This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] [MELT] Fix loading of .c files
Le 29/08/2011 16:57, Alexandre Lissy a écrit :
> Le 29/08/2011 16:44, Alexandre Lissy a écrit :
>> Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
>>> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
>>>> Files containing the MELT translated to C code were not loaded correctly
>>>> due to bad search path when working with an installed plugin. Present
>>>> commit fixes this by:
>>>> - using a melt find to search in current dir
>>>> - ensuring the melt_modulename in melt descriptor only contains the
>>>> module name, and no path component
>>>
>>> Are you sure of that patch? It seems that it loops on the hello world tiny test I added in the MELT branch rev 178154
>>>
>>> Does it work for you when merging with latest MELT updates (ie when applying that patch to rev 178154 of the MELT branch?
>>>
>>> (I am investigating)
>>>
>>> Cheers.
>>>
>>
>> Okay, once fixing your Makefile it's okay:
>>
>> $ LC_ALL=C time -p gcc -c @melt-sayhello.args
>> cc1: note: MELT is bootstrapping so ignore builtin source directory
>> /usr/lib64/gcc/x86_64-mandriva-linux-gnu/4.6.1/plugin/melt-sources
>> cc1: note: MELT is bootstrapping so ignore builtin module directory
>> /usr/lib64/gcc/x86_64-mandriva-linux-gnu/4.6.1/plugin/melt-modules
>> cc1: note: MELT generating C code of module melt-sayhello
>> cc1: note: MELT generated same file melt-sayhello.c in
>> /home/alex/BuildSystem/gcc-plugin-melt/BUILD/melt-0.8.99-plugin-for-gcc-4.6
>> cc1: note: MELT generated descriptor file melt-sayhello+meltdesc.c
>> cc1: note: MELT generated C code of module melt-sayhello with 0
>> secondary files in 10 CPU millisec.
>> real 0.06
>> user 0.02
>> sys 0.02
>>
>> The bug was: -fplugin-arg-melt-arg=melt-modules. I suspect it's this
>> line in melt-build.tpl:
>> $(meltarg_arg)=$< -frandom-seed=$(shell md5sum melt-sayhello.melt | cut
>> -b-24) \
>>
>> The '$<' part ... But I have no idea of your intentions.
>>
>
> I suspect '$<' should be replaced by 'melt-sayhello.melt' to get
> -fplugin-arg-melt-arg=melt-sayhello.melt
>
> At least, it goes well doing this ...
>
I am currently trying the attached patch.
diff --git a/gcc/melt-build.tpl b/gcc/melt-build.tpl
index 10462a4..55fec0d 100644
--- a/gcc/melt-build.tpl
+++ b/gcc/melt-build.tpl
@@ -562,18 +562,22 @@ melt-modules/[+base+].quicklybuilt.so: melt-sources/[+base+].c \
################
#@ [+ (. (tpl-file-line))+]
-melt-tiny-tests: melt-modules melt-sources melt-all-modules melt-all-sources
+melt-sayhello.melt:
+ @date +'(code_chunk say%YM%mhello #{printf("hello_from_MELT on %c pid %%d\n", (int) getpid());}#)' > $@
+#@ [+ (. (tpl-file-line))+]
+
+#@ [+ (. (tpl-file-line))+]
+melt-tiny-tests: melt-sayhello.melt melt-modules melt-sources melt-all-modules melt-all-sources
# test that a helloworld can be translated
- @date +'(code_chunk say%YM%mhello #{printf("hello_from_MELT on %c pid %%d\n", (int) getpid());}#)' > melt-sayhello.melt
@echo $(MELTCCAPPLICATION1ARGS) \
- $(meltarg_arg)=$< -frandom-seed=$(shell md5sum melt-sayhello.melt | cut -b-24) \
+ $(meltarg_arg)=$< -frandom-seed=$(shell md5sum $< | cut -b-24) \
$(meltarg_module_path)=$(realpath melt-modules) \
$(meltarg_source_path)=$(realpath melt-sources) \
- $(meltarg_workdir)=melt-workdir $(meltarg_inhibitautobuild) \
- $(meltarg_output)=melt-sayhello.c empty-file-for-melt.c > melt-sayhello.args-tmp
- @mv melt-sayhello.args-tmp melt-sayhello.args
- @echo -n melt-sayhello.args: ; cat melt-sayhello.args ; echo "***** doing " melt-sayhello
- $(melt_make_cc1) @melt-sayhello.args
+ $(meltarg_workdir)=melt-workdir $(meltarg_inhibitautobuild) $(meltarg_bootstrapping) \
+ $(meltarg_output)=$(basename $<).c empty-file-for-melt.c > $(basename $<).args-tmp
+ @mv $(basename $<).args-tmp $(basename $<).args
+ @echo -n $(basename $<).args: ; cat $(basename $<).args ; echo "***** doing " $(basename $<)
+ $(melt_make_cc1) @$(basename $<).args
#@ [+ (. (tpl-file-line))+]