Status of linuxppc_2.5

Status of linuxppc_2.5

[Pantelis Antoniou]

[-- Attachment #1: Type: text/plain, Size: 349 bytes --]

Hi

I'm trying to compile linuxppc_2.5
and the following patch is needed to
make it compile.

The configuration is from
make pmac_config.

I didn't test it though, I don't have
a test pmac system, my intention is to
test it on a 8xx system.

BTW, are there any gotchas, or anything
I should pay attention moving
from 2.4 -> 2.5?

Regards

Pantelis


[-- Attachment #2: compile-2.5.diff --]
[-- Type: text/plain, Size: 3025 bytes --]

===== arch/ppc/kernel/traps.c 1.21 vs edited =====
--- 1.21/arch/ppc/kernel/traps.c        Mon Jan  6 09:02:42 2003
+++ edited/arch/ppc/kernel/traps.c      Wed Jan  8 09:23:43 2003
@@ -174,7 +174,7 @@
                               (*nip & 0x100)? "OUT to": "IN from",
                               regs->gpr[rb] - _IO_BASE, nip);
                        regs->msr |= MSR_RI;
-                       regs->nip = fixup;
+                       regs->nip = entry->fixup;
                        return;
                }
        }
===== drivers/block/swim3.c 1.22 vs edited =====
--- 1.22/drivers/block/swim3.c  Fri Dec  6 03:10:45 2002
+++ edited/drivers/block/swim3.c        Wed Jan  8 10:40:20 2003
@@ -26,6 +26,7 @@
 #include <linux/ioctl.h>
 #include <linux/blk.h>
 #include <linux/devfs_fs_kernel.h>
+#include <linux/interrupt.h>
 #include <asm/io.h>
 #include <asm/dbdma.h>
 #include <asm/prom.h>
===== drivers/macintosh/mediabay.c 1.10 vs edited =====
--- 1.10/drivers/macintosh/mediabay.c   Tue Nov 12 01:47:53 2002
+++ edited/drivers/macintosh/mediabay.c Wed Jan  8 10:53:47 2003
@@ -36,6 +36,7 @@
 #include <asm/keylargo.h>
 #include <linux/adb.h>
 #include <linux/pmu.h>
+#include <linux/ide.h>

 #ifdef CONFIG_PMAC_PBOOK
 static int mb_notify_sleep(struct pmu_sleep_notifier *self, int when);
@@ -599,7 +600,7 @@
                if (bay->cd_index >= 0) {
                        printk(KERN_DEBUG "Unregistering mb %d ide, index:%d\n", i,
                               bay->cd_index);
-                       ide_unregister(&ide_hwifs[bay->cd_index]);
+                       ide_unregister(bay->cd_index);
                        bay->cd_index = -1;
                }
                if (bay->cd_retry) {
===== drivers/usb/host/ohci-pci.c 1.6 vs edited =====
--- 1.6/drivers/usb/host/ohci-pci.c     Sun Dec  1 04:21:11 2002
+++ edited/drivers/usb/host/ohci-pci.c  Wed Jan  8 10:49:46 2003
@@ -18,6 +18,7 @@
 #include <asm/machdep.h>
 #include <asm/pmac_feature.h>
 #include <asm/pci-bridge.h>
+#include <asm/prom.h>
 #ifndef CONFIG_PM
 #      define CONFIG_PM
 #endif
===== fs/nfs/write.c 1.36 vs edited =====
--- 1.36/fs/nfs/write.c Sat Dec 21 08:29:02 2002
+++ edited/fs/nfs/write.c       Wed Jan  8 10:38:32 2003
@@ -828,7 +828,9 @@
        struct nfs_write_data   *data = (struct nfs_write_data *) task->tk_calldata;
        struct nfs_writeargs    *argp = &data->args;
        struct nfs_writeres     *resp = &data->res;
+#if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4)
        struct inode            *inode = data->inode;
+#endif
        struct nfs_page         *req;
        struct page             *page;

===== include/linux/ide.h 1.32 vs edited =====
--- 1.32/include/linux/ide.h    Sat Dec 28 21:09:38 2002
+++ edited/include/linux/ide.h  Wed Jan  8 10:47:16 2003
@@ -334,6 +334,9 @@
  */
 int ide_register_hw(hw_regs_t *hw, struct hwif_s **hwifp);

+/* XXX panto */
+void ide_unregister (unsigned int index);
+
 /*
  * Set up hw_regs_t structure before calling ide_register_hw (optional)
  */

Re: Status of linuxppc_2.5

[Boris Bezlaj]

On Wed, Jan 08, 2003 at 11:18:10AM +0200, Pantelis Antoniou wrote:

> BTW, are there any gotchas, or anything
> I should pay attention moving
> from 2.4 -> 2.5?

AFAIK, you need the new module-init-tools for loading/unloading modules.

I see there is a little patch for swim3 driver. Do you have any problems
with swim3 driver in 2.4 kernel when writing to floppy? What is the status
in v2.5?

--
With best regards,
			Boris B.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Benjamin Herrenschmidt]

On Wed, 2003-01-08 at 18:32, Boris Bezlaj wrote:
> On Wed, Jan 08, 2003 at 11:18:10AM +0200, Pantelis Antoniou wrote:
>
> > BTW, are there any gotchas, or anything
> > I should pay attention moving
> > from 2.4 -> 2.5?
>
> AFAIK, you need the new module-init-tools for loading/unloading modules.
>
> I see there is a little patch for swim3 driver. Do you have any problems
> with swim3 driver in 2.4 kernel when writing to floppy? What is the status
> in v2.5?

What patch ? I'm interested ;)

So far, swim3 in 2.5 is not up-to-date to new BIO semantics in 2.5. I've wanted
to do that for some time now but didn't find time yet.

Ben.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Boris Bezlaj]

On Wed, Jan 08, 2003 at 07:18:35PM +0100, Benjamin Herrenschmidt wrote:
>
> What patch ? I'm interested ;)

It just adds +#include <linux/interrupt.h> to swim3.c

> So far, swim3 in 2.5 is not up-to-date to new BIO semantics in 2.5. I've wanted
> to do that for some time now but didn't find time yet.

But does the writing work in 2.4.x ? For me it works flawlessly in 2.2.x,
but in 2.4 there are some problems with settling (only when writing; read
is fine), which im trying to figure out and fix.

--

With best regards,

		Boris B.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Benjamin Herrenschmidt]

On Wed, 2003-01-08 at 19:22, Boris Bezlaj wrote:
> On Wed, Jan 08, 2003 at 07:18:35PM +0100, Benjamin Herrenschmidt wrote:
> >
> > What patch ? I'm interested ;)
>
> It just adds +#include <linux/interrupt.h> to swim3.c
>
> > So far, swim3 in 2.5 is not up-to-date to new BIO semantics in 2.5. I've wanted
> > to do that for some time now but didn't find time yet.
>
> But does the writing work in 2.4.x ? For me it works flawlessly in 2.2.x,
> but in 2.4 there are some problems with settling (only when writing; read
> is fine), which im trying to figure out and fix.

Did didn't try that driver very hard lately. Please cc me any fix you
may have.

Note also that there is a darwin driver for swim3 floating around that
gets a lot of code from MacOS 9 and that appear to document a bunch
of black magic stuffs, so that may be a good source of information. I
think it's on mklinux.org


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Boris Bezlaj]

On Wed, Jan 08, 2003 at 07:34:22PM +0100, Benjamin Herrenschmidt wrote:
> >
> > But does the writing work in 2.4.x ? For me it works flawlessly in 2.2.x,
> > but in 2.4 there are some problems with settling (only when writing; read
> > is fine), which im trying to figure out and fix.
>
> Did didn't try that driver very hard lately. Please cc me any fix you
> may have.

None for now..didn't found the problem yet..i will let you know when
(if) i produce a fix.
>
> Note also that there is a darwin driver for swim3 floating around that
> gets a lot of code from MacOS 9 and that appear to document a bunch
> of black magic stuffs, so that may be a good source of information. I
> think it's on mklinux.org

It is..looks tasty too. I will look into it.

--

With best regards,

		Boris B.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Benjamin Herrenschmidt]

On Wed, 2003-01-08 at 23:15, Michel Lanners wrote:
> On   8 Jan, this message from Pantelis Antoniou echoed through cyberspace:
> > I'm trying to compile linuxppc_2.5
> > and the following patch is needed to
> > make it compile.
>
> Tried to compile my standard kernel compile a while back, and found the
> following things broken (didn't have time to try fixing yet...):
>
> - 53c94 Mac SCSI
> - Matroxfb
>
> I also didn't try very hard to compile in face of these problems; there
> may be others still.
>
> Sorry I can't provide more info, but maybe this discussion makes ore
> people interested in 2.5 status, where I think a number of PPC-related
> things need to be seriously looked at, hopefully before it is declared
> 2.6...

Yup, mostly drivers though. Part of this is me lacking time or beeing
too lazy (pick your preferred one) as I want to rework most of the pmac
drivers to better fit in the new model. Some of it is in progress
locally though, hopefully I'll push something a couple of weeks from
now.

Ben.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Michel Lanners]

On   8 Jan, this message from Pantelis Antoniou echoed through cyberspace:
> I'm trying to compile linuxppc_2.5
> and the following patch is needed to
> make it compile.

Tried to compile my standard kernel compile a while back, and found the
following things broken (didn't have time to try fixing yet...):

- 53c94 Mac SCSI
- Matroxfb

I also didn't try very hard to compile in face of these problems; there
may be others still.

Sorry I can't provide more info, but maybe this discussion makes ore
people interested in 2.5 status, where I think a number of PPC-related
things need to be seriously looked at, hopefully before it is declared
2.6...

Cheers

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Paul Mackerras]

Michel Lanners writes:

> Tried to compile my standard kernel compile a while back, and found the
> following things broken (didn't have time to try fixing yet...):
>
> - 53c94 Mac SCSI

I just pushed a fix (untested) for mac53c94 scsi to linuxppc-2.5.
If you try it out, let me know how it goes.

Paul.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Pantelis Antoniou]

Benjamin Herrenschmidt wrote:

>On Wed, 2003-01-08 at 18:32, Boris Bezlaj wrote:
>
>
>>On Wed, Jan 08, 2003 at 11:18:10AM +0200, Pantelis Antoniou wrote:
>>
>>
>>
>>>BTW, are there any gotchas, or anything
>>>I should pay attention moving
>>>from 2.4 -> 2.5?
>>>
>>>
>>AFAIK, you need the new module-init-tools for loading/unloading modules.
>>
>>I see there is a little patch for swim3 driver. Do you have any problems
>>with swim3 driver in 2.4 kernel when writing to floppy? What is the status
>>in v2.5?
>>
>
>What patch ? I'm interested ;)
>
>So far, swim3 in 2.5 is not up-to-date to new BIO semantics in 2.5. I've wanted
>to do that for some time now but didn't find time yet.
>
>Ben.
>
>
>
>
>
I'm sorry, the patch is just trivial stuff needed
to get it to compile. I don't even have hardware
that applies to swim3 driver. I just added the
include for the header and modified the call
to ide_unregister to use the new calling convention.

BTW ide_unregister was not exported in linux/ide.h
and so I added the prototype there. Maybe someone
should notify the maintainer of the IDE layer that
since this function is needed by the swim3 driver
it should be visible in the header?
Or we should use a new infrastructure?

As I mentioned since I don't have a power mac testing
system, I'm not willing to sacrifice my workstation
to the gods of 2.5 just yet.

What I want to do is to use 2.5 for my board which
uses a 8xx, and there I'm up against greater problems...
My major obstacle is the 8xx uart driver.

I'm not up to speed on 2.5 yet, but what
happened to DECLARE_TASK_QUEUE? Apparently alot of
things changed significantly.

Is there any document or whatever to help me
clean up the confusion?

Finally, if I am permitted to say, it appears that we are
a little behind in the 2.5 train. The release date
is getting closer, and I'm not sure that there
are any people using 2.5 in PPC systems.
IMHO we should pick up the pace a little bit.

Regards

Pantelis


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Brad Boyer]

On Thu, Jan 09, 2003 at 10:13:40AM +0200, Pantelis Antoniou wrote:
> I'm sorry, the patch is just trivial stuff needed
> to get it to compile. I don't even have hardware
> that applies to swim3 driver. I just added the
> include for the header and modified the call
> to ide_unregister to use the new calling convention.
>
> BTW ide_unregister was not exported in linux/ide.h
> and so I added the prototype there. Maybe someone
> should notify the maintainer of the IDE layer that
> since this function is needed by the swim3 driver
> it should be visible in the header?
> Or we should use a new infrastructure?

Is there a reason that the swim3 driver needs something
from the IDE layer? I haven't tried 2.5 on my 7600 yet,
but when I'm compiling a kernel for it, I never include
IDE, since the motherboard doesn't have IDE hardware.
Any machine old enough to use the swim3 driver is slow
enough without any help...

	Brad Boyer
	flar@allandria.com


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Status of linuxppc_2.5

[Michel Lanners]

On   9 Jan, this message from Paul Mackerras echoed through cyberspace:
>> Tried to compile my standard kernel compile a while back, and found the
>> following things broken (didn't have time to try fixing yet...):
>>
>> - 53c94 Mac SCSI
>
> I just pushed a fix (untested) for mac53c94 scsi to linuxppc-2.5.
> If you try it out, let me know how it goes.

Thanks, at least it compiles :-) I'll have to see how I can test 2.5
kernels since most of my partitions are on LVM1 and that one has just
been removed in 2.5...

Cheers

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/