similar to: [LLVMdev] llvm-gcc fortran bootstrap broken

On Wednesday 30 July 2008 18:13:27 Duncan Sands wrote: > On x86-64 linux, in stage 2, I get: > > c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -Wold-style-definition -Wmissing-format-attribute -fno-common -DHAVE_CONFIG_H -o f951 \ > fortran/arith.o

On Jul 30, 2008, at 11:39 AM, Duncan Sands wrote: > On Wednesday 30 July 2008 18:13:27 Duncan Sands wrote: >> On x86-64 linux, in stage 2, I get: >> >> c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings -Wstrict- >> prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno- >> variadic-macros -Wno-overlength-strings -Wold-style-definition - >>

And how about this one for handling the C specific invoke_impl_ptr_type? Index: gcc-4.2.llvm/gcc/c-common.h =================================================================== --- gcc-4.2.llvm.orig/gcc/c-common.h 2008-07-30 21:01:49.000000000 +0200 +++ gcc-4.2.llvm/gcc/c-common.h 2008-07-30 21:01:56.000000000 +0200 @@ -1164,7 +1164,6 @@ extern struct block_sema_info *cur_block; extern tree

And how about this one so as not to include a C specific header in llvm-backend (!!!) and not to have llvm-backend use a C specific flag (flag_no_builtin)? Index: gcc-4.2.llvm/gcc/c-opts.c =================================================================== --- gcc-4.2.llvm.orig/gcc/c-opts.c 2008-07-30 21:25:28.000000000 +0200 +++ gcc-4.2.llvm/gcc/c-opts.c 2008-07-30 21:26:17.000000000 +0200 @@

This is a forward of a question from the SourceForge LLVM forum. If anyone has experience with Cygwin and building the C front-end there, please reply and CC Alex. --- Post by Alex Vinokur below --- =============== Windows 2000 GNU gcc 3.3.3 (cygwin special) =============== ./configure ./make I have got compilation error ------------------------------- make[1]: Entering directory

How to solve this problem? $ make make[1]: Entering directory `/tmp/llvm3/llvm/build/libiberty' make[2]: Entering directory `/tmp/llvm3/llvm/build/libiberty/testsuite' make[2]: Nothing to be done for `all'. make[2]: Leaving directory `/tmp/llvm3/llvm/build/libiberty/testsuite' make[1]: Leaving directory `/tmp/llvm3/llvm/build/libiberty' make[1]: Entering directory

I'm getting a build error when trying to build gcc4 from sources. This is for the recent 1.9 release. How I built llvm-1.9: ----------------------------- tar zxf llvm-1.9.tar.gz cd llvm-1.9/ ./configure --prefix=/custom/llvm-1.9 make ENABLE_OPTIMIZED=1 OPTIMIZE_OPTION='-O2' tools-only make install How I built gcc4: ----------------------------- export

Hi, in trying to build an llvm-gfortran on linux from today's gcc4 SVN, I get the following link errors that didn't show up on OS X. This is also using LLVM CVS from this morning. They appear to mostly be related to CodeWarrior compatibility code, but a couple are less obvious. What's the best way to fix this? I'm stuck without my powerbook for at least three weeks and I'd

Hi everyone, I am using linux/ppc and after porting llvm's jit for this architecture, I am currently trying to compile llvm-gcc on it. I have gcc 4.1.1 and use llvm-gcc from cvs. Compilation fails and the message is: gcc -O2 -DIN_GCC -DDEFAULT_TARGET_MACHINE=\"powerpc-unknown-linux-gnu\" -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long

There was a patch that went thru for this recently from Rafael Espindola. The fix is on the mirror. Index: gcc/dwarf2out.c =================================================================== --- gcc/dwarf2out.c (revision 120589) +++ gcc/dwarf2out.c (working copy) @@ -14361,9 +14361,8 @@ s->refcount++; /* Avoid unnecessarily putting strings that are used less than twice in the hash

Hi, After a svn/cvs update I get the following errors when building the frontend. Anyone that sees the same ? http://gcc.gnu.org/svn/gcc/branches/apple/trunk/gcc/expr.h /* APPLE LOCAL begin radar 4441049 */ extern tree objc_v2_component_ref_field_offset (tree exp); extern tree objc_v2_bitfield_ivar_bitpos (tree exp); /* APPLE LOCAL end radar 4441049 */ mentions APPLE, is it only working on

The curent llvm svn (r54623) is unable to link the gfortran compiler in llvm-gcc-4.2 svn. I am getting the error... c++ -g -O2 -mdynamic-no-pic -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -Wold-style-definition -Wmissing-format-attribute -mdynamic-no-pic -DHAVE_CONFIG_H -o f951 \

Hi Jack, > The curent llvm svn (r54623) is unable to link the gfortran > compiler in llvm-gcc-4.2 svn. I am getting the error... ... > Undefined symbols: > "_create_init_utf16_var", referenced from: > _darwin_build_constant_cfstring in libbackend.a(darwin.o) this is probably due to recent Apple changes. Fortran builds on linux. > Why is c++ being used

Attached is a port of the new ARM eabi from gcc 4.1 to the llvm-gcc branch. With this patch I am able to bootstrap the 4.0 branch using the new eabi. The llvm-gcc branch fails with ------------------------------------ internal compiler error: in prune_unused_types_update_strings, at dwarf2out.c:14372 ------------------------------------ But I believe that this is an unrelated problem and I will

Thanks for the suggestion, Andre. I downloaded GNAT GPL 2007, and followed the directions on your web page. Now I get a lot of C++ errors. In fact, the output I captured to a log file is over six megabytes just from the start of the problem command to the termination of the make. I'm obviously not going to include it all here, but the command and the first few errors are copied below.

> LLVM-GCC 4.0 fails on Cygwin, possibly due to lack of symbolic links. I am trying with llvm-gcc-4.2 from svn. Aaron

A clean build of revisions 12 or 13 of llvm-gcc4 produces: make[1]: *** No rule to make target `llvm-backend.c', needed by `llvm-backend.o'. Stop. While the gcc/Makefile.in has: llvm-backend.o : llvm-backend.cpp $(ALL_HEADERS) #line 2440 it also has: # APPLE LOCAL begin LLVM #line 1018 ifneq ($(LLVMOBJDIR),) OBJS-common += llvm-backend.o llvm-convert.o

Hello, I have an analysis pass and a target machine class, both in different dynamically loadable libraries (plugins). There is no problem to use the analysis pass (opt -load=libanalysis.dylib -barpass foo.bc), but is it possible to get data from the analysis pass (by getAnalysis) while using the backend? AFAIK it is only allowed to have one plugin argument for llc (like: llc

Hi The fix is very simple: the defaulting of this macro to 0 in gcc/config/arm/arm.h, present in 2.2 has been omitted in 2.5. The attached patch puts the defaulting clause back, the same as it was in 2.2 M -------------- next part -------------- A non-text attachment was scrubbed... Name: 0004-arm-default-MACHO_DYNAMIC_NO_PIC_P.patch Type: text/x-diff Size: 385 bytes Desc: not available

Hello >   The fix is very simple: the defaulting of this macro to 0 in > gcc/config/arm/arm.h, present in 2.2 has been omitted in 2.5. > > The attached patch puts the defaulting clause back, the same as it was in 2.2 This is known problem which was fixed after 2.5 release. -- With best regards, Anton Korobeynikov Faculty of Mathematics and Mechanics, Saint Petersburg State