Re: [oe-commits] Joshua Lock : packaged-staging: post-process non-binary files to fix paths packages

Re: [oe-commits] Joshua Lock : packaged-staging: post-process non-binary files to fix paths packages

[Tom Rini]

On Thu, 2010-04-01 at 12:51 +0000, git version control wrote:
> Module: openembedded.git
> Branch: rpurdie/work-in-progress
> Commit: 5df536641049c5af6413e410aa20bb97880fe897
> URL:    http://gitweb.openembedded.net/?p=openembedded.git&a=commit;h=5df536641049c5af6413e410aa20bb97880fe897
> 
> Author: Joshua Lock <josh@linux.intel.com>
> Date:   Wed Mar 31 11:22:02 2010 +0100
> 
> packaged-staging: post-process non-binary files to fix paths packages
> 
> Add tooling from Poky to scan packages and fix up paths in libtool files, we
> then build on this to further post-process native packages (native, cross, sdk)
> and fix up any references to the STAGING_DIR in non-binary files.
> 
> Signed-off-by: Joshua Lock <josh@linux.intel.com>

There's a lot more than just .la files that get these paths wrong.  If
you use the scan_cmd you set on native/etc for everything you can fix
them all up.

-- 
Tom Rini <tom_rini@mentor.com>
Mentor Graphics Corporation

Re: [oe-commits] Joshua Lock : packaged-staging: post-process non-binary files to fix paths packages

[Richard Purdie]

On Thu, 2010-04-01 at 09:05 -0700, Tom Rini wrote:
> On Thu, 2010-04-01 at 12:51 +0000, git version control wrote:
> > Module: openembedded.git
> > Branch: rpurdie/work-in-progress
> > Commit: 5df536641049c5af6413e410aa20bb97880fe897
> > URL:    http://gitweb.openembedded.net/?p=openembedded.git&a=commit;h=5df536641049c5af6413e410aa20bb97880fe897
> > 
> > Author: Joshua Lock <josh@linux.intel.com>
> > Date:   Wed Mar 31 11:22:02 2010 +0100
> > 
> > packaged-staging: post-process non-binary files to fix paths packages
> > 
> > Add tooling from Poky to scan packages and fix up paths in libtool files, we
> > then build on this to further post-process native packages (native, cross, sdk)
> > and fix up any references to the STAGING_DIR in non-binary files.
> > 
> > Signed-off-by: Joshua Lock <josh@linux.intel.com>
> 
> There's a lot more than just .la files that get these paths wrong.  If
> you use the scan_cmd you set on native/etc for everything you can fix
> them all up.

It scans everything for -native packages and just .la files for target
packages. I'd argue that for target packages we're better off fixing the
problems at source since the real target packages are probably also
bust?

Cheers,

Richard

Re: [oe-commits] Joshua Lock : packaged-staging: post-process non-binary files to fix paths packages

[Tom Rini]

On Thu, 2010-04-01 at 17:34 +0100, Richard Purdie wrote:
> On Thu, 2010-04-01 at 09:05 -0700, Tom Rini wrote:
> > On Thu, 2010-04-01 at 12:51 +0000, git version control wrote:
> > > Module: openembedded.git
> > > Branch: rpurdie/work-in-progress
> > > Commit: 5df536641049c5af6413e410aa20bb97880fe897
> > > URL:    http://gitweb.openembedded.net/?p=openembedded.git&a=commit;h=5df536641049c5af6413e410aa20bb97880fe897
> > > 
> > > Author: Joshua Lock <josh@linux.intel.com>
> > > Date:   Wed Mar 31 11:22:02 2010 +0100
> > > 
> > > packaged-staging: post-process non-binary files to fix paths packages
> > > 
> > > Add tooling from Poky to scan packages and fix up paths in libtool files, we
> > > then build on this to further post-process native packages (native, cross, sdk)
> > > and fix up any references to the STAGING_DIR in non-binary files.
> > > 
> > > Signed-off-by: Joshua Lock <josh@linux.intel.com>
> > 
> > There's a lot more than just .la files that get these paths wrong.  If
> > you use the scan_cmd you set on native/etc for everything you can fix
> > them all up.
> 
> It scans everything for -native packages and just .la files for target
> packages. I'd argue that for target packages we're better off fixing the
> problems at source since the real target packages are probably also
> bust?

Ah, hmm.  It's possibly just the .la files on the target (which some of
us would like to see die and I can't argue against), I hadn't really
bothered to differentiate (and I don't think it's a noticeable time
diff).  Less logic also brings it closer to working for relocation of
everything.

-- 
Tom Rini <tom_rini@mentor.com>
Mentor Graphics Corporation

Re: [oe-commits] Joshua Lock : packaged-staging: post-process non-binary files to fix paths packages

[Richard Purdie]

On Thu, 2010-04-01 at 09:55 -0700, Tom Rini wrote:
> On Thu, 2010-04-01 at 17:34 +0100, Richard Purdie wrote:
> > > There's a lot more than just .la files that get these paths wrong.  If
> > > you use the scan_cmd you set on native/etc for everything you can fix
> > > them all up.
> > 
> > It scans everything for -native packages and just .la files for target
> > packages. I'd argue that for target packages we're better off fixing the
> > problems at source since the real target packages are probably also
> > bust?
> 
> Ah, hmm.  It's possibly just the .la files on the target (which some of
> us would like to see die and I can't argue against), I hadn't really
> bothered to differentiate (and I don't think it's a noticeable time
> diff).  Less logic also brings it closer to working for relocation of
> everything.

But at the cost our target packages are most likely buggy. We accept
the .la files are a problem and there is a solution (libtool sysroot
support) but we're not there yet.

I'd like to see the other problems addressed, not brushed under the
carpet so to speak.

You can trivially make this work on all files in your
local.conf/distro.conf anyway. I'd not like to see that the default.

Cheers,

Richard

Re: [oe-commits] Joshua Lock : packaged-staging: post-process non-binary files to fix paths packages

[Tom Rini]

On Thu, 2010-04-01 at 19:12 +0100, Richard Purdie wrote:
> On Thu, 2010-04-01 at 09:55 -0700, Tom Rini wrote:
> > On Thu, 2010-04-01 at 17:34 +0100, Richard Purdie wrote:
> > > > There's a lot more than just .la files that get these paths wrong.  If
> > > > you use the scan_cmd you set on native/etc for everything you can fix
> > > > them all up.
> > > 
> > > It scans everything for -native packages and just .la files for target
> > > packages. I'd argue that for target packages we're better off fixing the
> > > problems at source since the real target packages are probably also
> > > bust?
> > 
> > Ah, hmm.  It's possibly just the .la files on the target (which some of
> > us would like to see die and I can't argue against), I hadn't really
> > bothered to differentiate (and I don't think it's a noticeable time
> > diff).  Less logic also brings it closer to working for relocation of
> > everything.
> 
> But at the cost our target packages are most likely buggy. We accept
> the .la files are a problem and there is a solution (libtool sysroot
> support) but we're not there yet.
> 
> I'd like to see the other problems addressed, not brushed under the
> carpet so to speak.
> 
> You can trivially make this work on all files in your
> local.conf/distro.conf anyway. I'd not like to see that the default.

My worry is that we're missing something important.  What level of
relocation torture have you put this through?  I'm thinking on if I can
merge what you've got into my version easily enough to give it a workout
over the weekend.

-- 
Tom Rini <tom_rini@mentor.com>
Mentor Graphics Corporation