From anton at samba.org  Tue Nov  5 14:19:42 2002
From: anton at samba.org (Anton Blanchard)
Date: Tue, 5 Nov 2002 14:19:42 +1100
Subject: inline syscalls in glibc
Message-ID: <20021105031942.GA5190@krispykreme>


Hi,

Paul just pointed out to me that glibc 2.3 wont work on 2.5 ppc32
kernels because it has a broken inline syscall macro which doesnt
treat a syscall as a function call.

Steve: are we OK on the 64 bit port? In 2.5 Im looking at optimising
the syscall entry/exit like ppc32 and its going to be painful if it
will require an update to glibc just for that.

Anton

** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/



From amodra at bigpond.net.au  Tue Nov  5 15:29:54 2002
From: amodra at bigpond.net.au (Alan Modra)
Date: Tue, 5 Nov 2002 14:59:54 +1030
Subject: inline syscalls in glibc
In-Reply-To: <20021105031942.GA5190@krispykreme>
References: <20021105031942.GA5190@krispykreme>
Message-ID: <20021105042954.GC967@bubble.sa.bigpond.net.au>


On Tue, Nov 05, 2002 at 02:19:42PM +1100, Anton Blanchard wrote:
>
> Hi,
>
> Paul just pointed out to me that glibc 2.3 wont work on 2.5 ppc32
> kernels because it has a broken inline syscall macro which doesnt
> treat a syscall as a function call.
>
> Steve: are we OK on the 64 bit port? In 2.5 Im looking at optimising
> the syscall entry/exit like ppc32 and its going to be painful if it
> will require an update to glibc just for that.

See sysdeps/unix/sysv/linux/powerpc/powerpc64/syscall.h.  There's
a small matter of turning #if 1 into #if 0.

--
Alan Modra
IBM OzLabs - Linux Technology Centre

** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From benh at kernel.crashing.org  Tue Nov  5 21:25:14 2002
From: benh at kernel.crashing.org (Benjamin Herrenschmidt)
Date: Tue, 5 Nov 2002 11:25:14 +0100
Subject: inline syscalls in glibc
In-Reply-To: <20021105031942.GA5190@krispykreme>
References: <20021105031942.GA5190@krispykreme>
Message-ID: <20021105102514.26537@192.168.4.1>


>Paul just pointed out to me that glibc 2.3 wont work on 2.5 ppc32
>kernels because it has a broken inline syscall macro which doesnt
>treat a syscall as a function call.
>
>Steve: are we OK on the 64 bit port? In 2.5 Im looking at optimising
>the syscall entry/exit like ppc32 and its going to be painful if it
>will require an update to glibc just for that.

Doh ! Well, can we still fix glibc for ppc32 before anybody really
use it or it is really too late ?

Ben.


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From sjmunroe at us.ibm.com  Tue Nov 12 11:07:10 2002
From: sjmunroe at us.ibm.com (Steve Munroe)
Date: Mon, 11 Nov 2002 18:07:10 -0600
Subject: inline syscalls in glibc
Message-ID: <OF5B157CE9.94B84A4C-ON86256C6F.00000A7F@rchland.ibm.com>


The code is there but the #if 1 needs to change. We still need to run on
the 2.4 kernels. Perhaps change the:" #if 1 " to something based on
LINUX_VERSION_CODE?

I'll have to play with this a bit to be sure.




                      Anton Blanchard
                      <anton at samba.org>                   To:       Alan Modra <amodra at bigpond.net.au>
                      Sent by:                            cc:       linuxppc64-dev at lists.linuxppc.org
                      owner-linuxppc64-dev at lists.l        Subject:  Re: inline syscalls in glibc
                      inuxppc.org


                      12/31/01 07:10 AM






> See sysdeps/unix/sysv/linux/powerpc/powerpc64/syscall.h.  There's
> a small matter of turning #if 1 into #if 0.

Thanks Alan.

Steve: is this fixed in glibc CVS?

Anton


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From amodra at bigpond.net.au  Tue Nov 12 14:09:24 2002
From: amodra at bigpond.net.au (Alan Modra)
Date: Tue, 12 Nov 2002 13:39:24 +1030
Subject: inline syscalls in glibc
In-Reply-To: <OF5B157CE9.94B84A4C-ON86256C6F.00000A7F@rchland.ibm.com>
References: <OF5B157CE9.94B84A4C-ON86256C6F.00000A7F@rchland.ibm.com>
Message-ID: <20021112030924.GN997@bubble.sa.bigpond.net.au>


On Mon, Nov 11, 2002 at 06:07:10PM -0600, Steve Munroe wrote:
>
> The code is there but the #if 1 needs to change. We still need to run on
> the 2.4 kernels. Perhaps change the:" #if 1 " to something based on
> LINUX_VERSION_CODE?

No, it shouldn't depend on kernel version, as assuming that registers
are used when not will always be safe.  The reason I didn't enable the
code initially was that gcc-3.0.x (or maybe it was 3.1.x, I forget)
couldn't handle more than 10 asm args.  Whatever the version, I don't
want to support compilation of glibc with old gcc versions anyway..

--
Alan Modra
IBM OzLabs - Linux Technology Centre

** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From dist-lists at veronicas-therapy.com  Thu Nov 14 08:02:06 2002
From: dist-lists at veronicas-therapy.com (dist-lists at veronicas-therapy.com)
Date: Wed, 13 Nov 2002 15:02:06 -0600 (CST)
Subject: Is it possible to run Linux on a 7028-6E1 ?
Message-ID: <4339.205.182.74.247.1037221326.squirrel@www.e-securenetworks.net>


Has anyone had success in installing Linux on a RS6k 6E1 - pSeries 610 ?

Regards


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From jan at smith.pp.se  Thu Nov 14 08:56:19 2002
From: jan at smith.pp.se (Jan Smith)
Date: 13 Nov 2002 22:56:19 +0100
Subject: Is it possible to run Linux on a 7028-6E1 ?
In-Reply-To: <4339.205.182.74.247.1037221326.squirrel@www.e-securenetworks.net>
References: 
	 <4339.205.182.74.247.1037221326.squirrel@www.e-securenetworks.net>
Message-ID: <1037224579.2008.8.camel@frodo.smith.pp.se>


On Wed, 2002-11-13 at 22:02, dist-lists at veronicas-therapy.com wrote:
> Has anyone had success in installing Linux on a RS6k 6E1 - pSeries 610 ?

Yes I've installed both Redhat Linux v7.1 and YellowDog v2.1 and 2.2 on a 2CPU
pSeries 610 with 512Mb RAM. There is no problem with the main board but maybe
will there problems with a lot of expansion cards. I've heard one or two weeks
ago that IBM has started to support many of the PCI cards that weren't supported
earlier. Many of the graphics card is OK but I had problem with a SCSI-RAID
controller that I needed for a customer. We had to change our minds so that
machine are using software RAID instead of hardware RAID today.

Try it on it's not that hard to install Linux on most of the pSeries 6xx. I've
nowadays tried model 610 and 640 with success on both.

   /Jan Smith


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From aprasad at in.ibm.com  Sat Nov 16 00:27:39 2002
From: aprasad at in.ibm.com (Anil K Prasad)
Date: Fri, 15 Nov 2002 18:57:39 +0530
Subject: where do I find 2.5 series kernel for ppc64
Message-ID: <OF823634CD.E4C0077F-ON65256C72.00481608@in.ibm.com>


Hi All,

Where from I get latest 2.5 series linux kernel patch for ppc64?

Thanks,
Anil.


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/



From anton at samba.org  Sat Nov 16 04:27:18 2002
From: anton at samba.org (Anton Blanchard)
Date: Sat, 16 Nov 2002 04:27:18 +1100
Subject: where do I find 2.5 series kernel for ppc64
In-Reply-To: <OF823634CD.E4C0077F-ON65256C72.00481608@in.ibm.com>
References: <OF823634CD.E4C0077F-ON65256C72.00481608@in.ibm.com>
Message-ID: <20021115172718.GA4203@krispykreme>


Hi,

> Where from I get latest 2.5 series linux kernel patch for ppc64?

You can get the latest 2.5 tree via bitkeeper from
bk://ppc.bkbits.net/for-linus-ppc64

I regularly get Linus to update from there, so his tree sometimes
builds. However in the last few days a number of big changes (eg Rusty's
in kernel module linker) went into Linus' tree so its unlikely to compile.
I should have fixes for that in the next day or so.

There are a significant number of bugs in older toolchains. You will
need a recent toolchain, best bet is to grab cvs version of binutils and
the cvs 3.1 branch of gcc. Apply Alan Modra's latest gcc patch available
on ftp://ftp.linuxppc64.org/pub/people/amodra/gcc-3.2/

Apply the patch below to the kernel source tree. I still havent got
these bits merged into Linus' tree.

Anton

===== drivers/pci/probe.c 1.15 vs edited =====
--- 1.15/drivers/pci/probe.c	Sat Oct  5 11:06:13 2002
+++ edited/drivers/pci/probe.c	Sat Oct  5 11:47:03 2002
@@ -152,7 +152,7 @@
 		limit |= (io_limit_hi << 16);
 	}

-	if (base && base <= limit) {
+	if (base <= limit) {
 		res->flags = (io_base_lo & PCI_IO_RANGE_TYPE_MASK) | IORESOURCE_IO;
 		res->start = base;
 		res->end = limit + 0xfff;
===== include/linux/compiler.h 1.6 vs edited =====
--- 1.6/include/linux/compiler.h	Sun Sep  1 12:52:31 2002
+++ edited/include/linux/compiler.h	Thu Sep 19 17:40:30 2002
@@ -17,6 +17,6 @@
    shouldn't recognize the original var, and make assumptions about it */
 #define RELOC_HIDE(ptr, off)					\
   ({ unsigned long __ptr;					\
-    __asm__ ("" : "=g"(__ptr) : "0"(ptr));		\
+    __asm__ ("" : "=r"(__ptr) : "0"(ptr));		\
     (typeof(ptr)) (__ptr + (off)); })
 #endif /* __LINUX_COMPILER_H */


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/




From anton at samba.org  Sat Nov 16 05:12:24 2002
From: anton at samba.org (Anton Blanchard)
Date: Sat, 16 Nov 2002 05:12:24 +1100
Subject: where do I find 2.5 series kernel for ppc64
In-Reply-To: <20021115172718.GA4203@krispykreme>
References: <OF823634CD.E4C0077F-ON65256C72.00481608@in.ibm.com> <20021115172718.GA4203@krispykreme>
Message-ID: <20021115181224.GB4497@krispykreme>



> There are a significant number of bugs in older toolchains. You will
> need a recent toolchain, best bet is to grab cvs version of binutils and
> the cvs 3.1 branch of gcc. Apply Alan Modra's latest gcc patch available
          ^^^
> on ftp://ftp.linuxppc64.org/pub/people/amodra/gcc-3.2/

That should read 3.2.

Anton

** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/