public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/23687] New: Crosscompiler looks for files it never built
@ 2005-09-01 22:26 imipak at yahoo dot com
2005-09-01 22:28 ` [Bug target/23687] " pinskia at gcc dot gnu dot org
2005-09-13 7:50 ` rsandifo at gcc dot gnu dot org
0 siblings, 2 replies; 4+ messages in thread
From: imipak at yahoo dot com @ 2005-09-01 22:26 UTC (permalink / raw)
To: gcc-bugs
I can compile the initial version of GCC 4.1.0 on a Pentium 4 with a target of
an SB1 MIPS64 bigendian processor, as required to bootstrap the toolchain
compilation. On attempting to use this cross-compiler, it reports that it
cannot find crti.o. (This was tested in a number of ways, including compiling
glibc as that is the next stage in building the cross compiler toolchain.)
The source for crti is in the config directory for the MIPS processor but is
not being accessed.I can compile this manually and copying the resultant file
into the proper place.
Using the cross-compiler now, I then get the error that the cross-compiler
cannot find crt1.o. There is no sourcecode for crt1 in the MIPS directory. The
only other crt file in that directory is for crtn - this file will not compile
at all.
The main compiler I am using for building the cross-compiler is GCC 3.4.3 and
was built using the following options:
Configured with: ../configure --prefix=/usr --mandir=/usr/share/man
--infodir=/usr/share/info --enable-shared --enable-threads=posix
--disable-checking --with-system-zlib --enable-__cxa_atexit
--disable-libunwind-exceptions --enable-java-awt=gtk --host=i386-redhat-linux
Thread model: posix
gcc version 3.4.3 20050227 (Red Hat 3.4.3-22.1)
Copying the .o files from a third-party toolchain, the cross-compiler will
progress a good deal further. However, it will fail when cross-compiling either
glibc or newlib. In the case of newlib, it crashes
at ../gcc/libstdc++-v3/libmath/stubs.c
../../../../gcc/libstdc++-v3/libmath/stubs.c: In function 'floorl':
/usr/mips/mipsisa64sb1-unknown-linux-gnu/include/bits/mathinline.h:530: error:
impossible register constraint in 'asm'
/usr/mips/mipsisa64sb1-unknown-linux-gnu/include/bits/mathinline.h:530: error:
impossible register constraint in 'asm'
../../../../gcc/libstdc++-v3/libmath/stubs.c:203: error: unrecognizable insn:
(insn 20 19 21
0 /usr/mips/mipsisa64sb1-unknown-linux-gnu/include/bits/mathinline.h:530 (set
(reg/v:DF 32 $f0 [orig:195 __value ] [195])
(asm_operands/v:DF ("frndint") ("=t") 0 [
(reg/v:DF 32 $f0 [orig:195 __value ] [195])
]
[
(asm_input:DF ("0"))
]
("/usr/mips/mipsisa64sb1-unknown-linux-gnu/include/bits/mathinline.h") 530)) -1
(insn_list:REG_DEP_TRUE 19 (nil))
(nil))
../../../../gcc/libstdc++-v3/libmath/stubs.c:203: internal compiler error: in
reload_cse_simplify_operands, at postreload.c:393
The cross-compiler was built using the following configuration settings:
Target: mipsisa64sb1-unknown-linux-gnu
Configured with: ../gcc/configure --target=mipsisa64sb1-unknown-linux-gnu
--prefix=/usr/mips --without-headers --with-newlib --with-gnu-as --with-gnu-ld
--with-local-prefix=/usr/mips/mipsisa64sb1-unknown-linux-gnu --disable-shared
--enable-languages=c
Thread model: posix
gcc version 4.1.0 20050901 (experimental)
I'm running out of ideas for workarounds to build a 64-bit bigendian MIPS
toolchain, so any known-to-work workarounds or patches would be much
appreciated.
--
Summary: Crosscompiler looks for files it never built
Product: gcc
Version: 4.1.0
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: imipak at yahoo dot com
CC: gcc-bugs at gcc dot gnu dot org
GCC build triplet: i386-linux-gnu
GCC host triplet: i386-linux-gnu
GCC target triplet: mipsisa64sb1-linux-gnu
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23687
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug target/23687] Crosscompiler looks for files it never built
2005-09-01 22:26 [Bug c/23687] New: Crosscompiler looks for files it never built imipak at yahoo dot com
@ 2005-09-01 22:28 ` pinskia at gcc dot gnu dot org
2005-09-13 7:50 ` rsandifo at gcc dot gnu dot org
1 sibling, 0 replies; 4+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2005-09-01 22:28 UTC (permalink / raw)
To: gcc-bugs
--
What |Removed |Added
----------------------------------------------------------------------------
Component|c |target
Keywords| |build
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23687
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug target/23687] Crosscompiler looks for files it never built
2005-09-01 22:26 [Bug c/23687] New: Crosscompiler looks for files it never built imipak at yahoo dot com
2005-09-01 22:28 ` [Bug target/23687] " pinskia at gcc dot gnu dot org
@ 2005-09-13 7:50 ` rsandifo at gcc dot gnu dot org
1 sibling, 0 replies; 4+ messages in thread
From: rsandifo at gcc dot gnu dot org @ 2005-09-13 7:50 UTC (permalink / raw)
To: gcc-bugs
------- Additional Comments From rsandifo at gcc dot gnu dot org 2005-09-13 07:50 -------
Re crti.o and crtn.o: can I just check that you unpacked unmodified
FSF sources, configured with those arguments, ran "make", and that
"make" completed successfully, but that crti.o and crtn.o were
somehow not built? If so, I'll need a log of the make process.
FWIW, many people build cross mips*-linux-gnu compilers, and I don't
remember ever hearing of this kind of problem before. In case it helps
debugging, the rules to make crti.o and crtn.o are in config/mips/t-elf.
As discussed on the gcc lists, crt1.o and friends are provided
by glibc.
Using newlib for mips*-linux-gnu is basically unsupported. I wouldn't
recommend trying it unless you're really keen on sorting through the
fallout yourself ;)
If you want us to look at the ICE, we'll need preprocessed
input. Please see the bug-reporting instructions for details.
--
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |WAITING
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23687
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug target/23687] Crosscompiler looks for files it never built
[not found] <bug-23687-11278@http.gcc.gnu.org/bugzilla/>
@ 2005-12-17 0:26 ` pinskia at gcc dot gnu dot org
0 siblings, 0 replies; 4+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2005-12-17 0:26 UTC (permalink / raw)
To: gcc-bugs
------- Comment #2 from pinskia at gcc dot gnu dot org 2005-12-17 00:26 -------
No feedback in 3 months.
--
pinskia at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|WAITING |RESOLVED
Resolution| |INVALID
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23687
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2005-12-17 0:26 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-09-01 22:26 [Bug c/23687] New: Crosscompiler looks for files it never built imipak at yahoo dot com
2005-09-01 22:28 ` [Bug target/23687] " pinskia at gcc dot gnu dot org
2005-09-13 7:50 ` rsandifo at gcc dot gnu dot org
[not found] <bug-23687-11278@http.gcc.gnu.org/bugzilla/>
2005-12-17 0:26 ` pinskia at gcc dot gnu dot org
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).