[gentoo-catalyst] [PATCH 0/7] Pending branch patches (various)

[gentoo-catalyst] [PATCH 0/7] Pending branch patches (various)

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 936 bytes --]


These are several small patches fixing a few bugs nad upcoming 
fixes for the new portage ebuild (patches 3,4).  
Patch 6 fixes the setup_pkgmgr() adding the 'build' use flag 
to all stages calling the function.

Brian Dolbec (7):
  Remove unused urllib import.
  Remove unused variable new and an undefined variable s.
  stage1-controller.sh: Fix portage bin path hard coding
  chroot-functions.sh: Remove --nodeps option from portage update.
  chroot-functions.sh: Fix a mis-worded comment
  setup_pkgmgr(): WIP Make the 'build' use flag optional
  Fix a relative path bug

 catalyst/support.py                 | 12 ------------
 targets/stage1/stage1-chroot.sh     |  1 +
 targets/stage1/stage1-controller.sh |  6 +++---
 targets/stage2/stage2-chroot.sh     |  1 +
 targets/support/chroot-functions.sh | 15 +++++++++++----
 targets/support/functions.sh        |  2 +-
 6 files changed, 17 insertions(+), 20 deletions(-)

-- 
1.9.3

[gentoo-catalyst] [PATCH 1/7] Remove unused urllib import.

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 330 bytes --]


This urllib import was added in commit 64c16cae70da13de3c55d8555a2e4c5dcdf2fcad
to fix an issue, but it is not used.  So must have covered up the real bug.  Removing it now
that I've completed some testing and haven't come across anything I can attribute to it.
---
 catalyst/support.py | 4 ----
 1 file changed, 4 deletions(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Remove-unused-urllib-import.patch --]
[-- Type: text/x-patch; name="0001-Remove-unused-urllib-import.patch", Size: 404 bytes --]

diff --git a/catalyst/support.py b/catalyst/support.py
index 4fe4603..bc24130 100644
--- a/catalyst/support.py
+++ b/catalyst/support.py
@@ -23,10 +23,6 @@ except:
 # pids this process knows of.
 spawned_pids = []
 
-try:
-        import urllib
-except SystemExit, e:
-        raise
 
 def cleanup(pids,block_exceptions=True):
         """function to go through and reap the list of pids passed to it"""

[gentoo-catalyst] [PATCH 2/7] Remove unused variable new and an undefined variable s.

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 332 bytes --]


Commit b3475906d5f51a21ecaf4ff048002a2f44face52 introduced an
undefined variable "s".  The code must always be hitting the
general except: pass block.  Seems useless, if not maybe it'll
spit out a true error, so the real problem can be fixed.  Removed it all.
---
 catalyst/support.py | 8 --------
 1 file changed, 8 deletions(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0002-Remove-unused-variable-new-and-an-undefined-variable.patch --]
[-- Type: text/x-patch; name="0002-Remove-unused-variable-new-and-an-undefined-variable.patch", Size: 622 bytes --]

diff --git a/catalyst/support.py b/catalyst/support.py
index bc24130..d5dbfec 100644
--- a/catalyst/support.py
+++ b/catalyst/support.py
@@ -292,15 +292,7 @@ def spawn(mycommand,env={},raw_exit_code=False,opt_name=None,fd_pipes=None,retur
 				if trg_fd[x] == src_fd[x]:
 					continue
 				if trg_fd[x] in src_fd[x+1:]:
-					new=os.dup2(trg_fd[x],max(src_fd) + 1)
 					os.close(trg_fd[x])
-					try:
-						while True:
-							src_fd[s.index(trg_fd[x])]=new
-					except SystemExit, e:
-						raise
-					except:
-						pass
 
 			# transfer the fds to their final pre-exec position.
 			for x in range(0,len(trg_fd)):

[gentoo-catalyst] [PATCH 3/7] stage1-controller.sh: Fix portage bin path hard coding

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 225 bytes --]


Also rework to use find's filters instead of piping several times through grep.
Code supplied by: Douglas Freed <dwfreed>
---
 targets/stage1/stage1-controller.sh | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0003-stage1-controller.sh-Fix-portage-bin-path-hard-codin.patch --]
[-- Type: text/x-patch; name="0003-stage1-controller.sh-Fix-portage-bin-path-hard-codin.patch", Size: 699 bytes --]

diff --git a/targets/stage1/stage1-controller.sh b/targets/stage1/stage1-controller.sh
index 8dbd16b..1b2eee2 100755
--- a/targets/stage1/stage1-controller.sh
+++ b/targets/stage1/stage1-controller.sh
@@ -33,9 +33,9 @@ case $1 in
 		find . -iname "*.py[co]" -exec rm -f {} \;
 		# Cleanup all .a files except libgcc.a, *_nonshared.a and
 		# /usr/lib/portage/bin/*.a
-		find . -type f -iname "*.a" | grep -v 'libgcc.a' | \
-			grep -v 'nonshared.a' | grep -v '/usr/lib/portage/bin/' | \
-			grep -v 'libgcc_eh.a' | xargs rm -f
+		find . -path usr/lib/portage -prune -o -type f \
+			-iname '*.a' ! -name 'libgcc.a' ! -name '*nonshared.a' \
+			! -name 'libgcc_eh.a' -print | xargs rm -f
 	;;
 
 	*)

[gentoo-catalyst] [PATCH 4/7] chroot-functions.sh: Remove --nodeps option from portage update.

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 173 bytes --]


This option prevented new deps from being installed with a portage upgrade.
---
 targets/support/chroot-functions.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0004-chroot-functions.sh-Remove-nodeps-option-from-portag.patch --]
[-- Type: text/x-patch; name="0004-chroot-functions.sh-Remove-nodeps-option-from-portag.patch", Size: 596 bytes --]

diff --git a/targets/support/chroot-functions.sh b/targets/support/chroot-functions.sh
index 5c30537..ce56157 100755
--- a/targets/support/chroot-functions.sh
+++ b/targets/support/chroot-functions.sh
@@ -171,7 +171,7 @@ setup_pkgmgr(){
 	# just let emerge @system could merge it.
 	# Use --update or portage won't reinstall the same version.
 	[ -e /etc/portage/make.conf ] && echo 'USE="${USE} build"' >> /etc/portage/make.conf
-	run_merge --oneshot --nodeps --update sys-apps/portage
+	run_merge --oneshot --update sys-apps/portage
 	sed -i '/USE="${USE} build"/d' /etc/portage/make.conf
 }
 

[gentoo-catalyst] [PATCH 5/7] chroot-functions.sh: Fix a mis-worded comment

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 358 bytes --]


Fixes: 900554b0b067e76d3806e520357d91adbfa4fa0c
Author: Jorge Manuel B. S. Vicetto (jmbsvicetto) <jmbsvicetto@gentoo.org> (Mon 19 Dec 2011 11:14:05 PM PST)
subject: Make sure portage updates itself at the start of stage1 - thanks to Zac for noticing the issue.
---
 targets/support/chroot-functions.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0005-chroot-functions.sh-Fix-a-mis-worded-comment.patch --]
[-- Type: text/x-patch; name="0005-chroot-functions.sh-Fix-a-mis-worded-comment.patch", Size: 753 bytes --]

diff --git a/targets/support/chroot-functions.sh b/targets/support/chroot-functions.sh
index ce56157..3495f14 100755
--- a/targets/support/chroot-functions.sh
+++ b/targets/support/chroot-functions.sh
@@ -169,7 +169,7 @@ setup_pkgmgr(){
 	# We need to merge our package manager with USE="build" set in case it is
 	# portage to avoid frying our /etc/portage/make.conf file.  Otherwise, we could
 	# just let emerge @system could merge it.
-	# Use --update or portage won't reinstall the same version.
+	# Use --update or portage will reinstall the same version.
 	[ -e /etc/portage/make.conf ] && echo 'USE="${USE} build"' >> /etc/portage/make.conf
 	run_merge --oneshot --update sys-apps/portage
 	sed -i '/USE="${USE} build"/d' /etc/portage/make.conf

[gentoo-catalyst] [PATCH 6/7] setup_pkgmgr(): WIP Make the 'build' use flag optional

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 200 bytes --]

---
 targets/stage1/stage1-chroot.sh     |  1 +
 targets/stage2/stage2-chroot.sh     |  1 +
 targets/support/chroot-functions.sh | 13 ++++++++++---
 3 files changed, 12 insertions(+), 3 deletions(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0006-setup_pkgmgr-WIP-Make-the-build-use-flag-optional.patch --]
[-- Type: text/x-patch; name="0006-setup_pkgmgr-WIP-Make-the-build-use-flag-optional.patch", Size: 1758 bytes --]

diff --git a/targets/stage1/stage1-chroot.sh b/targets/stage1/stage1-chroot.sh
index ed83f38..7d8589f 100755
--- a/targets/stage1/stage1-chroot.sh
+++ b/targets/stage1/stage1-chroot.sh
@@ -21,6 +21,7 @@ then
 fi
 
 ## Setup seed pkgmgr to ensure latest
+export PKGMGR_BUILD='yes'
 clst_root_path=/ setup_pkgmgr
 
 # Update stage3
diff --git a/targets/stage2/stage2-chroot.sh b/targets/stage2/stage2-chroot.sh
index 61f5261..0f70a84 100755
--- a/targets/stage2/stage2-chroot.sh
+++ b/targets/stage2/stage2-chroot.sh
@@ -4,6 +4,7 @@ source /tmp/chroot-functions.sh
 
 # Setup the environment
 export FEATURES="${clst_myfeatures} nodoc noman noinfo -news"
+export PKGMGR_BUILD='yes'
 
 if [ "${clst_VERBOSE}" ]
 then
diff --git a/targets/support/chroot-functions.sh b/targets/support/chroot-functions.sh
index 3495f14..6260c69 100755
--- a/targets/support/chroot-functions.sh
+++ b/targets/support/chroot-functions.sh
@@ -170,9 +170,16 @@ setup_pkgmgr(){
 	# portage to avoid frying our /etc/portage/make.conf file.  Otherwise, we could
 	# just let emerge @system could merge it.
 	# Use --update or portage will reinstall the same version.
-	[ -e /etc/portage/make.conf ] && echo 'USE="${USE} build"' >> /etc/portage/make.conf
-	run_merge --oneshot --update sys-apps/portage
-	sed -i '/USE="${USE} build"/d' /etc/portage/make.conf
+	if [ -n "${PKGMGR_BUILD}" ];then
+		echo "Adding USE='${USE} build' to make.conf for portage build"
+		[ -e /etc/portage/make.conf ] && echo 'USE="${USE} build"' >> /etc/portage/make.conf
+		run_merge --oneshot --update sys-apps/portage
+		sed -i '/USE="${USE} build"/d' /etc/portage/make.conf
+	else
+		echo "Updating portage with normal USE="
+		run_merge --oneshot --update sys-apps/portage
+	fi
+
 }
 
 cleanup_distcc() {

[gentoo-catalyst] [PATCH 7/7] Fix a relative path bug

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 131 bytes --]


Conflicts:
	targets/support/functions.sh
---
 targets/support/functions.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0007-Fix-a-relative-path-bug.patch --]
[-- Type: text/x-patch; name="0007-Fix-a-relative-path-bug.patch", Size: 551 bytes --]

diff --git a/targets/support/functions.sh b/targets/support/functions.sh
index c8757d4..0894b35 100755
--- a/targets/support/functions.sh
+++ b/targets/support/functions.sh
@@ -34,7 +34,7 @@ exec_in_chroot(){
 
 	echo "Running ${file_name} in chroot:"
 	echo "    ${clst_CHROOT} ${chroot_path} ${destdir}/${file_name}"
-	${clst_CHROOT} ${chroot_path} ${destdir}/${file_name} || exit 1
+	${clst_CHROOT} ${chroot_path} .${destdir}/${file_name} || exit 1
 
 	delete_from_chroot ${destdir}/${file_name}
 	delete_from_chroot ${destdir}/chroot-functions.sh

Re: [gentoo-catalyst] [PATCH 3/7] stage1-controller.sh: Fix portage bin path hard coding

[Rick "Zero_Chaos" Farina]

[-- Attachment #1: Type: text/plain, Size: 465 bytes --]

On 09/01/2014 10:31 PM, Brian Dolbec wrote:
> 
> Also rework to use find's filters instead of piping several times through grep.
> Code supplied by: Douglas Freed <dwfreed>
> ---
>  targets/stage1/stage1-controller.sh | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
Was this patch tested and confirmed equivalent to the original in function?

From experience, nothing untested is correct no matter how correct it looks.

-Zero_Chaos


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]

[gentoo-catalyst] [PATCH v2] stage1-controller.sh: Remove some old poor cleaning code

[Brian Dolbec]

This code had portage bin path hard coded.  That path needed to be
changed for a new portage ebuild and install system.
After testing the origianl code and comparing it with some updated code
supplied by Douglas Freed.  It turned out both code chunks resulted in
nothing being cleaned.

Tested and confirmed by zero_chaos.
---
 targets/stage1/stage1-controller.sh | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/targets/stage1/stage1-controller.sh b/targets/stage1/stage1-controller.sh
index 8dbd16b..d029545 100755
--- a/targets/stage1/stage1-controller.sh
+++ b/targets/stage1/stage1-controller.sh
@@ -31,11 +31,6 @@ case $1 in
 		rm -rf usr/share/{man,doc,info}/*
 		# Zap all .pyc and .pyo files
 		find . -iname "*.py[co]" -exec rm -f {} \;
-		# Cleanup all .a files except libgcc.a, *_nonshared.a and
-		# /usr/lib/portage/bin/*.a
-		find . -type f -iname "*.a" | grep -v 'libgcc.a' | \
-			grep -v 'nonshared.a' | grep -v '/usr/lib/portage/bin/' | \
-			grep -v 'libgcc_eh.a' | xargs rm -f
 	;;
 
 	*)
-- 
1.9.3

Re: [gentoo-catalyst] [PATCH 7/7] Fix a relative path bug

[W. Trevor King]

[-- Attachment #1: Type: text/plain, Size: 533 bytes --]

On Mon, Sep 01, 2014 at 07:31:41PM -0700, Brian Dolbec wrote:
> Conflicts:
> 	targets/support/functions.sh

Presumably these are conflicts from rebasing from 3.0.  I don't think
we care about those, so I'd remove this note from the commit message.

What are the symptoms of the bug this fixes?  How do I trigger it in
Catalysts without this patch?

Cheers,
Trevor

-- 
This email may be signed or encrypted with GnuPG (http://www.gnupg.org).
For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-catalyst] [PATCH v2 3/7] stage1-controller.sh: Remove some old poor cleaning code

[Brian Dolbec]

On Mon,  1 Sep 2014 22:05:45 -0700
Brian Dolbec <dolsen@gentoo.org> wrote:

This code had portage bin path hard coded.  That path needed to be
changed for a new portage ebuild and install system.
After testing the origianl code and comparing it with some updated
code supplied by Douglas Freed.  It turned out both code chunks
resulted in nothing being cleaned.

Tested and confirmed by zero_chaos.
---
 targets/stage1/stage1-controller.sh | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/targets/stage1/stage1-controller.sh
b/targets/stage1/stage1-controller.sh index 8dbd16b..d029545 100755
--- a/targets/stage1/stage1-controller.sh
+++ b/targets/stage1/stage1-controller.sh
@@ -31,11 +31,6 @@ case $1 in
 		rm -rf usr/share/{man,doc,info}/*
 		# Zap all .pyc and .pyo files
 		find . -iname "*.py[co]" -exec rm -f {} \;
-		# Cleanup all .a files except libgcc.a, *_nonshared.a and
-		# /usr/lib/portage/bin/*.a
-		find . -type f -iname "*.a" | grep -v 'libgcc.a' | \
-			grep -v 'nonshared.a' | grep -v '/usr/lib/portage/bin/' | \
-			grep -v 'libgcc_eh.a' | xargs rm -f
 	;;
 
 	*)



-- 
Brian Dolbec <dolsen>

Re: [gentoo-catalyst] [PATCH 1/7] Remove unused urllib import.

[W. Trevor King]

[-- Attachment #1: Type: text/plain, Size: 1908 bytes --]

Previous submissions for this patch: [1,2].  It's already been acked
by Matt [3] and me [4].  I acked it again on IRC on 2014-06-15:

  08:02 < wking> dol-sen: I don't see how c718315 (Remove unused
    urllib import, 2013-02-11) could possibly break anything ;).
  08:02 < wking> But I'm going through stages one through three anyway
    just to be sure ;)
  08:03 < wking> Are you going to mail these to the list, or are you
    planning on polishing them first, and then submitting them to the
    list?
  08:03 <@dol-sen> they need some work
  …
  18:16 < wking> dol-sen: c718315 is now 135 of 263 in my stage 3, so
    I think it's good to merge to master (after floating on the list,
    if that's still Catalyst policy).
  21:50 <@jmbsvicetto> dol-sen / wking: please keep reviewing those
    patches with any help in the catalyst ml. I'll likely be tied for
    a while and won't have time to go over them, but I trust your
    judgement
  21:50 <@jmbsvicetto> dol-sen: feel free to commit them whenever
    you're happy.
  21:51 <@jmbsvicetto> Zero_Chaos: ^^ Please try to review them when
    you can to ensure you don't get any surprises
  21:52 <@jmbsvicetto> dol-sen: I can try to setup automated builds
    with the rewrite branch on my server so we can get automated
    feedback about it

So after three submissions and three acks (one on this exact c718315
commit), I think we can safely land this one ;).

Cheers,
Trevor

[1]: http://article.gmane.org/gmane.linux.gentoo.catalyst/2158
[2]: http://article.gmane.org/gmane.linux.gentoo.catalyst/2364
[3]: http://thread.gmane.org/gmane.linux.gentoo.catalyst/2361/focus=2378
[4]: http://thread.gmane.org/gmane.linux.gentoo.catalyst/2361/focus=2425

-- 
This email may be signed or encrypted with GnuPG (http://www.gnupg.org).
For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-catalyst] [PATCH 7/7] Fix a relative path bug

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 985 bytes --]

On Mon, 1 Sep 2014 22:06:34 -0700
"W. Trevor King" <wking@tremily.us> wrote:

> On Mon, Sep 01, 2014 at 07:31:41PM -0700, Brian Dolbec wrote:
> > Conflicts:
> > 	targets/support/functions.sh
> 
> Presumably these are conflicts from rebasing from 3.0.  I don't think
> we care about those, so I'd remove this note from the commit message.
> 
> What are the symptoms of the bug this fixes?  How do I trigger it in
> Catalysts without this patch?
> 
> Cheers,
> Trevor
> 

Yeah, sorry, I was leaving those in place while fixing rebase errors...
it helps to find them.  I'll clean them out once review is done.

This is only half the original problem.  The other half is fixed in a
previous commit.  What it orginally did was assign destdir with a
leading '.' then when I made some other seemingly unrelated change,
moving the '.' from the assignment to the usage fixed the relative path
issue.  It drove me nuts trying to find it.

-- 
Brian Dolbec <dolsen>


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 620 bytes --]

Re: [gentoo-catalyst] [PATCH v2] stage1-controller.sh: Remove some old poor cleaning code

[W. Trevor King]

[-- Attachment #1: Type: text/plain, Size: 604 bytes --]

On Mon, Sep 01, 2014 at 10:05:45PM -0700, Brian Dolbec wrote:
> Tested and confirmed by zero_chaos.

I haven't looked at the body of this commit yet, but Git standardized
this sort of note with Tested-by.  For example:

  Tested-by: Rick Zero_Chaos Farina <zerochaos@gentoo.org>

We don't need to follow that convention, but I think following
available conventions is a good idea unless we have a good reason not
to.

Cheers,
Trevor

-- 
This email may be signed or encrypted with GnuPG (http://www.gnupg.org).
For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-catalyst] [PATCH 2/7] Remove unused variable new and an undefined variable s.

[W. Trevor King]

[-- Attachment #1: Type: text/plain, Size: 802 bytes --]

This has already been submitted [1,2] and acked by Matt [3] and me
[4].  I'm testing now to double check.

Cheers,
Trevor

[1]: http://article.gmane.org/gmane.linux.gentoo.catalyst/2159
     id:1361956980.20292.12.camel@big_daddy.dol-sen.ca
[2]: http://article.gmane.org/gmane.linux.gentoo.catalyst/2362
     id:1386991708-9385-3-git-send-email-dolsen@gentoo.org
[3]: http://thread.gmane.org/gmane.linux.gentoo.catalyst/2361/focus=2378
     id:CAEdQ38GN1Co65i30U8peXbouwsK_dBF5k=YRq4d9WRbWCWEoPA@mail.gmail.com
[4]: http://thread.gmane.org/gmane.linux.gentoo.catalyst/2361/focus=2425
     id:20131214235358.GD25409@odin.tremily.us

-- 
This email may be signed or encrypted with GnuPG (http://www.gnupg.org).
For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-catalyst] [PATCH 2/7] Remove unused variable new and an undefined variable s.

[W. Trevor King]

[-- Attachment #1: Type: text/plain, Size: 474 bytes --]

On Mon, Sep 01, 2014 at 10:43:16PM -0700, W. Trevor King wrote:
> This has already been submitted [1,2] and acked by Matt [3] and me
> [4].  I'm testing now to double check.

I got this (c43de8722) through stage3 from scratch successfully, so
it can get both my Reviewed-by and Tested-by ;).

Cheers,
Trevor

-- 
This email may be signed or encrypted with GnuPG (http://www.gnupg.org).
For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-catalyst] [PATCH 7/7] Fix a relative path bug

[Brian Dolbec]

[-- Attachment #1: Type: text/plain, Size: 1620 bytes --]

On Mon, 1 Sep 2014 22:26:40 -0700
Brian Dolbec <dolsen@gentoo.org> wrote:

> On Mon, 1 Sep 2014 22:06:34 -0700
> "W. Trevor King" <wking@tremily.us> wrote:
> 
> > 
> > What are the symptoms of the bug this fixes?  How do I trigger it in
> > Catalysts without this patch?
> > 
> > Cheers,
> > Trevor
> > 
> 
> Yeah, sorry, I was leaving those in place while fixing rebase
> errors... it helps to find them.  I'll clean them out once review is
> done.
> 
> This is only half the original problem.  The other half is fixed in a
> previous commit.  What it orginally did was assign destdir with a
> leading '.' then when I made some other seemingly unrelated change,
> moving the '.' from the assignment to the usage fixed the relative
> path issue.  It drove me nuts trying to find it.
> 

Here is the full change, already in the 2.X branch and 2.0.17 release.

diff --git a/targets/support/functions.sh b/targets/support/functions.sh
index 311ed7b..fba855c 100755
@@ -20,7 +20,7 @@ exec_in_chroot(){
 # and executes it.
 	local file_name=$(basename ${1})
 	local subdir=${2}
-	local destdir=".${subdir}/tmp"
+	local destdir="${subdir}/tmp"
 
 	echo "Copying ${file_name} to ${destdir}"
 	copy_to_chroot ${1} ${destdir}
@@ -33,7 +33,7 @@ exec_in_chroot(){
 	chmod +x ${chroot_path}/${destdir}/${file_name}
 
 	echo "Running ${file_name} in chroot ${chroot_path}"
-	${clst_CHROOT} ${chroot_path} ${destdir}/${file_name} || exit 1
+	${clst_CHROOT} ${chroot_path} .${destdir}/${file_name} || exit
1 
 	delete_from_chroot ${destdir}/${file_name}
 

-- 
Brian Dolbec <dolsen>


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 605 bytes --]

Re: [gentoo-catalyst] [PATCH 6/7] setup_pkgmgr(): WIP Make the 'build' use flag optional

[Brian Dolbec]

On Mon,  1 Sep 2014 19:31:40 -0700
Brian Dolbec <dolsen@gentoo.org> wrote:

> ---
>  targets/stage1/stage1-chroot.sh     |  1 +
>  targets/stage2/stage2-chroot.sh     |  1 +
>  targets/support/chroot-functions.sh | 13 ++++++++++---
>  3 files changed, 12 insertions(+), 3 deletions(-)
> 

Here is an updated patch based on discussions with Rick.

From 52024ff70bf044f026330efd5a167139649d529a Mon Sep 17 00:00:00 2001
From: Brian Dolbec <dolsen@gentoo.org>
Date: Mon, 1 Sep 2014 15:10:38 -0700
Subject: [PATCH] setup_pkgmgr(): Make the 'build' use flag passed in
To: gentoo-catalyst@lists.gentoo.org

The "build" USE flag is only needed for the stage1 build.
It also causes other errors in later stages.
This makes setup_pkgmgr() takes an optional USE flag string
parameter to be added to the USE variable.
---
 targets/stage1/stage1-chroot.sh     |  2 +-
 targets/support/chroot-functions.sh | 13 ++++++++++---
 2 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/targets/stage1/stage1-chroot.sh b/targets/stage1/stage1-chroot.sh
index ed83f38..eccbd3f 100755
--- a/targets/stage1/stage1-chroot.sh
+++ b/targets/stage1/stage1-chroot.sh
@@ -21,7 +21,7 @@ then
 fi
 
 ## Setup seed pkgmgr to ensure latest
-clst_root_path=/ setup_pkgmgr
+clst_root_path=/ setup_pkgmgr "build"
 
 # Update stage3
 if [ -n "${clst_update_seed}" ]; then
diff --git a/targets/support/chroot-functions.sh b/targets/support/chroot-functions.sh
index 3495f14..0659ee0 100755
--- a/targets/support/chroot-functions.sh
+++ b/targets/support/chroot-functions.sh
@@ -170,9 +170,16 @@ setup_pkgmgr(){
 	# portage to avoid frying our /etc/portage/make.conf file.  Otherwise, we could
 	# just let emerge @system could merge it.
 	# Use --update or portage will reinstall the same version.
-	[ -e /etc/portage/make.conf ] && echo 'USE="${USE} build"' >> /etc/portage/make.conf
-	run_merge --oneshot --update sys-apps/portage
-	sed -i '/USE="${USE} build"/d' /etc/portage/make.conf
+	if [ -n "$1" ];then
+		echo "Adding USE='${USE} $1' to make.conf for portage build"
+		[ -e /etc/portage/make.conf ] && echo 'USE="${USE} $1"' >> /etc/portage/make.conf
+		run_merge --oneshot --update sys-apps/portage
+		sed -i '/USE="${USE} $1"/d' /etc/portage/make.conf
+	else
+		echo "Updating portage with USE='${USE}'"
+		run_merge --oneshot --update sys-apps/portage
+	fi
+
 }
 
 cleanup_distcc() {
-- 
2.1.0



-- 
Brian Dolbec <dolsen>

Re: [gentoo-catalyst] [PATCH v2] stage1-controller.sh: Remove some old poor cleaning code

[Brian Dolbec]

On Mon,  1 Sep 2014 22:05:45 -0700
Brian Dolbec <dolsen@gentoo.org> wrote:

> This code had portage bin path hard coded.  That path needed to be
> changed for a new portage ebuild and install system.
> After testing the origianl code and comparing it with some updated
> code supplied by Douglas Freed.  It turned out both code chunks
> resulted in nothing being cleaned.
> 
> Tested and confirmed by zero_chaos.

I have gone over things more and tested the new find command. It does
work on my host system.  However, the question remains... DOES this
particular cleaning operation NEED to be performed?

With current the tree snapshot for my testing 20140829.  It does not
find anything to clean.  BUT, will that remain the same in the future
as pkgs are bumped?

For safety, I'd be inclined to keep the find command (v1 of the
patch) and clean any it does find just in case.

-- 
Brian Dolbec <dolsen>

Re: [gentoo-catalyst] [PATCH v2] stage1-controller.sh: Remove some old poor cleaning code

[Rick "Zero_Chaos" Farina]

[-- Attachment #1: Type: text/plain, Size: 1496 bytes --]

On 09/09/2014 02:38 PM, Brian Dolbec wrote:
> On Mon,  1 Sep 2014 22:05:45 -0700
> Brian Dolbec <dolsen@gentoo.org> wrote:
> 
>> This code had portage bin path hard coded.  That path needed to be
>> changed for a new portage ebuild and install system.
>> After testing the origianl code and comparing it with some updated
>> code supplied by Douglas Freed.  It turned out both code chunks
>> resulted in nothing being cleaned.
>>
>> Tested and confirmed by zero_chaos.
> 
> I have gone over things more and tested the new find command. It does
> work on my host system.  However, the question remains... DOES this
> particular cleaning operation NEED to be performed?
> 
> With current the tree snapshot for my testing 20140829.  It does not
> find anything to clean.  BUT, will that remain the same in the future
> as pkgs are bumped?
> 
> For safety, I'd be inclined to keep the find command (v1 of the
> patch) and clean any it does find just in case.
> 

If we truly need to remove these files, I don't think catalyst was ever
the place to do this.  We have USE=static and USE=static-libs for a
reason.  Randomly removing files from the filesystem was a hack then,
and if we are cleaning this up let's just remove it.  If I want to build
things with USE=static or USE=static-libs then catalyst shouldn't be
pointlessly crippling my builds.

v2 means one less horrible hack in catalyst, and one at a time is the
only way will remove all the horror.

-Zero


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]

Re: [gentoo-catalyst] [PATCH v2] stage1-controller.sh: Remove some old poor cleaning code

[Anthony G. Basile]

On 09/10/14 13:16, Rick "Zero_Chaos" Farina wrote:
> On 09/09/2014 02:38 PM, Brian Dolbec wrote:
>> On Mon,  1 Sep 2014 22:05:45 -0700
>> Brian Dolbec <dolsen@gentoo.org> wrote:
>>
>>> This code had portage bin path hard coded.  That path needed to be
>>> changed for a new portage ebuild and install system.
>>> After testing the origianl code and comparing it with some updated
>>> code supplied by Douglas Freed.  It turned out both code chunks
>>> resulted in nothing being cleaned.
>>>
>>> Tested and confirmed by zero_chaos.
>>
>> I have gone over things more and tested the new find command. It does
>> work on my host system.  However, the question remains... DOES this
>> particular cleaning operation NEED to be performed?
>>
>> With current the tree snapshot for my testing 20140829.  It does not
>> find anything to clean.  BUT, will that remain the same in the future
>> as pkgs are bumped?
>>
>> For safety, I'd be inclined to keep the find command (v1 of the
>> patch) and clean any it does find just in case.
>>
>
> If we truly need to remove these files, I don't think catalyst was ever
> the place to do this.  We have USE=static and USE=static-libs for a
> reason.  Randomly removing files from the filesystem was a hack then,
> and if we are cleaning this up let's just remove it.  If I want to build
> things with USE=static or USE=static-libs then catalyst shouldn't be
> pointlessly crippling my builds.
>
> v2 means one less horrible hack in catalyst, and one at a time is the
> only way will remove all the horror.
>
> -Zero
>

I just did some testing here.  You have to be very careful not to remove 
the wrong .a's else you break the toolchain.  So I agree, remove the 
.a's in the ebuilds switched on USE=statlic-libs, and not in catalyst. 
The biggest .a's in stage1 are libpythonXX.a which are 3 to 4 MB, but 
the others are like a few hundred K's at most.  The gain is a smaller 
stage, but the downside is headaches making sure we don't break something.

So, to slim down stage1, see if you can get USE=static-libs in python's 
ebuilds and build USE=-static-libs.  That itself will reduce that stage 
by about 8 MBs which is more than all the "expendable" .a's put together.

-- 
Anthony G. Basile, Ph. D.
Chair of Information Technology
D'Youville College
Buffalo, NY 14201
(716) 829-8197

Re: [gentoo-catalyst] [PATCH 0/7] Pending branch patches (various)

[Brian Dolbec]

On Mon,  1 Sep 2014 19:31:34 -0700
Brian Dolbec <dolsen@gentoo.org> wrote:

> 
> These are several small patches fixing a few bugs nad upcoming 
> fixes for the new portage ebuild (patches 3,4).  
> Patch 6 fixes the setup_pkgmgr() adding the 'build' use flag 
> to all stages calling the function.
> 
> Brian Dolbec (7):
>   Remove unused urllib import.
>   Remove unused variable new and an undefined variable s.
>   stage1-controller.sh: Fix portage bin path hard coding
>   chroot-functions.sh: Remove --nodeps option from portage update.
>   chroot-functions.sh: Fix a mis-worded comment
>   setup_pkgmgr(): WIP Make the 'build' use flag optional
>   Fix a relative path bug
> 
>  catalyst/support.py                 | 12 ------------
>  targets/stage1/stage1-chroot.sh     |  1 +
>  targets/stage1/stage1-controller.sh |  6 +++---
>  targets/stage2/stage2-chroot.sh     |  1 +
>  targets/support/chroot-functions.sh | 15 +++++++++++----
>  targets/support/functions.sh        |  2 +-
>  6 files changed, 17 insertions(+), 20 deletions(-)
> 

Merged into master...

prepare for a bunch more... moohahahaha :)

-- 
Brian Dolbec <dolsen>