Just for feedback:
Today's CVS snapshot "binutils-2.17.50", after applying all the workarounds
reported today applied, builds fine with the configure options below
on sparc-sun-solaris2.6, installs fine, and the resulting gas and gld
seem to start off well in bootstrapping gcc...
Thanks, and gree
Hi
Build of snapshot binutils-2.17.50 fails (using gcc-4.1.2)
when compiling binutils/srconv.c
Workaround: initialize
dpt.dunno = 0;
Arto
--
gcc -DHAVE_CONFIG_H -I. -I/build/binutils-2.17.50/binutils -I. -D_GNU
Hi
Already reported for distribution binutils-2.17,
but build also fails with todays CVS snapshot binutils-2.17.50:
When using "configure --enable-targets=all"
then compiling opcodes/m32c-asm.c fails in cgen-types.h
because no file stdint.h can be found anywhere on Solaris.
(neither 2.6 nor Solar
Hi
Building of snapshot binutils-2.17.50 fails
(when using "configure --enable-targets=all"
on Solaris 2.6) in file opcodes/arm-dis.h
Workaround: initialize
enum map_type type = 0;
(to be verified that this is no problem...)
Arto
---
gcc -DHAVE_CON
Hi
For 2 more files, build of snapshot binutils-2.17.50 fails (on Solaris 2.6)
when compiling bfd/vms-hdr.c and bfd/xsym.c because alloca is undefined.
Strangely enough, the first file (bfd/vms-hdr.c) even has
#ifdef HAVE_ALLOCA_H
#include
#endif
Therfore, it is probably the configure (when usi
Hi
Build of snapshot binutils-2.17.50 fails (using gcc-4.1.2)
when compiling bfd/pe-sh.c inside coffcode.h line 4911
because of the following difference regarding CALC_ADDEND()
in file bfd/coff-sh.c which is included by bfd/pe-sh.c
/* This is the same as the macro in coffcode.h, except that it
Hi
3 -Werror failures in bfd/ieee.c (full output of first one below,
the other ones essentially the same...)
bfd/ieee.c:381: warning: 'result' may be used uninitialized
bfd/ieee.c:373: warning: 'x' may be used uninitialized
bfd/ieee.c:771: warning: 'value' may be used uninitialized
Workaround:
Hi
in snapshot binutils-2.17.50 the build fails in bfd/elf32-m68hc1x.c
because alloca is not defined (when using "configure --enable-targets=all"
on sun-sparc-solaris2.6)
Workaround:
#include
on my platform (Solaris 2.6)
Greets,
Arto
-
BTW, same error also on line 3089...
(after setting this also to 0, file compiles alright).
Arto
- Begin Included Message -
>From [EMAIL PROTECTED] Fri Apr 20 22:11:48 2007
To: bug-binutils@gnu.org
Subject: build of binutils-2.17.50 snapshot fails: bfd/ecoff.c
Hi
With the compiler gc
Hi
With the compiler gcc-4.1.2 detects that in file bfd/ecoff.c
the uninitialized int rehash on line 3760 is not always set
by function ecoff_armap_hash() and the build fails (see below)
due to -Werror.
best regards,
Arto
--
...
...
/b
What is the point of building in a different directory, if the
original directory still needs to be modified?
Arto
> From anirkko Thu Apr 19 10:25:16 2007
> To: bug-binutils@gnu.org
> Subject: binutils-2.17 build tries to modify distribution directory
>
>
> Hi
> When building b
Hi
When building binutils-2.17 in a different obj directory (as recommended),
with the original distribution directory write protected/owned by a different
user, one notices that the build process tries to modify the distribution
directory (with messages such as "Permission denied" and "touch: can
and again:
Using gcc-4.1.2, build of binutils-2.17 also fails in gas/read.c
due to -Werror warning:
Here, it seems to me that the code is ok. Maybe the compiler
assumes that flag_mri can change between the paired uses of stopc,
possibly because stopc is global and might be modified by another
thr
Hi again.
Interesting, gcc-4.1.2 seems to discover initialization problems
even across function calls, triggering another -Werror failure in
build of binutils-2.17:
In file binutils/wrstabs.c the initialization is lacking in stab_tag_type(),
and the compiler tells us this, but points out line 12
Hi again
Building binutils-2.17 also fails with --enable-64-bit-bfd (see output below)
on an Ultra-1 (ultrasparc, running Solaris 2.6 on a 32bit-kernel).
Best regards,
Arto
-
...
...
gcc -W -Wall -Wstrict-prototypes -Wmis
Hi
Building binutils-2.17 with --enable-targets=all fails, apparently
because of a missing stdint.h file (see output below), which I can't find
anywhere on my system (Solaris 2.6). Should the configure script not check
for the presence of this, and possibly substitute if missing? Or are you
suppos
The build of binutils-2.17 fails with gcc-4.1.2, as far as I can tell
because configure turns on -Werror by default (see output below).
Maybe gcc-4.1.2 warns about stuff that previous versions (gcc-3.x.x) did not?
Have no idea whether the warning points out a real problem or not.
Maybe not, there
17 matches
Mail list logo
