Skip to contentSkip to conversation

gcc-crosssdk-initial can not compile. Get message: this target does not support --with-float

DISCUSSION Martin Ertsås 2 participants 7 messages Jun 4, 2012 – Jun 11, 2012

lore
Share & tools
lore
#0 lore raw

Hi.

When trying to compile meta-toolchain, it crashes during gcc-cross-initial with the error:

this target does not support --with-float
Valied --with options are: arch arch_32 arch_64 cpu cpu_32 cpu_64 tune tune_32 tune_64

The reason for this is a commit that happened 2012-05-01 in openembedded core, called "gcc: Grok for callconvention-hard to enable hard float". Reverting this patches makes it compile, and still I get the --mfloat-abi=hard when compiling with the resulting toolchain.

I'm compiling on a x64 Linux machine, with the target being a cortexa9 processor. I have the DEFAULTTUNE set to cortexa9hf-neon. What should the patch do, and what will happen if I disable it?

Best regard
Martin

#1 lore raw
Quoted 8 lines · Martin Ertsås

Hi.

When trying to compile meta-toolchain, it crashes during gcc-cross-initial with the error:

this target does not support --with-float
Valied --with options are: arch arch_32 arch_64 cpu cpu_32 cpu_64 tune tune_32 tune_64

can you post log.do_configure of this package somewhere ?

Quoted 17 lines

The reason for this is a commit that happened 2012-05-01 in openembedded core, called "gcc: Grok for callconvention-hard to enable hard float". Reverting this patches makes it compile, and still I get the --mfloat-abi=hard when compiling with the resulting toolchain.

I'm compiling on a x64 Linux machine, with the target being a cortexa9 processor. I have the DEFAULTTUNE set to cortexa9hf-neon. What should the patch do, and what will happen if I disable it?

Best regard
Martin

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#2 lore raw

Sure. Will do it first thing in the morning. Think this actually happens in compile though, for some reason. Do you want me to paste it and give a link, or is it ok with an attachment?

- Martin

Quoted 32 lines · Khem Raj
Quoted 8 lines · Martin Ertsås

Hi.

When trying to compile meta-toolchain, it crashes during gcc-cross-initial with the error:

this target does not support --with-float
Valied --with options are: arch arch_32 arch_64 cpu cpu_32 cpu_64 tune tune_32 tune_64

can you post log.do_configure of this package somewhere ?

Quoted 9 lines

The reason for this is a commit that happened 2012-05-01 in openembedded core, called "gcc: Grok for callconvention-hard to enable hard float". Reverting this patches makes it compile, and still I get the --mfloat-abi=hard when compiling with the resulting toolchain.

I'm compiling on a x64 Linux machine, with the target being a cortexa9 processor. I have the DEFAULTTUNE set to cortexa9hf-neon. What should the patch do, and what will happen if I disable it?

Quoted 7 lines

Best regard
Martin

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#3 lore raw
Quoted 1 line · Martin Ertsås

Sure. Will do it first thing in the morning. Think this actually happens

in compile though, for some reason. Do you want me to paste it and give a link, or is it ok with an attachment?

Paste in is ok

Quoted 6 lines

- Martin

Quoted 2 lines · Khem Raj

On Sun, Jun 3, 2012 at 11:49 PM, Martin Ertsås<[email protected]>

Quoted 40 lines · wrote:
Quoted 29 lines
Quoted 9 lines

Hi.

When trying to compile meta-toolchain, it crashes during gcc-cross-initial with the error:

this target does not support --with-float
Valied --with options are: arch arch_32 arch_64 cpu cpu_32 cpu_64 tune tune_32 tune_64

can you post log.do_configure of this package somewhere ?

Quoted 9 lines

The reason for this is a commit that happened 2012-05-01 in openembedded core, called "gcc: Grok for callconvention-hard to enable hard float". Reverting this patches makes it compile, and still I get the --mfloat-abi=hard when compiling with the resulting toolchain.

I'm compiling on a x64 Linux machine, with the target being a cortexa9 processor. I have the DEFAULTTUNE set to cortexa9hf-neon. What should the patch do, and what will happen if I disable it?

Quoted 7 lines

Best regard
Martin

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

Quoted 4 lines

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#4 lore raw

The configure log is at: http://pastebin.com/93h1rw7G
while the compile log, which is what actually failes, is at:
http://pastebin.com/2iXYrCDa

Be warned that the compile log is quite big, a little over 1700 lines, and do seem to actually be a configure step, not a compilation step. This really puzzles me.

Thanks for the help

- Martin

Quoted 51 lines · Khem Raj
Quoted 1 line · Martin Ertsås

Sure. Will do it first thing in the morning. Think this actually happens

in compile though, for some reason. Do you want me to paste it and give a link, or is it ok with an attachment?

Paste in is ok

Quoted 4 lines

- Martin

Quoted 1 line · Khem Raj

On Sun, Jun 3, 2012 at 11:49 PM, Martin Ertsås<[email protected]>

Quoted 31 lines · wrote:
Quoted 22 lines
Quoted 8 lines

Hi.

When trying to compile meta-toolchain, it crashes during gcc-cross-initial with the error:

this target does not support --with-float
Valied --with options are: arch arch_32 arch_64 cpu cpu_32 cpu_64 tune tune_32 tune_64

can you post log.do_configure of this package somewhere ?

Quoted 12 lines

The reason for this is a commit that happened 2012-05-01 in openembedded core, called "gcc: Grok for callconvention-hard to enable hard float". Reverting this patches makes it compile, and still I get the --mfloat-abi=hard when compiling with the resulting toolchain.

I'm compiling on a x64 Linux machine, with the target being a cortexa9 processor. I have the DEFAULTTUNE set to cortexa9hf-neon. What should the patch do, and what will happen if I disable it?

Best regard
Martin

Quoted 8 lines
Quoted 4 lines

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

Quoted 5 lines

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#5 lore raw

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

Quoted 4 lines · Martin Ertsås

The configure log is at: http://pastebin.com/93h1rw7G while the compile log, which is what actually failes, is at:
http://pastebin.com/2iXYrCDa

can you try this patch please

http://patchwork.openembedded.org/patch/29589/

and report if it fixes the problem.

- -Khem
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk/TzUwACgkQuwUzVZGdMxScmwCdH3cD2RwQGjgtZmn3V3oRVWcH r0UAn0wptXuapT0nHtBVz3IbLSoi5R2A
=zhhD
-----END PGP SIGNATURE-----

#6 lore raw

This fixed it. Thanks.

- Martin

Quoted 18 lines · Khem Raj
Quoted 3 lines · Martin Ertsås

The configure log is at: http://pastebin.com/93h1rw7G while the compile log, which is what actually failes, is at:
http://pastebin.com/2iXYrCDa

can you try this patch please

http://patchwork.openembedded.org/patch/29589/

and report if it fixes the problem.

-Khem

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel