Re: EABI pain and frustration!

Re: EABI pain and frustration!

[pspreadborough]

I'm working from the Angstrom distribution. 

I'm using the arm-2009q3-67-arm-none-linux-gnueabi compiler because that's 
the EABI compiler I had to hand. I started out with the intention of moving to 
an EABI environment (no OABI_COMPAT), however when my EABI ramfs did 
not work (kernel booted but barfed when executing init or sh) I decided to try 
and EABI executable in the OABI environment.

When compiling the "HelloWorld" I used the "-mabi=aapcs-linux" compiler option 
because that's what the kernel uses. 


Thanks,

Pete

----- Original Message ----- 
From: "Khem Raj" <raj.khem@gmail.com> 
To: openembedded-devel@lists.openembedded.org 
Sent: Wednesday, February 3, 2010 10:59:42 AM GMT -05:00 US/Canada Eastern 
Subject: Re: [oe] EABI pain and frustration! 

On (03/02/10 15:43), Koen Kooi wrote: 
> -----BEGIN PGP SIGNED MESSAGE----- 
> Hash: SHA1 
> 
> On 03-02-10 15:03, pspreadborough@comcast.net wrote: 
> > 
> > My 2.6.29 kernel is built with the CodeSourcery arm-2009q3-67-arm-none-linux-gnueabi 
> > compiler and has the EABI and OABI_COMPAT config options enabled. Using this kernel 
> > I can boot my SA1110 iPaq into a OABI ramfs with no problems, however if I build a 
> > "HelloWorld" executable with the same compiler and try and run it I get an "Illegal Instruction" 
> > error. I build the test program using static linking, select the correct machine type and tune 
> > options but to no avail, I always get the illegal instruction. 

hmm just tune and machine opts wont be enough you have to chose right abi 
options too. because your userland is OABI thats what you should chose for 
helloworld too. 

> 
> Which MACHINE and DISTRO are you using in OE? And also: why CSL? 
> -----BEGIN PGP SIGNATURE----- 
> Version: GnuPG v1.4.5 (Darwin) 
> 
> iD8DBQFLaYuoMkyGM64RGpERAiBXAJ9EKxw1hWGKcmJGo5LlEsDmHYJ/wACfZ1Id 
> 3Jz3mIP5xiMAez+V8l63TQ0= 
> =aaua 
> -----END PGP SIGNATURE----- 
> 
> 
> _______________________________________________ 
> Openembedded-devel mailing list 
> Openembedded-devel@lists.openembedded.org 
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel 

_______________________________________________ 
Openembedded-devel mailing list 
Openembedded-devel@lists.openembedded.org 
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel 

Re: EABI pain and frustration!

[Koen Kooi]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03-02-10 19:32, pspreadborough@comcast.net wrote:
> 
> I'm working from the Angstrom distribution. 
> 
> I'm using the arm-2009q3-67-arm-none-linux-gnueabi compiler

Why not use the default compiler for angstrom, that supports eabi on
strongarm for a while now:
http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=ee961dd30c417911edae2a0ba025e73c5a96958b
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFLacLHMkyGM64RGpERAkhJAKC2AhLXMJvkoeZL/Z3jE7ScIdB5PACgnvGH
YN0hP0ylyKluyZjJkTVL89E=
=R2bj
-----END PGP SIGNATURE-----

Re: EABI pain and frustration!

[pspreadborough]

----- "Travis Stratman" <tstratman@emacinc.com> wrote: 
> On Fri, 2010-02-05 at 13:41 -0800, Khem Raj wrote: 
> > On Fri, Feb 5, 2010 at 12:43 PM, Travis Stratman <tstratman@emacinc.com> wrote: 
> > > On Fri, 2010-02-05 at 10:00 -0800, Khem Raj wrote: 
> > >> On Fri, Feb 5, 2010 at 4:29 AM, Detlef Vollmann <dv@vollmann.ch> wrote: 
> > >> > On 02/03/10 15:03, pspreadborough@comcast.net wrote: 
> > >> >> 
> > >> >> My 2.6.29 kernel is built with the CodeSourcery 
> > >> >> arm-2009q3-67-arm-none-linux-gnueabi compiler and has the EABI and 
> > >> >> OABI_COMPAT config options enabled. Using this kernel I can boot my SA1110 
> > >> >> iPaq into a OABI ramfs with no problems, however if I build a "HelloWorld" 
> > >> >> executable with the same compiler and try and run it I get an "Illegal 
> > >> >> Instruction" error. I build the test program using static linking, select 
> > >> >> the correct machine type and tune options but to no avail, I always get the 
> > >> >> illegal instruction. 
> > >> 
> > >> You need gcc 4.4 to get EABI working on ARMv4 and there are additional patches 
> > >> to gcc on top of gcc 4.4 see 
> > >> gcc-armv4-pass-fix-v4bx-to-ld.patch for gcc in OE. So as Koen suggested earlier 
> > >> if you try with OE compiler you might have better chance of getting it 
> > >> to work as many 
> > >> OE devs have tried it. 
> > >> btw. are you passing march=armv4 to the compiler when doing the compile ? 
> > > 
> > > >From my understanding of this issue he has an OABI rootfs and is trying 
> > > to run an EABI executable on top of it. This just doesn't work period 
> > > AFAIK; the kernel can boot either OABI or EABI with the compatibility 
> > > flags config'd in, but all of the binaries on the filesystem must match. 
> > > 
> > > I'll have to look into the patch that you referenced above because I 
> > > have been using EABI (built through OE) for both the kernel and 
> > > filesystem with armv4t for a few years now. First w/ GCC 4.1.1 and now 
> > > with 4.2.4 and I've never run into issues. Does the issue only manifest 
> > > in certain situations? 
> > 
> > Well you said armv4t thats a whole different story than armv4 (without 
> > 't') for EABI 
> > AFAICT SA1110 is based on ARMv4 not ARMv4t. 
> 
> You are correct, my mistake. 

I have built a EABI rootfs but I get a kernel panic when it attempts
to run init, hence my attempts to verify my EABI executables on a OABI 
fs.

Perhaps it's time for me to try the OE EABI compiler. Can you advise me 
on which OE recipes to use for this?

Regards,

Pete 

Re: EABI pain and frustration!

[Khem Raj]

On Fri, Feb 5, 2010 at 5:37 PM,  <pspreadborough@comcast.net> wrote:
>
> ----- "Travis Stratman" <tstratman@emacinc.com> wrote:
>> On Fri, 2010-02-05 at 13:41 -0800, Khem Raj wrote:
>> > On Fri, Feb 5, 2010 at 12:43 PM, Travis Stratman <tstratman@emacinc.com> wrote:
>> > > On Fri, 2010-02-05 at 10:00 -0800, Khem Raj wrote:
>> > >> On Fri, Feb 5, 2010 at 4:29 AM, Detlef Vollmann <dv@vollmann.ch> wrote:
>> > >> > On 02/03/10 15:03, pspreadborough@comcast.net wrote:
>> > >> >>
>> > >> >> My 2.6.29 kernel is built with the CodeSourcery
>> > >> >> arm-2009q3-67-arm-none-linux-gnueabi compiler and has the EABI and
>> > >> >> OABI_COMPAT config options enabled. Using this kernel I can boot my SA1110
>> > >> >> iPaq into a OABI ramfs with no problems, however if I build a "HelloWorld"
>> > >> >> executable with the same compiler and try and run it I get an "Illegal
>> > >> >> Instruction" error. I build the test program using static linking, select
>> > >> >> the correct machine type and tune options but to no avail, I always get the
>> > >> >> illegal instruction.
>> > >>
>> > >> You need gcc 4.4 to get EABI working on ARMv4 and there are additional patches
>> > >> to gcc on top of gcc 4.4 see
>> > >> gcc-armv4-pass-fix-v4bx-to-ld.patch for gcc in OE. So as Koen suggested earlier
>> > >> if you try with OE compiler you might have better chance of getting it
>> > >> to work as many
>> > >> OE devs have tried it.
>> > >> btw. are you passing march=armv4 to the compiler when doing the compile ?
>> > >
>> > > >From my understanding of this issue he has an OABI rootfs and is trying
>> > > to run an EABI executable on top of it. This just doesn't work period
>> > > AFAIK; the kernel can boot either OABI or EABI with the compatibility
>> > > flags config'd in, but all of the binaries on the filesystem must match.
>> > >
>> > > I'll have to look into the patch that you referenced above because I
>> > > have been using EABI (built through OE) for both the kernel and
>> > > filesystem with armv4t for a few years now. First w/ GCC 4.1.1 and now
>> > > with 4.2.4 and I've never run into issues. Does the issue only manifest
>> > > in certain situations?
>> >
>> > Well you said armv4t thats a whole different story than armv4 (without
>> > 't') for EABI
>> > AFAICT SA1110 is based on ARMv4 not ARMv4t.
>>
>> You are correct, my mistake.
>
> I have built a EABI rootfs but I get a kernel panic when it attempts
> to run init, hence my attempts to verify my EABI executables on a OABI
> fs.
>
> Perhaps it's time for me to try the OE EABI compiler. Can you advise me
> on which OE recipes to use for this?

use 4.4.2 or 4.4.3

>
> Regards,
>
> Pete
>
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>

Re: EABI pain and frustration!

[pspreadborough]

----- "Khem Raj" <raj.khem@gmail.com> wrote: 

> > 
> > Perhaps it's time for me to try the OE EABI compiler. Can you advise me 
> > on which OE recipes to use for this? 
> 
> use 4.4.2 or 4.4.3 
> 

Mmmm, I don't see any 4.4.? targets:

bitbake -s

gcc                                      0:4.2.4-r3                     
gcc-canadian-sdk                         0:4.2.4-r2                     
gcc-cross                      0:4.3.2+csl-arm-2008q3-39-r1           0:4.2.4-r5
gcc-cross-initial                        0:4.2.4-r5                     
gcc-cross-intermediate                   0:4.2.4-r5                     
gcc-cross-kernel-3.3.3                   0:3.3.3-r6                     
gcc-cross-kernel-3.3.4                   0:3.3.4-r9                     
gcc-cross-kernel-3.4.4                  0:3.4.4-r13                     
gcc-cross-kernel-4.2.1         0:4.2.1+csl-arm-2007q3-53-r4                     
gcc-cross-kernel-4.2.4                   0:4.2.4-r5                     
gcc-cross-sdk                            0:4.2.4-r5                     
gcc-native                              0:3.4.6-r10                     
gcc-noemu                                0:4.1.2-r0 
                 
> > 
> > Regards, 
> > 
> > Pete 
> > 
> > 
> _______________________________________________ 
> Openembedded-devel mailing list 
> Openembedded-devel@lists.openembedded.org 
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel 
> 

Re: EABI pain and frustration!

[Khem Raj]

On Sat, Feb 6, 2010 at 8:38 AM,  <pspreadborough@comcast.net> wrote:
>
> ----- "Khem Raj" <raj.khem@gmail.com> wrote:
>
>> >
>> > Perhaps it's time for me to try the OE EABI compiler. Can you advise me
>> > on which OE recipes to use for this?
>>
>> use 4.4.2 or 4.4.3
>>
>
> Mmmm, I don't see any 4.4.? targets:
>

are you using .dev branch ?

> bitbake -s
>
> gcc                                      0:4.2.4-r3
> gcc-canadian-sdk                         0:4.2.4-r2
> gcc-cross                      0:4.3.2+csl-arm-2008q3-39-r1           0:4.2.4-r5
> gcc-cross-initial                        0:4.2.4-r5
> gcc-cross-intermediate                   0:4.2.4-r5
> gcc-cross-kernel-3.3.3                   0:3.3.3-r6
> gcc-cross-kernel-3.3.4                   0:3.3.4-r9
> gcc-cross-kernel-3.4.4                  0:3.4.4-r13
> gcc-cross-kernel-4.2.1         0:4.2.1+csl-arm-2007q3-53-r4
> gcc-cross-kernel-4.2.4                   0:4.2.4-r5
> gcc-cross-sdk                            0:4.2.4-r5
> gcc-native                              0:3.4.6-r10
> gcc-noemu                                0:4.1.2-r0
>
>> >
>> > Regards,
>> >
>> > Pete
>> >
>> >
>> _______________________________________________
>> Openembedded-devel mailing list
>> Openembedded-devel@lists.openembedded.org
>> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>>
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>

EABI pain and frustration!

[pspreadborough]

My 2.6.29 kernel is built with the CodeSourcery arm-2009q3-67-arm-none-linux-gnueabi 
compiler and has the EABI and OABI_COMPAT config options enabled. Using this kernel 
I can boot my SA1110 iPaq into a OABI ramfs with no problems, however if I build a 
"HelloWorld" executable with the same compiler and try and run it I get an "Illegal Instruction" 
error. I build the test program using static linking, select the correct machine type and tune 
options but to no avail, I always get the illegal instruction. 


Any help or suggestions would be greatly appreciated. 

Cheers, 

Pete 

Re: EABI pain and frustration!

[Koen Kooi]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03-02-10 15:03, pspreadborough@comcast.net wrote:
> 
> My 2.6.29 kernel is built with the CodeSourcery arm-2009q3-67-arm-none-linux-gnueabi 
> compiler and has the EABI and OABI_COMPAT config options enabled. Using this kernel 
> I can boot my SA1110 iPaq into a OABI ramfs with no problems, however if I build a 
> "HelloWorld" executable with the same compiler and try and run it I get an "Illegal Instruction" 
> error. I build the test program using static linking, select the correct machine type and tune 
> options but to no avail, I always get the illegal instruction. 

Which MACHINE and DISTRO are you using in OE? And also: why CSL?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFLaYuoMkyGM64RGpERAiBXAJ9EKxw1hWGKcmJGo5LlEsDmHYJ/wACfZ1Id
3Jz3mIP5xiMAez+V8l63TQ0=
=aaua
-----END PGP SIGNATURE-----

Re: EABI pain and frustration!

[Khem Raj]

On (03/02/10 15:43), Koen Kooi wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> On 03-02-10 15:03, pspreadborough@comcast.net wrote:
> > 
> > My 2.6.29 kernel is built with the CodeSourcery arm-2009q3-67-arm-none-linux-gnueabi 
> > compiler and has the EABI and OABI_COMPAT config options enabled. Using this kernel 
> > I can boot my SA1110 iPaq into a OABI ramfs with no problems, however if I build a 
> > "HelloWorld" executable with the same compiler and try and run it I get an "Illegal Instruction" 
> > error. I build the test program using static linking, select the correct machine type and tune 
> > options but to no avail, I always get the illegal instruction. 

hmm just tune and machine opts wont be enough you have to chose right abi
options too. because your userland is OABI thats what you should chose for
helloworld too.

> 
> Which MACHINE and DISTRO are you using in OE? And also: why CSL?
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (Darwin)
> 
> iD8DBQFLaYuoMkyGM64RGpERAiBXAJ9EKxw1hWGKcmJGo5LlEsDmHYJ/wACfZ1Id
> 3Jz3mIP5xiMAez+V8l63TQ0=
> =aaua
> -----END PGP SIGNATURE-----
> 
> 
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

Re: EABI pain and frustration!

[Detlef Vollmann]

On 02/03/10 15:03, pspreadborough@comcast.net wrote:
> My 2.6.29 kernel is built with the CodeSourcery arm-2009q3-67-arm-none-linux-gnueabi 
> compiler and has the EABI and OABI_COMPAT config options enabled. Using this kernel 
> I can boot my SA1110 iPaq into a OABI ramfs with no problems, however if I build a 
> "HelloWorld" executable with the same compiler and try and run it I get an "Illegal Instruction" 
> error. I build the test program using static linking, select the correct machine type and tune 
> options but to no avail, I always get the illegal instruction. 
Just checking: are you sure this is an EABI/OABI problem and not a
Thumb code problem?

   Detlef

Re: EABI pain and frustration!

[pspreadborough]

----- "Detlef Vollmann" <dv@vollmann.ch> wrote:

> On 02/03/10 15:03, pspreadborough@comcast.net wrote:
> > My 2.6.29 kernel is built with the CodeSourcery
> arm-2009q3-67-arm-none-linux-gnueabi 
> > compiler and has the EABI and OABI_COMPAT config options enabled.
> Using this kernel 
> > I can boot my SA1110 iPaq into a OABI ramfs with no problems,
> however if I build a 
> > "HelloWorld" executable with the same compiler and try and run it I
> get an "Illegal Instruction" 
> > error. I build the test program using static linking, select the
> correct machine type and tune 
> > options but to no avail, I always get the illegal instruction. 
> Just checking: are you sure this is an EABI/OABI problem and not a
> Thumb code problem?

I don't think so as I have the no-thumb compiler option enabled.

Re: EABI pain and frustration!

[Khem Raj]

On Fri, Feb 5, 2010 at 4:29 AM, Detlef Vollmann <dv@vollmann.ch> wrote:
> On 02/03/10 15:03, pspreadborough@comcast.net wrote:
>>
>> My 2.6.29 kernel is built with the CodeSourcery
>> arm-2009q3-67-arm-none-linux-gnueabi compiler and has the EABI and
>> OABI_COMPAT config options enabled. Using this kernel I can boot my SA1110
>> iPaq into a OABI ramfs with no problems, however if I build a "HelloWorld"
>> executable with the same compiler and try and run it I get an "Illegal
>> Instruction" error. I build the test program using static linking, select
>> the correct machine type and tune options but to no avail, I always get the
>> illegal instruction.

You need gcc 4.4 to get EABI working on ARMv4 and there are additional patches
to gcc on top of gcc 4.4 see
gcc-armv4-pass-fix-v4bx-to-ld.patch for gcc in OE. So as Koen suggested earlier
if you try with OE compiler you might have better chance of getting it
to work as many
OE devs have tried it.
btw. are you passing march=armv4 to the compiler when doing the compile ?

>
> Just checking: are you sure this is an EABI/OABI problem and not a
> Thumb code problem?
>
>  Detlef
>
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>

Re: EABI pain and frustration!

[Travis Stratman]

On Fri, 2010-02-05 at 10:00 -0800, Khem Raj wrote:
> On Fri, Feb 5, 2010 at 4:29 AM, Detlef Vollmann <dv@vollmann.ch> wrote:
> > On 02/03/10 15:03, pspreadborough@comcast.net wrote:
> >>
> >> My 2.6.29 kernel is built with the CodeSourcery
> >> arm-2009q3-67-arm-none-linux-gnueabi compiler and has the EABI and
> >> OABI_COMPAT config options enabled. Using this kernel I can boot my SA1110
> >> iPaq into a OABI ramfs with no problems, however if I build a "HelloWorld"
> >> executable with the same compiler and try and run it I get an "Illegal
> >> Instruction" error. I build the test program using static linking, select
> >> the correct machine type and tune options but to no avail, I always get the
> >> illegal instruction.
> 
> You need gcc 4.4 to get EABI working on ARMv4 and there are additional patches
> to gcc on top of gcc 4.4 see
> gcc-armv4-pass-fix-v4bx-to-ld.patch for gcc in OE. So as Koen suggested earlier
> if you try with OE compiler you might have better chance of getting it
> to work as many
> OE devs have tried it.
> btw. are you passing march=armv4 to the compiler when doing the compile ?

From my understanding of this issue he has an OABI rootfs and is trying
to run an EABI executable on top of it. This just doesn't work period
AFAIK; the kernel can boot either OABI or EABI with the compatibility
flags config'd in, but all of the binaries on the filesystem must match.

I'll have to look into the patch that you referenced above because I
have been using EABI (built through OE) for both the kernel and
filesystem with armv4t for a few years now. First w/ GCC 4.1.1 and now
with 4.2.4 and I've never run into issues. Does the issue only manifest
in certain situations?

Thanks,

TAS

Re: EABI pain and frustration!

[Khem Raj]

On Fri, Feb 5, 2010 at 12:43 PM, Travis Stratman <tstratman@emacinc.com> wrote:
> On Fri, 2010-02-05 at 10:00 -0800, Khem Raj wrote:
>> On Fri, Feb 5, 2010 at 4:29 AM, Detlef Vollmann <dv@vollmann.ch> wrote:
>> > On 02/03/10 15:03, pspreadborough@comcast.net wrote:
>> >>
>> >> My 2.6.29 kernel is built with the CodeSourcery
>> >> arm-2009q3-67-arm-none-linux-gnueabi compiler and has the EABI and
>> >> OABI_COMPAT config options enabled. Using this kernel I can boot my SA1110
>> >> iPaq into a OABI ramfs with no problems, however if I build a "HelloWorld"
>> >> executable with the same compiler and try and run it I get an "Illegal
>> >> Instruction" error. I build the test program using static linking, select
>> >> the correct machine type and tune options but to no avail, I always get the
>> >> illegal instruction.
>>
>> You need gcc 4.4 to get EABI working on ARMv4 and there are additional patches
>> to gcc on top of gcc 4.4 see
>> gcc-armv4-pass-fix-v4bx-to-ld.patch for gcc in OE. So as Koen suggested earlier
>> if you try with OE compiler you might have better chance of getting it
>> to work as many
>> OE devs have tried it.
>> btw. are you passing march=armv4 to the compiler when doing the compile ?
>
> >From my understanding of this issue he has an OABI rootfs and is trying
> to run an EABI executable on top of it. This just doesn't work period
> AFAIK; the kernel can boot either OABI or EABI with the compatibility
> flags config'd in, but all of the binaries on the filesystem must match.
>
> I'll have to look into the patch that you referenced above because I
> have been using EABI (built through OE) for both the kernel and
> filesystem with armv4t for a few years now. First w/ GCC 4.1.1 and now
> with 4.2.4 and I've never run into issues. Does the issue only manifest
> in certain situations?

Well you said armv4t thats a whole different story than armv4 (without
't') for EABI
AFAICT SA1110 is based on ARMv4 not ARMv4t.

>
> Thanks,
>
> TAS
>
>
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>

Re: EABI pain and frustration!

[Travis Stratman]

On Fri, 2010-02-05 at 13:41 -0800, Khem Raj wrote:
> On Fri, Feb 5, 2010 at 12:43 PM, Travis Stratman <tstratman@emacinc.com> wrote:
> > On Fri, 2010-02-05 at 10:00 -0800, Khem Raj wrote:
> >> On Fri, Feb 5, 2010 at 4:29 AM, Detlef Vollmann <dv@vollmann.ch> wrote:
> >> > On 02/03/10 15:03, pspreadborough@comcast.net wrote:
> >> >>
> >> >> My 2.6.29 kernel is built with the CodeSourcery
> >> >> arm-2009q3-67-arm-none-linux-gnueabi compiler and has the EABI and
> >> >> OABI_COMPAT config options enabled. Using this kernel I can boot my SA1110
> >> >> iPaq into a OABI ramfs with no problems, however if I build a "HelloWorld"
> >> >> executable with the same compiler and try and run it I get an "Illegal
> >> >> Instruction" error. I build the test program using static linking, select
> >> >> the correct machine type and tune options but to no avail, I always get the
> >> >> illegal instruction.
> >>
> >> You need gcc 4.4 to get EABI working on ARMv4 and there are additional patches
> >> to gcc on top of gcc 4.4 see
> >> gcc-armv4-pass-fix-v4bx-to-ld.patch for gcc in OE. So as Koen suggested earlier
> >> if you try with OE compiler you might have better chance of getting it
> >> to work as many
> >> OE devs have tried it.
> >> btw. are you passing march=armv4 to the compiler when doing the compile ?
> >
> > >From my understanding of this issue he has an OABI rootfs and is trying
> > to run an EABI executable on top of it. This just doesn't work period
> > AFAIK; the kernel can boot either OABI or EABI with the compatibility
> > flags config'd in, but all of the binaries on the filesystem must match.
> >
> > I'll have to look into the patch that you referenced above because I
> > have been using EABI (built through OE) for both the kernel and
> > filesystem with armv4t for a few years now. First w/ GCC 4.1.1 and now
> > with 4.2.4 and I've never run into issues. Does the issue only manifest
> > in certain situations?
> 
> Well you said armv4t thats a whole different story than armv4 (without
> 't') for EABI
> AFAICT SA1110 is based on ARMv4 not ARMv4t.

You are correct, my mistake.

Thanks,

TAS

Re: EABI pain and frustration!

[Phil Blundell]

On Fri, 2010-02-05 at 14:43 -0600, Travis Stratman wrote:
> >From my understanding of this issue he has an OABI rootfs and is trying
> to run an EABI executable on top of it. This just doesn't work period
> AFAIK; the kernel can boot either OABI or EABI with the compatibility
> flags config'd in, but all of the binaries on the filesystem must match.

I don't think this is strictly true: there shouldn't be any particular
reason why OABI and EABI binaries couldn't coexist on the same
filesystem.  Of course, you would need to have both sets of runtimes
installed (ld.so and libc at a mininum, plus whatever other libraries
are needed by your apps) so this would not be completely trivial to
arrange, but I can't think of anything that would impose a hard
requirement for everything on the system to be using the same ABI.

p.

Re: EABI pain and frustration!

[pspreadborough]

----- "Phil Blundell" <philb@gnu.org> wrote: 
> On Fri, 2010-02-05 at 14:43 -0600, Travis Stratman wrote: 
> > >From my understanding of this issue he has an OABI rootfs and is trying 
> > to run an EABI executable on top of it. This just doesn't work period 
> > AFAIK; the kernel can boot either OABI or EABI with the compatibility 
> > flags config'd in, but all of the binaries on the filesystem must match. 
> 
> I don't think this is strictly true: there shouldn't be any particular 
> reason why OABI and EABI binaries couldn't coexist on the same 
> filesystem. Of course, you would need to have both sets of runtimes 
> installed (ld.so and libc at a mininum, plus whatever other libraries 
> are needed by your apps) so this would not be completely trivial to 
> arrange, but I can't think of anything that would impose a hard 
> requirement for everything on the system to be using the same ABI. 

I used a statically linked executable in order to avoid having to install 
the support binaries. 

Pete