* [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building
2012-10-31 1:55 [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Joe Hershberger
@ 2012-10-31 1:55 ` Joe Hershberger
2012-10-31 20:22 ` Simon Glass
2012-12-07 15:50 ` Tom Rini
2012-10-31 20:21 ` [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Simon Glass
` (3 subsequent siblings)
4 siblings, 2 replies; 14+ messages in thread
From: Joe Hershberger @ 2012-10-31 1:55 UTC (permalink / raw)
To: u-boot
--continue will allow you to <ctrl-c> the MAKEALL and pick up where
you left off.
--rebuild-errors will allow you to rebuild only those boards which
had trouble on the last run of MAKEALL, allowing you to quickly test
a simple fix on just those boards.
Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
---
MAKEALL | 41 +++++++++++++++++++++++++++++++++--------
1 file changed, 33 insertions(+), 8 deletions(-)
diff --git a/MAKEALL b/MAKEALL
index 1f88dc5..c7ae27d 100755
--- a/MAKEALL
+++ b/MAKEALL
@@ -20,6 +20,8 @@ usage()
-m, --maintainers List all targets and maintainer email
-M, --mails List all targets and all affilated emails
-C, --check Enable build checking
+ -n, --continue Continue (skip boards already built)
+ -r, --rebuild-errors Rebuild any boards that errored
-h, --help This help output
Selections by these options are logically ANDed; if the same option
@@ -52,8 +54,8 @@ usage()
exit ${ret}
}
-SHORT_OPTS="ha:c:v:s:lmMC"
-LONG_OPTS="help,arch:,cpu:,vendor:,soc:,list,maintainers,mails,check"
+SHORT_OPTS="ha:c:v:s:lmMCnr"
+LONG_OPTS="help,arch:,cpu:,vendor:,soc:,list,maintainers,mails,check,continue,rebuild-errors"
# Option processing based on util-linux-2.13/getopt-parse.bash
@@ -73,6 +75,8 @@ SELECTED=''
ONLY_LIST=''
PRINT_MAINTS=''
MAINTAINERS_ONLY=''
+CONTINUE=''
+REBUILD_ERRORS=''
while true ; do
case "$1" in
@@ -115,6 +119,12 @@ while true ; do
-C|--check)
CHECK='C=1'
shift ;;
+ -n|--continue)
+ CONTINUE='y'
+ shift ;;
+ -r|--rebuild-errors)
+ REBUILD_ERRORS='y'
+ shift ;;
-l|--list)
ONLY_LIST='y'
shift ;;
@@ -198,7 +208,9 @@ fi
OUTPUT_PREFIX="${BUILD_DIR}"
[ -d ${LOG_DIR} ] || mkdir "${LOG_DIR}" || exit 1
-find "${LOG_DIR}/" -type f -exec rm -f {} +
+if [ "$CONTINUE" != 'y' -a "$REBUILD_ERRORS" != 'y' ] ; then
+ find "${LOG_DIR}/" -type f -exec rm -f {} +
+fi
LIST=""
@@ -208,6 +220,7 @@ ERR_LIST=""
WRN_CNT=0
WRN_LIST=""
TOTAL_CNT=0
+SKIP_CNT=0
CURRENT_CNT=0
OLDEST_IDX=1
RC=0
@@ -728,12 +741,20 @@ build_targets() {
: $((CURRENT_CNT += 1))
rm -f "${donep}${TOTAL_CNT}"
rm -f "${skipp}${TOTAL_CNT}"
- if [ $BUILD_MANY == 1 ] ; then
- build_target ${t} ${TOTAL_CNT} &
+ if [ "$CONTINUE" = 'y' -a -e ${LOG_DIR}/$t.MAKELOG ] ; then
+ : $((SKIP_CNT += 1))
+ touch "${donep}${TOTAL_CNT}"
+ elif [ "$REBUILD_ERRORS" = 'y' -a ! -e ${LOG_DIR}/$t.ERR ] ; then
+ : $((SKIP_CNT += 1))
+ touch "${donep}${TOTAL_CNT}"
else
- CUR_TGT="${t}"
- build_target ${t} ${TOTAL_CNT}
- CUR_TGT=''
+ if [ $BUILD_MANY == 1 ] ; then
+ build_target ${t} ${TOTAL_CNT} &
+ else
+ CUR_TGT="${t}"
+ build_target ${t} ${TOTAL_CNT}
+ CUR_TGT=''
+ fi
fi
fi
@@ -786,8 +807,12 @@ print_stats() {
rm -f ${LOG_DIR}/${CUR_TGT}.MAKELOG ${LOG_DIR}/${CUR_TGT}.ERR
fi
+ : $((TOTAL_CNT -= ${SKIP_CNT}))
echo ""
echo "--------------------- SUMMARY ----------------------------"
+ if [ "$CONTINUE" = 'y' -o "$REBUILD_ERRORS" = 'y' ] ; then
+ echo "Boards skipped: ${SKIP_CNT}"
+ fi
echo "Boards compiled: ${TOTAL_CNT}"
if [ ${ERR_CNT} -gt 0 ] ; then
echo "Boards with errors: ${ERR_CNT} (${ERR_LIST} )"
--
1.7.11.5
^ permalink raw reply related [flat|nested] 14+ messages in thread* [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building
2012-10-31 1:55 ` [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building Joe Hershberger
@ 2012-10-31 20:22 ` Simon Glass
2012-12-07 15:50 ` Tom Rini
1 sibling, 0 replies; 14+ messages in thread
From: Simon Glass @ 2012-10-31 20:22 UTC (permalink / raw)
To: u-boot
On Tue, Oct 30, 2012 at 6:55 PM, Joe Hershberger <joe.hershberger@ni.com> wrote:
> --continue will allow you to <ctrl-c> the MAKEALL and pick up where
> you left off.
>
> --rebuild-errors will allow you to rebuild only those boards which
> had trouble on the last run of MAKEALL, allowing you to quickly test
> a simple fix on just those boards.
>
> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
> ---
> MAKEALL | 41 +++++++++++++++++++++++++++++++++--------
> 1 file changed, 33 insertions(+), 8 deletions(-)
>
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building
2012-10-31 1:55 ` [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building Joe Hershberger
2012-10-31 20:22 ` Simon Glass
@ 2012-12-07 15:50 ` Tom Rini
1 sibling, 0 replies; 14+ messages in thread
From: Tom Rini @ 2012-12-07 15:50 UTC (permalink / raw)
To: u-boot
On Tue, Oct 30, 2012 at 08:55:21PM -0500, Joe Hershberger wrote:
> --continue will allow you to <ctrl-c> the MAKEALL and pick up where
> you left off.
>
> --rebuild-errors will allow you to rebuild only those boards which
> had trouble on the last run of MAKEALL, allowing you to quickly test
> a simple fix on just those boards.
>
> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Applied to u-boot/master, thanks!
--
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20121207/096e6355/attachment.pgp>
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-10-31 1:55 [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Joe Hershberger
2012-10-31 1:55 ` [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building Joe Hershberger
@ 2012-10-31 20:21 ` Simon Glass
2012-11-01 2:32 ` Marek Vasut
` (2 subsequent siblings)
4 siblings, 0 replies; 14+ messages in thread
From: Simon Glass @ 2012-10-31 20:21 UTC (permalink / raw)
To: u-boot
On Tue, Oct 30, 2012 at 6:55 PM, Joe Hershberger <joe.hershberger@ni.com> wrote:
> When building in parallel, make sure that we look up the children
> based on the the actual process group id instead of just assuming
> that the MAKEALL pid is the process group id.
>
> Also ensure that logs from incomplete builds are deleted in the
> process.
>
> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
> ---
> MAKEALL | 26 ++++++++++++++++++++++++--
> 1 file changed, 24 insertions(+), 2 deletions(-)
>
> diff --git a/MAKEALL b/MAKEALL
> index 84a5c92..1f88dc5 100755
> --- a/MAKEALL
> +++ b/MAKEALL
> @@ -610,6 +610,13 @@ list_target() {
> donep="${LOG_DIR}/._done_"
> skipp="${LOG_DIR}/._skip_"
>
> +build_target_killed() {
> + echo "Aborted $target build."
> + # Remove the logs for this board since it was aborted
> + rm -f ${LOG_DIR}/$target.MAKELOG ${LOG_DIR}/$target.ERR
> + exit
> +}
> +
> build_target() {
> target=$1
> build_idx=$2
> @@ -622,6 +629,7 @@ build_target() {
> if [ $BUILD_MANY == 1 ] ; then
> output_dir="${OUTPUT_PREFIX}/${target}"
> mkdir -p "${output_dir}"
> + trap build_target_killed TERM
> else
> output_dir="${OUTPUT_PREFIX}"
> fi
> @@ -640,6 +648,8 @@ build_target() {
> fi
>
> if [ $BUILD_MANY == 1 ] ; then
> + trap - TERM
> +
> ${MAKE} -s tidy
>
> if [ -s ${LOG_DIR}/${target}.ERR ] ; then
> @@ -721,7 +731,9 @@ build_targets() {
> if [ $BUILD_MANY == 1 ] ; then
> build_target ${t} ${TOTAL_CNT} &
> else
> + CUR_TGT="${t}"
> build_target ${t} ${TOTAL_CNT}
> + CUR_TGT=''
> fi
> fi
>
> @@ -745,7 +757,11 @@ build_targets() {
> #-----------------------------------------------------------------------
>
> kill_children() {
> - kill -- "-$1"
> + local pgid=`ps -p $$ --no-headers -o "%r" | tr -d ' '`
> + local children=`pgrep -g $pgid | grep -v $$ | grep -v $pgid`
> +
> + kill $children 2> /dev/null
> + wait $children 2> /dev/null
>
> exit
> }
> @@ -753,6 +769,9 @@ kill_children() {
> print_stats() {
> if [ "$ONLY_LIST" == 'y' ] ; then return ; fi
>
> + # Only count boards that completed
> + : $((TOTAL_CNT = `find ${skipp}* 2> /dev/null | wc -l`))
> +
> rm -f ${donep}* ${skipp}*
>
> if [ $BUILD_MANY == 1 ] && [ -e "${OUTPUT_PREFIX}/ERR" ] ; then
> @@ -762,6 +781,9 @@ print_stats() {
> WRN_LIST=`grep -riwL error ${OUTPUT_PREFIX}/ERR/`
> WRN_LIST=`for f in $WRN_LIST ; do echo -n " $(basename $f)" ; done`
> WRN_CNT=`echo $WRN_LIST | wc -w | awk '{print $1}'`
> + else
> + # Remove the logs for any board that was interrupted
> + rm -f ${LOG_DIR}/${CUR_TGT}.MAKELOG ${LOG_DIR}/${CUR_TGT}.ERR
> fi
>
> echo ""
> @@ -776,7 +798,7 @@ print_stats() {
> echo "----------------------------------------------------------"
>
> if [ $BUILD_MANY == 1 ] ; then
> - kill_children $$ &
> + kill_children
> fi
>
> exit $RC
> --
> 1.7.11.5
>
^ permalink raw reply [flat|nested] 14+ messages in thread* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-10-31 1:55 [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Joe Hershberger
2012-10-31 1:55 ` [U-Boot] [PATCH 2/2] MAKEALL: Add options for incremental building Joe Hershberger
2012-10-31 20:21 ` [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Simon Glass
@ 2012-11-01 2:32 ` Marek Vasut
2012-11-01 3:02 ` Tom Rini
2012-11-01 4:14 ` Andy Fleming
2012-11-03 22:54 ` Tom Rini
2012-12-07 15:50 ` Tom Rini
4 siblings, 2 replies; 14+ messages in thread
From: Marek Vasut @ 2012-11-01 2:32 UTC (permalink / raw)
To: u-boot
Dear Joe Hershberger,
> When building in parallel, make sure that we look up the children
> based on the the actual process group id instead of just assuming
> that the MAKEALL pid is the process group id.
>
> Also ensure that logs from incomplete builds are deleted in the
> process.
>
> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
[...]
Nice $subject.
btw. is it possible to improve the u-boot build process parallelization?
Best regards,
Marek Vasut
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-11-01 2:32 ` Marek Vasut
@ 2012-11-01 3:02 ` Tom Rini
2012-11-01 4:14 ` Andy Fleming
1 sibling, 0 replies; 14+ messages in thread
From: Tom Rini @ 2012-11-01 3:02 UTC (permalink / raw)
To: u-boot
On Thu, Nov 01, 2012 at 03:32:30AM +0100, Marek Vasut wrote:
> Dear Joe Hershberger,
>
> > When building in parallel, make sure that we look up the children
> > based on the the actual process group id instead of just assuming
> > that the MAKEALL pid is the process group id.
> >
> > Also ensure that logs from incomplete builds are deleted in the
> > process.
> >
> > Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
> [...]
>
> Nice $subject.
>
> btw. is it possible to improve the u-boot build process parallelization?
Probably a little, but not a lot? It's be interesting to profile and
see where a single machine build has bottlenecks, if any.
--
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20121031/4bf34f00/attachment.pgp>
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-11-01 2:32 ` Marek Vasut
2012-11-01 3:02 ` Tom Rini
@ 2012-11-01 4:14 ` Andy Fleming
2012-11-01 14:35 ` Marek Vasut
1 sibling, 1 reply; 14+ messages in thread
From: Andy Fleming @ 2012-11-01 4:14 UTC (permalink / raw)
To: u-boot
On Wed, Oct 31, 2012 at 9:32 PM, Marek Vasut <marek.vasut@gmail.com> wrote:
> Dear Joe Hershberger,
>
>> When building in parallel, make sure that we look up the children
>> based on the the actual process group id instead of just assuming
>> that the MAKEALL pid is the process group id.
>>
>> Also ensure that logs from incomplete builds are deleted in the
>> process.
>>
>> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
> [...]
>
> Nice $subject.
>
> btw. is it possible to improve the u-boot build process parallelization?
In what way?
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-11-01 4:14 ` Andy Fleming
@ 2012-11-01 14:35 ` Marek Vasut
2012-11-01 15:23 ` Tom Rini
0 siblings, 1 reply; 14+ messages in thread
From: Marek Vasut @ 2012-11-01 14:35 UTC (permalink / raw)
To: u-boot
Dear Andy Fleming,
> On Wed, Oct 31, 2012 at 9:32 PM, Marek Vasut <marek.vasut@gmail.com> wrote:
> > Dear Joe Hershberger,
> >
> >> When building in parallel, make sure that we look up the children
> >> based on the the actual process group id instead of just assuming
> >> that the MAKEALL pid is the process group id.
> >>
> >> Also ensure that logs from incomplete builds are deleted in the
> >> process.
> >>
> >> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
> >
> > [...]
> >
> > Nice $subject.
> >
> > btw. is it possible to improve the u-boot build process parallelization?
>
> In what way?
I recall someone complained the build process wasn't parallelizable enough and
thus we need this stuff to run multiple builds at once in MAKEALL.
Best regards,
Marek Vasut
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-11-01 14:35 ` Marek Vasut
@ 2012-11-01 15:23 ` Tom Rini
0 siblings, 0 replies; 14+ messages in thread
From: Tom Rini @ 2012-11-01 15:23 UTC (permalink / raw)
To: u-boot
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 11/01/12 07:35, Marek Vasut wrote:
> Dear Andy Fleming,
>
>> On Wed, Oct 31, 2012 at 9:32 PM, Marek Vasut
>> <marek.vasut@gmail.com> wrote:
[snip]
>>> btw. is it possible to improve the u-boot build process
>>> parallelization?
>>
>> In what way?
>
> I recall someone complained the build process wasn't
> parallelizable enough and thus we need this stuff to run multiple
> builds at once in MAKEALL.
I don't know if the first half of that statement is strictly true.
It'd be interesting to profile a build on one of the very large boxes
but I think the issue is we only have ~300 objects being built. It's
hard to run up a load of 32 when you only have that many things to
build. But multiplying that by the number of boards in a arch or some
cpu or soc combinations is easier to run up the load, except around
the tails. But, profiling, please!
- --
Tom
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://www.enigmail.net/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=4XRP
-----END PGP SIGNATURE-----
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-10-31 1:55 [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Joe Hershberger
` (2 preceding siblings ...)
2012-11-01 2:32 ` Marek Vasut
@ 2012-11-03 22:54 ` Tom Rini
2012-11-05 0:13 ` Joe Hershberger
2012-12-07 15:50 ` Tom Rini
4 siblings, 1 reply; 14+ messages in thread
From: Tom Rini @ 2012-11-03 22:54 UTC (permalink / raw)
To: u-boot
On Tue, Oct 30, 2012 at 08:55:20PM -0500, Joe Hershberger wrote:
> When building in parallel, make sure that we look up the children
> based on the the actual process group id instead of just assuming
> that the MAKEALL pid is the process group id.
>
> Also ensure that logs from incomplete builds are deleted in the
> process.
>
> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Now I see:
/home/trini/bin/uboot-build.sh: line 110: 2024 Terminated
/usr/bin/time -o $TIMEFILE -f %e ./MAKEALL $SOC $MACHINE > $LOG 2>&1
With my MAKEALL wrapper. Anything we can do about that?
--
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20121103/1b7ad819/attachment.pgp>
^ permalink raw reply [flat|nested] 14+ messages in thread* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-11-03 22:54 ` Tom Rini
@ 2012-11-05 0:13 ` Joe Hershberger
2012-11-05 0:53 ` Tom Rini
0 siblings, 1 reply; 14+ messages in thread
From: Joe Hershberger @ 2012-11-05 0:13 UTC (permalink / raw)
To: u-boot
Hi Tom,
On Sat, Nov 3, 2012 at 5:54 PM, Tom Rini <trini@ti.com> wrote:
> On Tue, Oct 30, 2012 at 08:55:20PM -0500, Joe Hershberger wrote:
>
>> When building in parallel, make sure that we look up the children
>> based on the the actual process group id instead of just assuming
>> that the MAKEALL pid is the process group id.
>>
>> Also ensure that logs from incomplete builds are deleted in the
>> process.
>>
>> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
>
> Now I see:
> /home/trini/bin/uboot-build.sh: line 110: 2024 Terminated
> /usr/bin/time -o $TIMEFILE -f %e ./MAKEALL $SOC $MACHINE > $LOG 2>&1
>
> With my MAKEALL wrapper. Anything we can do about that?
I see that as well... I did a little research and it seems like you
have to jump through some pretty serious hoops to get rid of that
message. I don't think it's worth the effort.
Does it have an impact on something in your wrapper?
-Joe
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-11-05 0:13 ` Joe Hershberger
@ 2012-11-05 0:53 ` Tom Rini
0 siblings, 0 replies; 14+ messages in thread
From: Tom Rini @ 2012-11-05 0:53 UTC (permalink / raw)
To: u-boot
On Sun, Nov 04, 2012 at 06:13:03PM -0600, Joe Hershberger wrote:
> Hi Tom,
>
> On Sat, Nov 3, 2012 at 5:54 PM, Tom Rini <trini@ti.com> wrote:
> > On Tue, Oct 30, 2012 at 08:55:20PM -0500, Joe Hershberger wrote:
> >
> >> When building in parallel, make sure that we look up the children
> >> based on the the actual process group id instead of just assuming
> >> that the MAKEALL pid is the process group id.
> >>
> >> Also ensure that logs from incomplete builds are deleted in the
> >> process.
> >>
> >> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
> >
> > Now I see:
> > /home/trini/bin/uboot-build.sh: line 110: 2024 Terminated
> > /usr/bin/time -o $TIMEFILE -f %e ./MAKEALL $SOC $MACHINE > $LOG 2>&1
> >
> > With my MAKEALL wrapper. Anything we can do about that?
>
> I see that as well... I did a little research and it seems like you
> have to jump through some pretty serious hoops to get rid of that
> message. I don't think it's worth the effort.
>
> Does it have an impact on something in your wrapper?
Purely cosemetic so I can live with the bugfixes as-is, thanks!
--
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20121104/90f40ffe/attachment.pgp>
^ permalink raw reply [flat|nested] 14+ messages in thread
* [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children
2012-10-31 1:55 [U-Boot] [PATCH 1/2] MAKEALL: Fix kill_children Joe Hershberger
` (3 preceding siblings ...)
2012-11-03 22:54 ` Tom Rini
@ 2012-12-07 15:50 ` Tom Rini
4 siblings, 0 replies; 14+ messages in thread
From: Tom Rini @ 2012-12-07 15:50 UTC (permalink / raw)
To: u-boot
On Tue, Oct 30, 2012 at 08:55:20PM -0500, Joe Hershberger wrote:
> When building in parallel, make sure that we look up the children
> based on the the actual process group id instead of just assuming
> that the MAKEALL pid is the process group id.
>
> Also ensure that logs from incomplete builds are deleted in the
> process.
>
> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Applied to u-boot/master, thanks!
--
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20121207/c57aa6f9/attachment.pgp>
^ permalink raw reply [flat|nested] 14+ messages in thread