update the workspace shortly. I have to clean up some of the changes I made to support the dump page size being smaller that the virtual memory default page size.

I hope to work out a put today with the EFI memory reading by lcrash, how about I send you a copy tomarrow with some of the code cleaned up and I'll check it in after I've made sure I haven't brooken ia32.

-piet

Quoted 24 lines

I've used kdb in the past for ia64, but was taking a look at getting lkcd working on RH 7.2 (2.4.9-18/21) for ia64.
I grabbed the latest stuff from sourceforge.net and I'm working through the patched build now and see that dump_base.c is ia64 unfriendly with its use of irq_stat (using a hacked patch originally from lkcd-2.4.15-1.diff).
I'm slogging through this, but I hate to reinvent the wheel if someone else has a working copy/patch (newer kernel version is fine). Anyone actually using this for newer kernels on ia64 (and has SGI totally dumped working on this)?

Thanks,

--Rob

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

Quoted 17 lines
Quoted 5 lines

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

Signature · 9 lines
Thanks

Jack Steiner    (651-683-5302)   (vnet 233-5302)      [email protected]


_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64

Signature · 3 lines
Thanks

Jack Steiner    (651-683-5302)   (vnet 233-5302)      [email protected]