Hi,
I've reviewed the thread and can propose a solution.
Let's see e.g. the dev.s ( from fuse.ko ). Currently with gcc-4.2 we get:
fuse_req_init_context:
movl$_proxy_pda+8, %edx #, tmp62
#APP
movl %gs:8,%ecx #, ret__
#NO_APP
movl344(%ecx), %ecx # .fsuid,
Hi,
I've reviewed the thread and can propose a solution.
Let's see e.g. the dev.s ( from fuse.ko ). Currently with gcc-4.2 we get:
fuse_req_init_context:
movl$_proxy_pda+8, %edx #, tmp62
#APP
movl %gs:8,%ecx #, ret__
#NO_APP
movl344(%ecx), %ecx #
[sorry for the delay. My DSL line at home was AWOL]
On Saturday 09 December 2006 09:41, Andrew Morton wrote:
> > void __init quirk_intel_irqbalance(void)
>
> Yes, that boots. Andi: you want me to send it to Mr T?
I'm about to send it.
-Andi
-
To unsubscribe from this list: send the line
On Fri, 8 Dec 2006 10:00:04 -0800
"Siddha, Suresh B" <[EMAIL PROTECTED]> wrote:
> On Fri, Dec 08, 2006 at 06:10:29PM +0100, Andi Kleen wrote:
> > Yes please check the mainline git tree.
>
> Ok. I think I am the culprit :(
>
> Andi, Attached patch should fix the panic issue that Andrew
On Fri, 8 Dec 2006 10:00:04 -0800
Siddha, Suresh B [EMAIL PROTECTED] wrote:
On Fri, Dec 08, 2006 at 06:10:29PM +0100, Andi Kleen wrote:
Yes please check the mainline git tree.
Ok. I think I am the culprit :(
Andi, Attached patch should fix the panic issue that Andrew encountered.
[sorry for the delay. My DSL line at home was AWOL]
On Saturday 09 December 2006 09:41, Andrew Morton wrote:
void __init quirk_intel_irqbalance(void)
Yes, that boots. Andi: you want me to send it to Mr T?
I'm about to send it.
-Andi
-
To unsubscribe from this list: send the line
On Friday 08 December 2006 22:22, Andi Kleen wrote:
> The trouble is when it's CSEd it actually causes worse code because
> a register is tied up. That might not be worth the advantage of having it?
>
> Hmm, maybe marking it volatile would help? Arkadiusz, does the following
> patch help?
Andi Kleen wrote:
> The trouble is when it's CSEd it actually causes worse code because
> a register is tied up. That might not be worth the advantage of having it?
>
I think so, definitely; without proxy_pda you need to make it asm
volatile+mem clobber, which completely eliminates all
On Friday 08 December 2006 22:09, Jeremy Fitzhardinge wrote:
> Andi Kleen wrote:
> > Looking at Arkadiusz' output file it looks like gcc 4.2 decided to CSE the
> > address :/
> >
> > movl$_proxy_pda+8, %edx #, tmp65
> >
> > Very sad, but legitimate.
> >
>
> Yes, that was my
Andi Kleen wrote:
> Looking at Arkadiusz' output file it looks like gcc 4.2 decided to CSE the
> address :/
>
> movl$_proxy_pda+8, %edx #, tmp65
>
> Very sad, but legitimate.
>
Yes, that was my conclusion too. Though in this case the code could be
cleaned up by cutting down on
On Friday 08 December 2006 21:35, Jeremy Fitzhardinge wrote:
> Arkadiusz Miskiewicz wrote:
> > LD .tmp_vmlinux1
> > arch/i386/kernel/built-in.o: In function `math_emulate':
> > (.text+0x3809): undefined reference to `_proxy_pda'
> >
>
> Hm, in theory nothing should ever generate a
Andi Kleen wrote:
>> binutils-2.17.50.0.8-1.i686
>> gcc-4.2.0-0.20061206r119598.2.i686
>>
>
> Hmm, that's not even a release -- afaik gcc 4.2 isn't out yet.
>
> Can you please do
>
> make arch/i386/math-emu/fpu_entry.i
> make arch/i386/math-emu/fpu_entry.s
>
> and send me the resulting .i and
Arkadiusz Miskiewicz wrote:
> LD .tmp_vmlinux1
> arch/i386/kernel/built-in.o: In function `math_emulate':
> (.text+0x3809): undefined reference to `_proxy_pda'
>
Hm, in theory nothing should ever generate a reference to _proxy_pda.
What compiler are you using?
J
-
To unsubscribe
On Friday 08 December 2006 19:00, Siddha, Suresh B wrote:
> On Fri, Dec 08, 2006 at 06:10:29PM +0100, Andi Kleen wrote:
> > Yes please check the mainline git tree.
>
> Ok. I think I am the culprit :(
>
> Andi, Attached patch should fix the panic issue that Andrew encountered.
> Andrew, please
> binutils-2.17.50.0.8-1.i686
> gcc-4.2.0-0.20061206r119598.2.i686
Hmm, that's not even a release -- afaik gcc 4.2 isn't out yet.
Can you please do
make arch/i386/math-emu/fpu_entry.i
make arch/i386/math-emu/fpu_entry.s
and send me the resulting .i and .s files privately?
-Andi
-
To
On Fri, Dec 08, 2006 at 06:10:29PM +0100, Andi Kleen wrote:
> Yes please check the mainline git tree.
Ok. I think I am the culprit :(
Andi, Attached patch should fix the panic issue that Andrew encountered.
Andrew, please confirm.
Andi, if you are applying Ingo's genapic changes and reverting
On Friday 08 December 2006 19:04, Andi Kleen wrote:
> > > > Something related (git tree fetched 1-2h ago) ?
> > >
> > > Probably. Please send your .config.
> >
> > #
> > # Automatically generated make config: don't edit
> > # Linux kernel version: 2.6.19
> > # Fri Dec 8 11:40:15 2006
> > #
> >
> > >
> > > Something related (git tree fetched 1-2h ago) ?
> >
> > Probably. Please send your .config.
>
> #
> # Automatically generated make config: don't edit
> # Linux kernel version: 2.6.19
> # Fri Dec 8 11:40:15 2006
> #
> CONFIG_X86_32=y
I built your config and it builds fine here with
> My old 4-way Intel Nocona-based SDV panics during boot with "APIC mode must
> be flat on this system" and I don't know how to make it stop. Help.
Hmm, i had these patches for week and didn't change anything. Weird.
>
> It didn't do this with your tree in 2.6.19-rc6-mm1 or 2.6.19-rc6-mm2,
On Friday 08 December 2006 17:41, Siddha, Suresh B wrote:
> On Fri, Dec 08, 2006 at 02:08:04AM -0800, Andrew Morton wrote:
> > On Fri, 8 Dec 2006 04:01:25 +0100
> > Andi Kleen <[EMAIL PROTECTED]> wrote:
> >
> > > [The merge already made it to Linus' tree. Sorry for sending this message
> > >
On Fri, Dec 08, 2006 at 02:08:04AM -0800, Andrew Morton wrote:
> On Fri, 8 Dec 2006 04:01:25 +0100
> Andi Kleen <[EMAIL PROTECTED]> wrote:
>
> > [The merge already made it to Linus' tree. Sorry for sending this message
> > late]
> >
> > Most of this is for both i386 and x86-64, unless when noted
On Fri, Dec 08, 2006 at 02:03:12PM +0100, Arkadiusz Miskiewicz wrote:
> On Friday 08 December 2006 13:51, Muli Ben-Yehuda wrote:
> > On Fri, Dec 08, 2006 at 01:04:23PM +0100, Arkadiusz Miskiewicz wrote:
> > > On Friday 08 December 2006 04:01, Andi Kleen wrote:
> > > > - Support for a Processor
On Fri, Dec 08, 2006 at 01:04:23PM +0100, Arkadiusz Miskiewicz wrote:
> On Friday 08 December 2006 04:01, Andi Kleen wrote:
>
> > - Support for a Processor Data Area (PDA) on i386. This makes
> > the code more similar to x86-64 and will allow some other
> > optimizations in the future.
>
> LD
On Friday 08 December 2006 04:01, Andi Kleen wrote:
> - Support for a Processor Data Area (PDA) on i386. This makes
> the code more similar to x86-64 and will allow some other
> optimizations in the future.
LD .tmp_vmlinux1
arch/i386/kernel/built-in.o: In function `math_emulate':
* Andrew Morton <[EMAIL PROTECTED]> wrote:
> My old 4-way Intel Nocona-based SDV panics during boot with "APIC mode
> must be flat on this system" and I don't know how to make it stop.
> Help.
>
> It didn't do this with your tree in 2.6.19-rc6-mm1 or 2.6.19-rc6-mm2,
> both of which included
On Fri, 8 Dec 2006 04:01:25 +0100
Andi Kleen <[EMAIL PROTECTED]> wrote:
> [The merge already made it to Linus' tree. Sorry for sending this message
> late]
>
> Most of this is for both i386 and x86-64, unless when noted
>
> These are just some high lights. As usual there are more
> smaller
On Fri, 8 Dec 2006 04:01:25 +0100
Andi Kleen [EMAIL PROTECTED] wrote:
[The merge already made it to Linus' tree. Sorry for sending this message
late]
Most of this is for both i386 and x86-64, unless when noted
These are just some high lights. As usual there are more
smaller
* Andrew Morton [EMAIL PROTECTED] wrote:
My old 4-way Intel Nocona-based SDV panics during boot with APIC mode
must be flat on this system and I don't know how to make it stop.
Help.
It didn't do this with your tree in 2.6.19-rc6-mm1 or 2.6.19-rc6-mm2,
both of which included
On Friday 08 December 2006 04:01, Andi Kleen wrote:
- Support for a Processor Data Area (PDA) on i386. This makes
the code more similar to x86-64 and will allow some other
optimizations in the future.
LD .tmp_vmlinux1
arch/i386/kernel/built-in.o: In function `math_emulate':
On Fri, Dec 08, 2006 at 01:04:23PM +0100, Arkadiusz Miskiewicz wrote:
On Friday 08 December 2006 04:01, Andi Kleen wrote:
- Support for a Processor Data Area (PDA) on i386. This makes
the code more similar to x86-64 and will allow some other
optimizations in the future.
LD
On Fri, Dec 08, 2006 at 02:03:12PM +0100, Arkadiusz Miskiewicz wrote:
On Friday 08 December 2006 13:51, Muli Ben-Yehuda wrote:
On Fri, Dec 08, 2006 at 01:04:23PM +0100, Arkadiusz Miskiewicz wrote:
On Friday 08 December 2006 04:01, Andi Kleen wrote:
- Support for a Processor Data Area
On Fri, Dec 08, 2006 at 02:08:04AM -0800, Andrew Morton wrote:
On Fri, 8 Dec 2006 04:01:25 +0100
Andi Kleen [EMAIL PROTECTED] wrote:
[The merge already made it to Linus' tree. Sorry for sending this message
late]
Most of this is for both i386 and x86-64, unless when noted
These
On Friday 08 December 2006 17:41, Siddha, Suresh B wrote:
On Fri, Dec 08, 2006 at 02:08:04AM -0800, Andrew Morton wrote:
On Fri, 8 Dec 2006 04:01:25 +0100
Andi Kleen [EMAIL PROTECTED] wrote:
[The merge already made it to Linus' tree. Sorry for sending this message
late]
Most
My old 4-way Intel Nocona-based SDV panics during boot with APIC mode must
be flat on this system and I don't know how to make it stop. Help.
Hmm, i had these patches for week and didn't change anything. Weird.
It didn't do this with your tree in 2.6.19-rc6-mm1 or 2.6.19-rc6-mm2, both
of
Something related (git tree fetched 1-2h ago) ?
Probably. Please send your .config.
#
# Automatically generated make config: don't edit
# Linux kernel version: 2.6.19
# Fri Dec 8 11:40:15 2006
#
CONFIG_X86_32=y
I built your config and it builds fine here with gcc 4.1/binutils
On Friday 08 December 2006 19:04, Andi Kleen wrote:
Something related (git tree fetched 1-2h ago) ?
Probably. Please send your .config.
#
# Automatically generated make config: don't edit
# Linux kernel version: 2.6.19
# Fri Dec 8 11:40:15 2006
#
CONFIG_X86_32=y
I built
On Fri, Dec 08, 2006 at 06:10:29PM +0100, Andi Kleen wrote:
Yes please check the mainline git tree.
Ok. I think I am the culprit :(
Andi, Attached patch should fix the panic issue that Andrew encountered.
Andrew, please confirm.
Andi, if you are applying Ingo's genapic changes and reverting
binutils-2.17.50.0.8-1.i686
gcc-4.2.0-0.20061206r119598.2.i686
Hmm, that's not even a release -- afaik gcc 4.2 isn't out yet.
Can you please do
make arch/i386/math-emu/fpu_entry.i
make arch/i386/math-emu/fpu_entry.s
and send me the resulting .i and .s files privately?
-Andi
-
To
Arkadiusz Miskiewicz wrote:
LD .tmp_vmlinux1
arch/i386/kernel/built-in.o: In function `math_emulate':
(.text+0x3809): undefined reference to `_proxy_pda'
Hm, in theory nothing should ever generate a reference to _proxy_pda.
What compiler are you using?
J
-
To unsubscribe from
Andi Kleen wrote:
binutils-2.17.50.0.8-1.i686
gcc-4.2.0-0.20061206r119598.2.i686
Hmm, that's not even a release -- afaik gcc 4.2 isn't out yet.
Can you please do
make arch/i386/math-emu/fpu_entry.i
make arch/i386/math-emu/fpu_entry.s
and send me the resulting .i and .s files
On Friday 08 December 2006 21:35, Jeremy Fitzhardinge wrote:
Arkadiusz Miskiewicz wrote:
LD .tmp_vmlinux1
arch/i386/kernel/built-in.o: In function `math_emulate':
(.text+0x3809): undefined reference to `_proxy_pda'
Hm, in theory nothing should ever generate a reference to
Andi Kleen wrote:
Looking at Arkadiusz' output file it looks like gcc 4.2 decided to CSE the
address :/
movl$_proxy_pda+8, %edx #, tmp65
Very sad, but legitimate.
Yes, that was my conclusion too. Though in this case the code could be
cleaned up by cutting down on the
Andi Kleen wrote:
The trouble is when it's CSEd it actually causes worse code because
a register is tied up. That might not be worth the advantage of having it?
I think so, definitely; without proxy_pda you need to make it asm
volatile+mem clobber, which completely eliminates all
On Friday 08 December 2006 22:22, Andi Kleen wrote:
The trouble is when it's CSEd it actually causes worse code because
a register is tied up. That might not be worth the advantage of having it?
Hmm, maybe marking it volatile would help? Arkadiusz, does the following
patch help?
[The merge already made it to Linus' tree. Sorry for sending this message
late]
Most of this is for both i386 and x86-64, unless when noted
These are just some high lights. As usual there are more
smaller optimizations, cleanups etc
- paravirt support for i386: the basic hooks for replacing
[The merge already made it to Linus' tree. Sorry for sending this message
late]
Most of this is for both i386 and x86-64, unless when noted
These are just some high lights. As usual there are more
smaller optimizations, cleanups etc
- paravirt support for i386: the basic hooks for replacing
46 matches
Mail list logo
