Message ID | alpine.LNX.2.00.1207031207270.17233@jbgna.fhfr.qr |
---|---|
State | New |
Headers | show |
On 07/03/2012 12:12 PM, Richard Guenther wrote: > On Tue, 3 Jul 2012, Richard Guenther wrote: > >> On Tue, 3 Jul 2012, Richard Guenther wrote: >> >>> On Tue, 3 Jul 2012, Richard Guenther wrote: >>> >>>> On Tue, 3 Jul 2012, Richard Guenther wrote: >>>> >>>>> On Mon, 2 Jul 2012, Nenad Vukicevic wrote: >>>>> >>>>>> On 6/27/2012 8:06 AM, Richard Guenther wrote: >>>>>>> This merges from the graphite branch the move of PPL to ISL, >>>>>>> and completes it where it was lacking - thanks to Micha. >>>>>>> It leaves unmerged the addition of a pluto-like ISL optimizer >>>>>>> as well as a bugfix for stride> 1 which did not come with >>>>>>> a testcase. >>>>>>> >>>>>>> With this patch (ontop of the one requiring ClooG 0.17.0) >>>>>>> we will require ISL 0.10 for enabling Graphite. >>>>>>> >>>>>>> I've bootstrapped and built various combinations with in-tree >>>>>>> and out-of-tree cloog and ISL, so I'm pretty confident that >>>>>>> this works. >>>>>>> >>>>>>> With out-of-tree ClooG and ISL a slightly older patch ontop of its >>>>>>> prerequesite passed bootstrap and testing on x86_64-unknown-linux-gnu. >>>>>>> >>>>>>> Currently re-bootstrapping and testing on x86_64-unknown-linux-gnu. >>>>>>> >>>>>>> Ok for trunk? >>>>>> >>>>>> After trying to build from the trunk I got this error on x86_64 platform: >>>>>> >>>>>> make[3]: Entering directory `/eng/upc/dev/nenad/bart/bld-trunk/cloog/test' >>>>>> CC generate_test.o >>>>>> cd ..; make libcloog-isl.la >>>>>> make[4]: Entering directory `/eng/upc/dev/nenad/bart/bld-trunk/cloog' >>>>>> CC libcloog_isl_la-domain.lo >>>>>> In file included from >>>>>> ../../gcc-trunk/cloog/include/cloog/isl/constraintset.h:4:0, >>>>>> from ../../gcc-trunk/cloog/include/cloog/isl/cloog.h:9, >>>>>> from ../../gcc-trunk/cloog/source/isl/domain.c:6: >>>>>> ../../gcc-trunk/cloog/include/cloog/isl/backend.h:4:28: fatal error: >>>> >>>> Looking again this is from the cloog build. If I repeat your tests >>>> (I suppose I didn't check non-bootstrap in-tree builds) it works just >>>> fine for me. The cloog/config.log should contain something like >>>> >>>> configure:11964: checking which isl to use >>>> configure:11966: result: system >>>> >>>> and all gcc invocations in the configury should have include paths >>>> set up to have the built isl/include first, then the source isl/include: >>>> >>>> configure:12073: gcc -c -g -O2 -I/abuild/rguenther/obj/./isl/include >>>> -I/space/rguenther/src/svn/trunk/isl/include conftest.c>&5 >>>> >>>> Did you really unpack isl into the source tree as well? >>>> >>>> Ah, I see what might happen - if you forget to have isl inside the >>>> tree we disable graphite but cloog still gets built, but will fail. >>>> I'll see to fix that. >>> >>> Fixed with the following. I'll test it with some combinations of >>> valid/invalid cloog/isl versions before installing. >> >> Hum. Ok, I installed the following, but in-tree builds seem to no >> longer work for some reason. CPPFLAGS get dropped in some way, >> even as they are present properly ... > > The following fixes that. Tested with various combinations of > in-tree/out-of-tree cloog/isl. > > Installed. Thanks Richard! Tobi
I'll try again. On the other hand I am pretty sure that I have ISL unpacked in my source tree. lrwxrwxrwx 1 nenad users 8 Jul 2 10:58 isl -> isl-0.10 drwxrwsr-x 7 nenad users 4096 Jun 4 09:40 isl-0.10 Thank you. On 7/3/2012 3:12 AM, Richard Guenther wrote: > The following fixes that. Tested with various combinations of > in-tree/out-of-tree cloog/isl. Installed. Richard.
I updated my trunk and was able to build. No problems at all. Thank you. Nenad On 7/3/2012 3:12 AM, Richard Guenther wrote: > On Tue, 3 Jul 2012, Richard Guenther wrote: > >> On Tue, 3 Jul 2012, Richard Guenther wrote: >> >>> On Tue, 3 Jul 2012, Richard Guenther wrote: >>> >>>> On Tue, 3 Jul 2012, Richard Guenther wrote: >>>> >>>>> On Mon, 2 Jul 2012, Nenad Vukicevic wrote: >>>>> >>>>>> On 6/27/2012 8:06 AM, Richard Guenther wrote: >>>>>>> This merges from the graphite branch the move of PPL to ISL, >>>>>>> and completes it where it was lacking - thanks to Micha. >>>>>>> It leaves unmerged the addition of a pluto-like ISL optimizer >>>>>>> as well as a bugfix for stride > 1 which did not come with >>>>>>> a testcase. >>>>>>> >>>>>>> With this patch (ontop of the one requiring ClooG 0.17.0) >>>>>>> we will require ISL 0.10 for enabling Graphite. >>>>>>> >>>>>>> I've bootstrapped and built various combinations with in-tree >>>>>>> and out-of-tree cloog and ISL, so I'm pretty confident that >>>>>>> this works. >>>>>>> >>>>>>> With out-of-tree ClooG and ISL a slightly older patch ontop of its >>>>>>> prerequesite passed bootstrap and testing on x86_64-unknown-linux-gnu. >>>>>>> >>>>>>> Currently re-bootstrapping and testing on x86_64-unknown-linux-gnu. >>>>>>> >>>>>>> Ok for trunk? >>>>>> After trying to build from the trunk I got this error on x86_64 platform: >>>>>> >>>>>> make[3]: Entering directory `/eng/upc/dev/nenad/bart/bld-trunk/cloog/test' >>>>>> CC generate_test.o >>>>>> cd ..; make libcloog-isl.la >>>>>> make[4]: Entering directory `/eng/upc/dev/nenad/bart/bld-trunk/cloog' >>>>>> CC libcloog_isl_la-domain.lo >>>>>> In file included from >>>>>> ../../gcc-trunk/cloog/include/cloog/isl/constraintset.h:4:0, >>>>>> from ../../gcc-trunk/cloog/include/cloog/isl/cloog.h:9, >>>>>> from ../../gcc-trunk/cloog/source/isl/domain.c:6: >>>>>> ../../gcc-trunk/cloog/include/cloog/isl/backend.h:4:28: fatal error: >>>> Looking again this is from the cloog build. If I repeat your tests >>>> (I suppose I didn't check non-bootstrap in-tree builds) it works just >>>> fine for me. The cloog/config.log should contain something like >>>> >>>> configure:11964: checking which isl to use >>>> configure:11966: result: system >>>> >>>> and all gcc invocations in the configury should have include paths >>>> set up to have the built isl/include first, then the source isl/include: >>>> >>>> configure:12073: gcc -c -g -O2 -I/abuild/rguenther/obj/./isl/include >>>> -I/space/rguenther/src/svn/trunk/isl/include conftest.c >&5 >>>> >>>> Did you really unpack isl into the source tree as well? >>>> >>>> Ah, I see what might happen - if you forget to have isl inside the >>>> tree we disable graphite but cloog still gets built, but will fail. >>>> I'll see to fix that. >>> Fixed with the following. I'll test it with some combinations of >>> valid/invalid cloog/isl versions before installing. >> Hum. Ok, I installed the following, but in-tree builds seem to no >> longer work for some reason. CPPFLAGS get dropped in some way, >> even as they are present properly ... > The following fixes that. Tested with various combinations of > in-tree/out-of-tree cloog/isl. > > Installed. > > Richard. > > 2012-06-03 Richard Guenther <rguenther@suse.de> > > * Makfile.def (isl): Remove not necessary extra_exports and > extra_make_flags. > (cloog): Use $$CPPFLAGS instead of ${CPPFLAGS}. > * Makefile.in: Regenerated. > > Index: Makefile.def > =================================================================== > --- Makefile.def (revision 189209) > +++ Makefile.def (working copy) > @@ -63,13 +63,11 @@ host_modules= { module= mpc; lib_path=.l > no_install= true; }; > host_modules= { module= isl; lib_path=.libs; bootstrap=true; > extra_configure_flags='--disable-shared --with-gmp-library=$$r/$(HOST_SUBDIR)/gmp/.libs --with-gmp-include=$$r/$(HOST_SUBDIR)/gmp --with-bits=gmp'; > - extra_exports='CPPFLAGS="${CPPFLAGS}"; export CPPFLAGS; LDFLAGS="$$LDFLAGS"; export LDFLAGS; '; > - extra_make_flags='CPPFLAGS="${CPPFLAGS}" LDFLAGS="$$LDFLAGS"'; > no_install= true; }; > host_modules= { module= cloog; lib_path=.libs; bootstrap=true; > extra_configure_flags='--disable-shared --with-gmp-library=$$r/$(HOST_SUBDIR)/gmp/.libs --with-gmp-include=$$r/$(HOST_SUBDIR)/gmp --with-bits=gmp --with-isl=system'; > - extra_exports='CPPFLAGS="-I$$r/$(HOST_SUBDIR)/isl/include -I$$s/isl/include ${CPPFLAGS}"; export CPPFLAGS; LDFLAGS="-L$$r/$(HOST_SUBDIR)/isl/.libs $$LDFLAGS"; export LDFLAGS; '; > - extra_make_flags='CPPFLAGS="${CPPFLAGS}" LDFLAGS="$$LDFLAGS" V=1'; > + extra_exports='CPPFLAGS="-I$$r/$(HOST_SUBDIR)/isl/include -I$$s/isl/include $$CPPFLAGS"; export CPPFLAGS; LDFLAGS="-L$$r/$(HOST_SUBDIR)/isl/.libs $$LDFLAGS"; export LDFLAGS; '; > + extra_make_flags='CPPFLAGS="$$CPPFLAGS" LDFLAGS="$$LDFLAGS" V=1'; > no_install= true; }; > host_modules= { module= libelf; lib_path=.libs; bootstrap=true; > extra_configure_flags='--disable-shared';
Hi Richard, What's the plan for 4.7 branch? Will you back port this patch to 4.7 and make it use ISL too? I am going to create a upstream GCC SVN branch from 4.7 for development on ARM embedded processors. If there will be some big changes for 4.7 in near future in terms of replacing PPL with ISL, I will delay the creation of my branch. Thanks. BR, Terry
On 07/04/2012 08:51 AM, Terry Guo wrote: > Hi Richard, > > What's the plan for 4.7 branch? Will you back port this patch to 4.7 and > make it use ISL too? I am going to create a upstream GCC SVN branch from 4.7 > for development on ARM embedded processors. If there will be some big > changes for 4.7 in near future in terms of replacing PPL with ISL, I will > delay the creation of my branch. Thanks. Hi Terry, I really don't think this makes a lot of sense. Those are huge changes that even bring different library requirements. Also, I don't see an urgent need to merge this back. ;-) If there is anybody who _needs_ a back port, I would be very interested in the reason. Cheers Tobi
On Tue, Jul 3, 2012 at 11:51 PM, Terry Guo <terry.guo@arm.com> wrote: > Hi Richard, > > What's the plan for 4.7 branch? Will you back port this patch to 4.7 and > make it use ISL too? I am going to create a upstream GCC SVN branch from 4.7 > for development on ARM embedded processors. If there will be some big > changes for 4.7 in near future in terms of replacing PPL with ISL, I will > delay the creation of my branch. Thanks. GCC has a policy of not backporting new features to release branches. This can be considered a new feature. In fact what might happen is disabling of the graphite support on the 4.7 branch instead. Is there a reason why you can't do development on the trunk? And then support a 4.7 for your own uses? At Cavium, we try to do development on an internal tree and then post them upstream. Though in the future we would like to do things upstream first and then backport features to a release branch that we handle internally. Thanks, Andrew Pinski
> -----Original Message----- > From: pinskia@gmail.com [mailto:pinskia@gmail.com] On Behalf Of Andrew > Pinski > Sent: Wednesday, July 04, 2012 2:58 PM > To: Terry Guo > Cc: Richard Guenther; gcc-patches@gcc.gnu.org; tobias@grosser.es; > sebpop@gmail.com; Michael Matz; Diego Novillo; Joey Ye > Subject: Re: [PATCH] Move Graphite from using PPL over to ISL > > On Tue, Jul 3, 2012 at 11:51 PM, Terry Guo <terry.guo@arm.com> wrote: > > Hi Richard, > > > > What's the plan for 4.7 branch? Will you back port this patch to 4.7 > and > > make it use ISL too? I am going to create a upstream GCC SVN branch > from 4.7 > > for development on ARM embedded processors. If there will be some big > > changes for 4.7 in near future in terms of replacing PPL with ISL, I > will > > delay the creation of my branch. Thanks. > > GCC has a policy of not backporting new features to release branches. > This can be considered a new feature. In fact what might happen is > disabling of the graphite support on the 4.7 branch instead. > > Is there a reason why you can't do development on the trunk? And then > support a 4.7 for your own uses? At Cavium, we try to do development > on an internal tree and then post them upstream. Though in the future > we would like to do things upstream first and then backport features > to a release branch that we handle internally. > Hi Tobi and Andrew, Thanks for your timely answers. I just saw Sebastian's comments: Yes, having GCC only depend on ISL and CLooG-ISL (and not depend anymore on PPL) is our plan for 4.7. from http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01161.html. As for development model, we do work as Andrew said, upstream first and then backport features. The 4.7 branch I mentioned is mainly because we want to make a tool chain release based on 4.7 with some fixes backported from trunk. BR, Terry
> -----Original Message----- > From: Terry Guo [mailto:terry.guo@arm.com] > Sent: Wednesday, July 04, 2012 15:12 > To: 'Andrew Pinski'; tobias@grosser.es > Cc: gcc-patches@gcc.gnu.org; Joey Ye > Subject: RE: [PATCH] Move Graphite from using PPL over to ISL > > > > > -----Original Message----- > > From: pinskia@gmail.com [mailto:pinskia@gmail.com] On Behalf Of > Andrew > > Pinski > > Sent: Wednesday, July 04, 2012 2:58 PM > > To: Terry Guo > > Cc: Richard Guenther; gcc-patches@gcc.gnu.org; tobias@grosser.es; > > sebpop@gmail.com; Michael Matz; Diego Novillo; Joey Ye > > Subject: Re: [PATCH] Move Graphite from using PPL over to ISL > > > > On Tue, Jul 3, 2012 at 11:51 PM, Terry Guo <terry.guo@arm.com> wrote: > > > Hi Richard, > > > > > > What's the plan for 4.7 branch? Will you back port this patch to > 4.7 > > and > > > make it use ISL too? I am going to create a upstream GCC SVN branch > > from 4.7 > > > for development on ARM embedded processors. If there will be some > big > > > changes for 4.7 in near future in terms of replacing PPL with ISL, > I > > will > > > delay the creation of my branch. Thanks. > > > > GCC has a policy of not backporting new features to release branches. > > This can be considered a new feature. In fact what might happen is > > disabling of the graphite support on the 4.7 branch instead. > > > > Is there a reason why you can't do development on the trunk? And > then > > support a 4.7 for your own uses? At Cavium, we try to do development > > on an internal tree and then post them upstream. Though in the > future > > we would like to do things upstream first and then backport features > > to a release branch that we handle internally. > > > > Hi Tobi and Andrew, > > Thanks for your timely answers. I just saw Sebastian's comments: > > Yes, having GCC only depend on ISL and CLooG-ISL (and not depend > anymore on PPL) is our plan for 4.7. > > from http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01161.html. > > As for development model, we do work as Andrew said, upstream first and > then backport features. The 4.7 branch I mentioned is mainly because we > want to make a tool chain release based on 4.7 with some fixes > backported from trunk. Don't think it make sense backporting to 4.7 either. - Joey
On Wed, Jul 4, 2012 at 12:11 AM, Terry Guo <terry.guo@arm.com> wrote: > > >> -----Original Message----- >> From: pinskia@gmail.com [mailto:pinskia@gmail.com] On Behalf Of Andrew >> Pinski >> Sent: Wednesday, July 04, 2012 2:58 PM >> To: Terry Guo >> Cc: Richard Guenther; gcc-patches@gcc.gnu.org; tobias@grosser.es; >> sebpop@gmail.com; Michael Matz; Diego Novillo; Joey Ye >> Subject: Re: [PATCH] Move Graphite from using PPL over to ISL >> >> On Tue, Jul 3, 2012 at 11:51 PM, Terry Guo <terry.guo@arm.com> wrote: >> > Hi Richard, >> > >> > What's the plan for 4.7 branch? Will you back port this patch to 4.7 >> and >> > make it use ISL too? I am going to create a upstream GCC SVN branch >> from 4.7 >> > for development on ARM embedded processors. If there will be some big >> > changes for 4.7 in near future in terms of replacing PPL with ISL, I >> will >> > delay the creation of my branch. Thanks. >> >> GCC has a policy of not backporting new features to release branches. >> This can be considered a new feature. In fact what might happen is >> disabling of the graphite support on the 4.7 branch instead. >> >> Is there a reason why you can't do development on the trunk? And then >> support a 4.7 for your own uses? At Cavium, we try to do development >> on an internal tree and then post them upstream. Though in the future >> we would like to do things upstream first and then backport features >> to a release branch that we handle internally. >> > > Hi Tobi and Andrew, > > Thanks for your timely answers. I just saw Sebastian's comments: > > Yes, having GCC only depend on ISL and CLooG-ISL (and not depend > anymore on PPL) is our plan for 4.7. > > from http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01161.html. Yes but those were made before 4.7 branched and was a release branch. The plan did not happen until almost a year later. Thanks, Andrew Pinski > > As for development model, we do work as Andrew said, upstream first and then backport features. The 4.7 branch I mentioned is mainly because we want to make a tool chain release based on 4.7 with some fixes backported from trunk. > > BR, > Terry > > >
> > Hi Tobi and Andrew, > > > > Thanks for your timely answers. I just saw Sebastian's comments: > > > > Yes, having GCC only depend on ISL and CLooG-ISL (and not depend > > anymore on PPL) is our plan for 4.7. > > > > from http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01161.html. > > Yes but those were made before 4.7 branched and was a release branch. > The plan did not happen until almost a year later. > Thanks Andrew. You solved all my questions and concerns. BR, Terry
Index: Makefile.def =================================================================== --- Makefile.def (revision 189209) +++ Makefile.def (working copy) @@ -63,13 +63,11 @@ host_modules= { module= mpc; lib_path=.l no_install= true; }; host_modules= { module= isl; lib_path=.libs; bootstrap=true; extra_configure_flags='--disable-shared --with-gmp-library=$$r/$(HOST_SUBDIR)/gmp/.libs --with-gmp-include=$$r/$(HOST_SUBDIR)/gmp --with-bits=gmp'; - extra_exports='CPPFLAGS="${CPPFLAGS}"; export CPPFLAGS; LDFLAGS="$$LDFLAGS"; export LDFLAGS; '; - extra_make_flags='CPPFLAGS="${CPPFLAGS}" LDFLAGS="$$LDFLAGS"'; no_install= true; }; host_modules= { module= cloog; lib_path=.libs; bootstrap=true; extra_configure_flags='--disable-shared --with-gmp-library=$$r/$(HOST_SUBDIR)/gmp/.libs --with-gmp-include=$$r/$(HOST_SUBDIR)/gmp --with-bits=gmp --with-isl=system'; - extra_exports='CPPFLAGS="-I$$r/$(HOST_SUBDIR)/isl/include -I$$s/isl/include ${CPPFLAGS}"; export CPPFLAGS; LDFLAGS="-L$$r/$(HOST_SUBDIR)/isl/.libs $$LDFLAGS"; export LDFLAGS; '; - extra_make_flags='CPPFLAGS="${CPPFLAGS}" LDFLAGS="$$LDFLAGS" V=1'; + extra_exports='CPPFLAGS="-I$$r/$(HOST_SUBDIR)/isl/include -I$$s/isl/include $$CPPFLAGS"; export CPPFLAGS; LDFLAGS="-L$$r/$(HOST_SUBDIR)/isl/.libs $$LDFLAGS"; export LDFLAGS; '; + extra_make_flags='CPPFLAGS="$$CPPFLAGS" LDFLAGS="$$LDFLAGS" V=1'; no_install= true; }; host_modules= { module= libelf; lib_path=.libs; bootstrap=true; extra_configure_flags='--disable-shared';