* [PATCH 20/20] Include config.h in MIN-CPPFLAGS @ 2014-08-21 18:00 Siddhesh Poyarekar 2014-08-22 5:31 ` [PATCH v1.1 " Siddhesh Poyarekar 0 siblings, 1 reply; 6+ messages in thread From: Siddhesh Poyarekar @ 2014-08-21 18:00 UTC (permalink / raw) To: libc-alpha [-- Attachment #1: Type: text/plain, Size: 1071 bytes --] This is needed when processing the Versions files since they could refer to macros defined in config.h. config.h was earlier included through libc-symbols.h but since MIN-CPPFLAGS does not include the latter anymore, it needs to at least include config.h. This was causing a difference in generated code on s390x. With this change, s390x code is also unchanged with this and other 19 patches (barring the IN_LIB patch of course). * Makeconfig (MIN-CPPFLAGS): Include config.h. --- Makeconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/Makeconfig b/Makeconfig index df26cd0..38f4851 100644 --- a/Makeconfig +++ b/Makeconfig @@ -845,6 +845,7 @@ override CXXFLAGS = $(c++-sysincludes) \ MIN-CPPFLAGS = $(config-extra-cppflags) $(CPPUNDEFS) $(CPPFLAGS-config) \ $($(subdir)-CPPFLAGS) \ $(+includes) $(defines) $(sysdep-CPPFLAGS) \ + -include $(common-objpfx)config.h \ $(CPPFLAGS-$(suffix $@)) \ $(foreach lib,$(libof-$(basename $(@F))) \ $(libof-$(<F)) $(libof-$(@F)),$(CPPFLAGS-$(lib))) \ -- 1.9.3 [-- Attachment #2: Type: application/pgp-signature, Size: 473 bytes --] ^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH v1.1 20/20] Include config.h in MIN-CPPFLAGS 2014-08-21 18:00 [PATCH 20/20] Include config.h in MIN-CPPFLAGS Siddhesh Poyarekar @ 2014-08-22 5:31 ` Siddhesh Poyarekar 2014-08-22 13:32 ` Stefan Liebler 0 siblings, 1 reply; 6+ messages in thread From: Siddhesh Poyarekar @ 2014-08-22 5:31 UTC (permalink / raw) To: libc-alpha [-- Attachment #1: Type: text/plain, Size: 1399 bytes --] On Thu, Aug 21, 2014 at 11:29:54PM +0530, Siddhesh Poyarekar wrote: > This is needed when processing the Versions files since they could > refer to macros defined in config.h. config.h was earlier included > through libc-symbols.h but since MIN-CPPFLAGS does not include the > latter anymore, it needs to at least include config.h. > > This was causing a difference in generated code on s390x. With this > change, s390x code is also unchanged with this and other 19 patches > (barring the IN_LIB patch of course). > > * Makeconfig (MIN-CPPFLAGS): Include config.h. config.h checks a few macros, most of which are set at command line, with the exception of _LIBC, which is set in libc-symbols.h. Set it on the commandline to be consistent with the behaviour we ought to get with libc-symbols.h included. * Makeconfig (MIN-CPPFLAGS): Include config.h. diff --git a/Makeconfig b/Makeconfig index df26cd0..c6eae06 100644 --- a/Makeconfig +++ b/Makeconfig @@ -845,6 +845,7 @@ override CXXFLAGS = $(c++-sysincludes) \ MIN-CPPFLAGS = $(config-extra-cppflags) $(CPPUNDEFS) $(CPPFLAGS-config) \ $($(subdir)-CPPFLAGS) \ $(+includes) $(defines) $(sysdep-CPPFLAGS) \ + -D_LIBC -include $(common-objpfx)config.h \ $(CPPFLAGS-$(suffix $@)) \ $(foreach lib,$(libof-$(basename $(@F))) \ $(libof-$(<F)) $(libof-$(@F)),$(CPPFLAGS-$(lib))) \ [-- Attachment #2: Type: application/pgp-signature, Size: 473 bytes --] ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v1.1 20/20] Include config.h in MIN-CPPFLAGS 2014-08-22 5:31 ` [PATCH v1.1 " Siddhesh Poyarekar @ 2014-08-22 13:32 ` Stefan Liebler 2014-08-22 14:22 ` Siddhesh Poyarekar 2014-08-22 18:31 ` Siddhesh Poyarekar 0 siblings, 2 replies; 6+ messages in thread From: Stefan Liebler @ 2014-08-22 13:32 UTC (permalink / raw) To: libc-alpha Hi Siddhesh, I´ve tested your is_in_module branch on s390x. Now there are only Wundef warnings with _POSIX*/_XBS*, which are mentioned in release wiki. Thanks. But building the testsuite fails while building nptl_db/db-symbols.v.iT because NOT_IN is not defined in structs.def: structs.def:81:12: error: missing binary operator before token "(" #if NOT_IN (libpthread) || TLS_TCB_AT_TP ^ The defining NOT_IN to zero of patch 19 is not done, because of patch v1.1 20. patch 19: +#ifndef _LIBC +# define NOT_IN(lib) (0) +#endif patch v1.1 20: + -D_LIBC -include $(common-objpfx)config.h \ After changing patch 19 to #ifndef NOT_IN # define NOT_IN(lib) (0) #endif the following tests are failing: FAIL: conform/ISO/stdlib.h/conform FAIL: conform/ISO11/stdlib.h/conform FAIL: conform/ISO99/stdlib.h/conform FAIL: conform/POSIX/mqueue.h/conform FAIL: conform/POSIX/stdlib.h/conform FAIL: conform/POSIX2008/mqueue.h/conform FAIL: conform/POSIX2008/stdlib.h/conform FAIL: conform/UNIX98/mqueue.h/conform FAIL: conform/XOPEN2K/stdlib.h/conform FAIL: conform/XOPEN2K8/mqueue.h/conform FAIL: conform/XOPEN2K8/stdlib.h/conform with either: >Namespace violation: "IS_IN" or: >../include/bits/stdlib-float.h:2:12: error: missing binary operator before token "(" #if NOT_IN (rtld) ^ FAIL: stdlib/isomac: >... stdlib.h system() returned nonzero ... I´ve also compared the obj-files and got the mentioned diffs of patch 5. Bye On 08/22/2014 07:31 AM, Siddhesh Poyarekar wrote: > On Thu, Aug 21, 2014 at 11:29:54PM +0530, Siddhesh Poyarekar wrote: >> This is needed when processing the Versions files since they could >> refer to macros defined in config.h. config.h was earlier included >> through libc-symbols.h but since MIN-CPPFLAGS does not include the >> latter anymore, it needs to at least include config.h. >> >> This was causing a difference in generated code on s390x. With this >> change, s390x code is also unchanged with this and other 19 patches >> (barring the IN_LIB patch of course). >> >> * Makeconfig (MIN-CPPFLAGS): Include config.h. > > config.h checks a few macros, most of which are set at command line, > with the exception of _LIBC, which is set in libc-symbols.h. Set it > on the commandline to be consistent with the behaviour we ought to get > with libc-symbols.h included. > > > * Makeconfig (MIN-CPPFLAGS): Include config.h. > > diff --git a/Makeconfig b/Makeconfig > index df26cd0..c6eae06 100644 > --- a/Makeconfig > +++ b/Makeconfig > @@ -845,6 +845,7 @@ override CXXFLAGS = $(c++-sysincludes) \ > MIN-CPPFLAGS = $(config-extra-cppflags) $(CPPUNDEFS) $(CPPFLAGS-config) \ > $($(subdir)-CPPFLAGS) \ > $(+includes) $(defines) $(sysdep-CPPFLAGS) \ > + -D_LIBC -include $(common-objpfx)config.h \ > $(CPPFLAGS-$(suffix $@)) \ > $(foreach lib,$(libof-$(basename $(@F))) \ > $(libof-$(<F)) $(libof-$(@F)),$(CPPFLAGS-$(lib))) \ > ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v1.1 20/20] Include config.h in MIN-CPPFLAGS 2014-08-22 13:32 ` Stefan Liebler @ 2014-08-22 14:22 ` Siddhesh Poyarekar 2014-08-22 18:31 ` Siddhesh Poyarekar 1 sibling, 0 replies; 6+ messages in thread From: Siddhesh Poyarekar @ 2014-08-22 14:22 UTC (permalink / raw) To: Stefan Liebler; +Cc: GNU C Library On 22 August 2014 19:01, Stefan Liebler <stli@linux.vnet.ibm.com> wrote: > Hi Siddhesh, > > I扉e tested your is_in_module branch on s390x. Thanks! > But building the testsuite fails while building nptl_db/db-symbols.v.iT > because NOT_IN is not defined in structs.def: > structs.def:81:12: error: missing binary operator before token "(" > #if NOT_IN (libpthread) || TLS_TCB_AT_TP > ^ > > The defining NOT_IN to zero of patch 19 is not done, because of patch v1.1 > 20. > > patch 19: > +#ifndef _LIBC > +# define NOT_IN(lib) (0) > +#endif > > patch v1.1 20: > + -D_LIBC -include $(common-objpfx)config.h \ > > > After changing patch 19 to > #ifndef NOT_IN > # define NOT_IN(lib) (0) > #endif I think I need to generate libc-modules.h without depending on soversions.i. > the following tests are failing: > FAIL: conform/ISO/stdlib.h/conform > FAIL: conform/ISO11/stdlib.h/conform > FAIL: conform/ISO99/stdlib.h/conform > FAIL: conform/POSIX/mqueue.h/conform > FAIL: conform/POSIX/stdlib.h/conform > FAIL: conform/POSIX2008/mqueue.h/conform > FAIL: conform/POSIX2008/stdlib.h/conform > FAIL: conform/UNIX98/mqueue.h/conform > FAIL: conform/XOPEN2K/stdlib.h/conform > FAIL: conform/XOPEN2K8/mqueue.h/conform > FAIL: conform/XOPEN2K8/stdlib.h/conform > with either: >>Namespace violation: "IS_IN" > or: OK, that's porobably a good reason to not try to define IS_IN or NOT_IN and just use _LIBC in the conditionals. I'll fix these things up and update the branch. Thanks, Siddhesh -- http://siddhesh.in ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v1.1 20/20] Include config.h in MIN-CPPFLAGS 2014-08-22 13:32 ` Stefan Liebler 2014-08-22 14:22 ` Siddhesh Poyarekar @ 2014-08-22 18:31 ` Siddhesh Poyarekar 2014-08-25 7:28 ` Stefan Liebler 1 sibling, 1 reply; 6+ messages in thread From: Siddhesh Poyarekar @ 2014-08-22 18:31 UTC (permalink / raw) To: Stefan Liebler; +Cc: libc-alpha [-- Attachment #1: Type: text/plain, Size: 4213 bytes --] On Fri, Aug 22, 2014 at 03:31:41PM +0200, Stefan Liebler wrote: > But building the testsuite fails while building nptl_db/db-symbols.v.iT > because NOT_IN is not defined in structs.def: > structs.def:81:12: error: missing binary operator before token "(" > #if NOT_IN (libpthread) || TLS_TCB_AT_TP Stefan, Can you please try this patch? It should fix everything except the conformtest failures. The fix for conformtest failures is fairly straightforward and I'll add it as a modification of another patch. Thanks, Siddhesh commit bc387f3bcc5059a32f8cd60336b06c5027f0cc88 Author: Siddhesh Poyarekar <siddhesh@redhat.com> Date: Fri Aug 22 23:21:13 2014 +0530 Use regular CPPFLAGS for everything except shlib-versions There was a circular dependency between shlib-versions.v.i and libc-modules, which I had attempted to break earlier by excluding libc-symbols.h from CPPFLAGS for %.v.i targets. This caused a few other problems like _LIBC not being defined in nptl_db/structs.def and ABI breakage in s390x (and probably also in m68k and alpha). This patch breaks libc-modules.h out of libc-symbols.h and includes it in the commandline instead. Additionally, the header is not included in the commandline when generating shlib-versions.v.i, but is present everywhere else. diff --git a/Makeconfig b/Makeconfig index c6eae06..ccd5d84 100644 --- a/Makeconfig +++ b/Makeconfig @@ -818,15 +818,20 @@ in-module = $(strip $(foreach lib,$(libof-$(basename $(@F))) $(libof-$(<F)) \ module-def = $(if $(in-module),$(in-module),-DIN_MODULE=MODULE_libc) +# We don't need libc-modules.h and the IN_MODULE definition for +# shlib-version.v.i. +module-cppflags = $(if $(filter shlib-versions.v.i,$(@F)),,-include \ + $(common-objpfx)libc-modules.h $(module-def)) + # These are the variables that the implicit compilation rules use. # Note that we can't use -std=* in CPPFLAGS, because it overrides # the implicit -lang-asm and breaks cpp behavior for .S files--notably # it causes cpp to stop predefining __ASSEMBLER__. CPPFLAGS = $(config-extra-cppflags) $(CPPUNDEFS) $(CPPFLAGS-config) \ $($(subdir)-CPPFLAGS) \ - $(+includes) $(defines) \ + $(+includes) $(defines) $(module-cppflags) \ -include $(..)include/libc-symbols.h $(sysdep-CPPFLAGS) \ - $(CPPFLAGS-$(suffix $@)) $(module-def) \ + $(CPPFLAGS-$(suffix $@)) \ $(foreach lib,$(libof-$(basename $(@F))) \ $(libof-$(<F)) $(libof-$(@F)),$(CPPFLAGS-$(lib))) \ $(CPPFLAGS-$(<F)) $(CPPFLAGS-$(@F)) $(CPPFLAGS-$(basename $(@F))) @@ -840,16 +845,6 @@ override CXXFLAGS = $(c++-sysincludes) \ $(filter-out %frame-pointer,$(+cflags)) $(sysdep-CFLAGS) \ $(CFLAGS-$(suffix $@)) $(CFLAGS-$(<F)) $(CFLAGS-$(@F)) -# Minimal CPPFLAGS for the initial compilations, i.e. syscall stubs, %.v.i, -# etc. -MIN-CPPFLAGS = $(config-extra-cppflags) $(CPPUNDEFS) $(CPPFLAGS-config) \ - $($(subdir)-CPPFLAGS) \ - $(+includes) $(defines) $(sysdep-CPPFLAGS) \ - -D_LIBC -include $(common-objpfx)config.h \ - $(CPPFLAGS-$(suffix $@)) \ - $(foreach lib,$(libof-$(basename $(@F))) \ - $(libof-$(<F)) $(libof-$(@F)),$(CPPFLAGS-$(lib))) \ - $(CPPFLAGS-$(<F)) $(CPPFLAGS-$(@F)) $(CPPFLAGS-$(basename $(@F))) # If everything is compiled with -fPIC (implicitly) we must tell this by # defining the PIC symbol. ifeq (yes,$(build-pic-default)) @@ -938,7 +933,7 @@ subdir-srcdirs = $(foreach dir,$(subdirs),\ %.v.i: $(common-objpfx)config.h $(..)Makeconfig sed '/^[ ]*%/!s/#.*$$//;/^[ ]*$$/d;s/^[ ]*%/#/' \ $(filter-out FORCE %.h $(..)Makeconfig,$^) \ - | $(CC) -E -undef $(MIN-CPPFLAGS) -x assembler-with-cpp - \ + | $(CC) -E -undef $(CPPFLAGS) -x assembler-with-cpp - \ > $@T mv -f $@T $@ %.v: %.v.i diff --git a/include/libc-symbols.h b/include/libc-symbols.h index 0081301..4155083 100644 --- a/include/libc-symbols.h +++ b/include/libc-symbols.h @@ -20,8 +20,6 @@ #ifndef _LIBC_SYMBOLS_H #define _LIBC_SYMBOLS_H 1 -#include "libc-modules.h" - #define IS_IN(lib) (IN_MODULE == MODULE_##lib) #define NOT_IN(lib) !IS_IN (lib) [-- Attachment #2: Type: application/pgp-signature, Size: 473 bytes --] ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v1.1 20/20] Include config.h in MIN-CPPFLAGS 2014-08-22 18:31 ` Siddhesh Poyarekar @ 2014-08-25 7:28 ` Stefan Liebler 0 siblings, 0 replies; 6+ messages in thread From: Stefan Liebler @ 2014-08-25 7:28 UTC (permalink / raw) To: libc-alpha On 08/22/2014 08:30 PM, Siddhesh Poyarekar wrote: > On Fri, Aug 22, 2014 at 03:31:41PM +0200, Stefan Liebler wrote: >> But building the testsuite fails while building nptl_db/db-symbols.v.iT >> because NOT_IN is not defined in structs.def: >> structs.def:81:12: error: missing binary operator before token "(" >> #if NOT_IN (libpthread) || TLS_TCB_AT_TP > > Stefan, > > Can you please try this patch? It should fix everything except the > conformtest failures. The fix for conformtest failures is fairly > straightforward and I'll add it as a modification of another patch. > > Thanks, > Siddhesh Hi Siddhesh, tested it on s390x. Now building the testsuite is okay. Thanks. The conformtests are failing as you said. Bye ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2014-08-25 7:28 UTC | newest] Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2014-08-21 18:00 [PATCH 20/20] Include config.h in MIN-CPPFLAGS Siddhesh Poyarekar 2014-08-22 5:31 ` [PATCH v1.1 " Siddhesh Poyarekar 2014-08-22 13:32 ` Stefan Liebler 2014-08-22 14:22 ` Siddhesh Poyarekar 2014-08-22 18:31 ` Siddhesh Poyarekar 2014-08-25 7:28 ` Stefan Liebler
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).