* problem installing gcc 2.95.2 on m68k-coff (fwd) @ 2000-09-21 7:26 Meir Hacohen 2000-09-21 13:06 ` Kai Ruottu 0 siblings, 1 reply; 9+ messages in thread From: Meir Hacohen @ 2000-09-21 7:26 UTC (permalink / raw) To: crossgcc, crossgcc I tried to install a GCC cross compiler 2.95 and 2.95.2 for target m68k-coff and I failed. I tried it at HPUX after I succeeded to install the native compiler. I got some compilation errors while the "make" is running. Following are some of the compilation errors I got : ******************************************************************************* gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc -I../../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include ../../gcc-2.95.2/gcc/config/m68k/m68k.c ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `valid_dbcc_comparison_p': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: `cc_prev_status' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: (Each undeclared identifier is reported only once ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: for each function it appears in.) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: `CC_NO_OVERFLOW' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `flags_in_68881': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:880: `cc_status' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `output_scc_di': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:1010: `CC_STATUS_INIT' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `output_btst': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:1208: `cc_status' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:1208: `CC_NOT_POSITIVE' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:1208: `CC_Z_IN_NOT_N' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:1210: `CC_NOT_NEGATIVE' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `output_move_qimode': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:1651: `CC_STATUS_INIT' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `notice_update_cc': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:2204: `cc_status' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:2210: `CC_STATUS_INIT' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:2284: `CC_NO_OVERFLOW' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c:2291: `CC_NOT_NEGATIVE' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `output_andsi3': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:3298: `CC_STATUS_INIT' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `output_iorsi3': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:3338: `CC_STATUS_INIT' undeclared (first use in this function) ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function `output_xorsi3': ../../gcc-2.95.2/gcc/config/m68k/m68k.c:3376: `CC_STATUS_INIT' undeclared (first use in this function) make[1]: *** [m68k.o] Error 1 make[1]: Leaving directory `/gnu/gcc-obj/gcc' make: *** [all-gcc] Error 2 ******************************************************************************* Is there a known problem in doing this ? I'll appreciate it very much if you can help me to solve this problem. Regards, Meir ******************************************************************************* Meir Hacohen Tel : +972-8-9133394 PLS-DMS-SIR team Fax : +972-8-9133264 Telrad, Israel e-mail: meir.hacohen@telrad.co.il P.O.B. 50, Lod ESN : 828-3394 ******************************************************************************* ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff (fwd) 2000-09-21 7:26 problem installing gcc 2.95.2 on m68k-coff (fwd) Meir Hacohen @ 2000-09-21 13:06 ` Kai Ruottu 2000-09-24 7:25 ` Meir Hacohen 2000-09-25 1:14 ` problem installing gcc 2.95.2 on m68k-coff (fwd) Meir Hacohen 0 siblings, 2 replies; 9+ messages in thread From: Kai Ruottu @ 2000-09-21 13:06 UTC (permalink / raw) To: Meir Hacohen; +Cc: crossgcc, crossgcc Meir Hacohen wrote: > > I tried to install a GCC cross compiler 2.95 and 2.95.2 for target > m68k-coff and I failed. I tried it at HPUX after I succeeded to install > the native compiler. I got some compilation errors while the "make" is > running. > > Following are some of the compilation errors I got : > > ******************************************************************************* > > gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc > -I../../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include > ../../gcc-2.95.2/gcc/config/m68k/m68k.c > ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function > `valid_dbcc_comparison_p': > ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: `cc_prev_status' undeclared > (first use in this function) > ******************************************************************************* > Is there a known problem in doing this ? No AFAIK, but your host can cause this somehow. The hint is : - the 'cc_prev_status' etc. are declared in the 'gcc-2.95.2/gcc/conditions.h'. Could it be possible that your native headers have a file with the same name and it will be used instead of the GCC-one? The 'gcc-2.95.2/gcc/config/m68k/m68k.c' includes it using '"'s instead of '<' and '>', so taking it from the system headers isn't expected... Your build-dir seems to be `/gnu/gcc-obj', and your sources to be in '/gnu/gcc-2.95.2'. Although the '-I../../gcc-2.95.2/gcc' is the second in the list and GCC should search from it after the 'gcc' build dir, and no environment settings for GCC could change this, not even C_INCLUDE_PATH : ------------------------------ clip ---------------------------------------- C_INCLUDE_PATH CPLUS_INCLUDE_PATH OBJC_INCLUDE_PATH These environment variables pertain to particular languages. Each variable's value is a colon-separated list of directories, much like PATH. When GCC searches for header files, it tries the directories listed in the variable for the language you are using, after the directories specified with -I but before the standard header file directories. ------------------------------ clip ---------------------------------------- somehow it takes the header somewhere else... Please try the '-v' option, using 'make CFLAGS="-v -O2"' and see what your 'cpp' gives for the search order for the headers... There must be something wrong now... If the search order is ok, perhaps your 'gcc/conditions.h' has been removed by an accident... BTW, the file(s) where the 'cc_prev_status' is declared could be found by a simple 'grep cc_prev_status *.h *.c' command, done in the 'gcc' source directory. Using 'grep' in cases like this is very useful... Cheers, Kai ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff (fwd) 2000-09-21 13:06 ` Kai Ruottu @ 2000-09-24 7:25 ` Meir Hacohen 2000-09-25 12:31 ` Kai Ruottu 2000-09-25 1:14 ` problem installing gcc 2.95.2 on m68k-coff (fwd) Meir Hacohen 1 sibling, 1 reply; 9+ messages in thread From: Meir Hacohen @ 2000-09-24 7:25 UTC (permalink / raw) To: karuottu; +Cc: Meir Hacohen, crossgcc, crossgcc Hello, Thanks for your help. I checked some things and now I got another error : ************************************************************************** gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc -I../ ../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include ../../gcc-2.95.2/gcc/ dbxout.c cc1: warning: -g is only supported when using GAS on this processor, cc1: warning: -g option disabled. ../../gcc-2.95.2/gcc/dbxout.c: In function `dbxout_start_new_source_file': ../../gcc-2.95.2/gcc/dbxout.c:511: `N_BINCL' undeclared (first use in this funct ion) ../../gcc-2.95.2/gcc/dbxout.c:511: (Each undeclared identifier is reported only once ../../gcc-2.95.2/gcc/dbxout.c:511: for each function it appears in.) ../../gcc-2.95.2/gcc/dbxout.c: In function `dbxout_resume_previous_source_file': ../../gcc-2.95.2/gcc/dbxout.c:523: `N_EINCL' undeclared (first use in this funct ion) ****************************************************************************** I look for this variable - N_BINCL - in some places, and I really do not find it. Can you help again ? Meir ******************************************************************************* Meir Hacohen Tel : +972-8-9133394 PLS-DMS-SIR team Fax : +972-8-9133264 Telrad, Israel e-mail: meir.hacohen@telrad.co.il P.O.B. 50, Lod ESN : 828-3394 ******************************************************************************* On Thu, 21 Sep 2000, Kai Ruottu wrote: > Meir Hacohen wrote: > > > > I tried to install a GCC cross compiler 2.95 and 2.95.2 for target > > m68k-coff and I failed. I tried it at HPUX after I succeeded to install > > the native compiler. I got some compilation errors while the "make" is > > running. > > > > Following are some of the compilation errors I got : > > > > ******************************************************************************* > > > > gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc > > -I../../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include > > ../../gcc-2.95.2/gcc/config/m68k/m68k.c > > ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function > > `valid_dbcc_comparison_p': > > ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: `cc_prev_status' undeclared > > (first use in this function) > > ******************************************************************************* > > Is there a known problem in doing this ? > > No AFAIK, but your host can cause this somehow. The hint is : > > - the 'cc_prev_status' etc. are declared in the 'gcc-2.95.2/gcc/conditions.h'. > Could it be possible that your native headers have a file with the same name > and it will be used instead of the GCC-one? > > The 'gcc-2.95.2/gcc/config/m68k/m68k.c' includes it using '"'s instead of > '<' and '>', so taking it from the system headers isn't expected... > > Your build-dir seems to be `/gnu/gcc-obj', and your sources to be in > '/gnu/gcc-2.95.2'. Although the '-I../../gcc-2.95.2/gcc' is the second > in the list and GCC should search from it after the 'gcc' build dir, and > no environment settings for GCC could change this, not even C_INCLUDE_PATH : > > ------------------------------ clip ---------------------------------------- > C_INCLUDE_PATH > CPLUS_INCLUDE_PATH > OBJC_INCLUDE_PATH > > These environment variables pertain to particular languages. Each variable's > value is a colon-separated list of directories, much like PATH. When GCC > searches for header files, it tries the directories listed in the variable for > the language you are using, after the directories specified with -I but before > the standard header file directories. > ------------------------------ clip ---------------------------------------- > > somehow it takes the header somewhere else... > > Please try the '-v' option, using 'make CFLAGS="-v -O2"' and see what > your 'cpp' gives for the search order for the headers... There must be > something wrong now... If the search order is ok, perhaps your > 'gcc/conditions.h' has been removed by an accident... > > BTW, the file(s) where the 'cc_prev_status' is declared could be found > by a simple 'grep cc_prev_status *.h *.c' command, done in the 'gcc' > source directory. Using 'grep' in cases like this is very useful... > > Cheers, Kai > > > ------ > Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ > Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com > ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff (fwd) 2000-09-24 7:25 ` Meir Hacohen @ 2000-09-25 12:31 ` Kai Ruottu 2000-10-26 6:24 ` problem installing gcc 2.95.2 on m68k-coff Meir Hacohen 0 siblings, 1 reply; 9+ messages in thread From: Kai Ruottu @ 2000-09-25 12:31 UTC (permalink / raw) To: Meir Hacohen; +Cc: crossgcc Meir Hacohen wrote: > > Thanks for your help. I checked some things and now I got another error : > gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc > -I../../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include Somehow the '-DCROSS_COMPILE' is not appearing here... So my bet is that something went wrong with your configure... This was seen already in your original message, but this case revealed the oddity... > ../../gcc-2.95.2/gcc/dbxout.c: In function `dbxout_start_new_source_file': > ../../gcc-2.95.2/gcc/dbxout.c:511: `N_BINCL' undeclared (first use in this > function) > ../../gcc-2.95.2/gcc/dbxout.c:523: `N_EINCL' undeclared (first use in this > function) > > I look for this variable - N_BINCL - in some places, and I really do not > find it. Can you help again ? These seem to be related to the stabs debug format. All cross-compilers should use the 'gstab.h' and the 'stab.def' it includes. The lines 170... in the 'dbxout.c' say: --------------------------- clip ------------------------------------- #if defined (USG) || !defined (HAVE_STAB_H) || defined (CROSS_COMPILE) #include "gstab.h" /* If doing DBX on sysV, use our own stab.h. */ #else #include <stab.h> --------------------------- clip ------------------------------------- If the 'CROSS_COMPILE' would be defined, as it should, there weren't problems with the N_BINCL and N_EINCL, because the 'stab.def' has them. So the configure command you used, seems to be the reason... Although you could just add the missing values (N_BINCL, N_EINCL) into your native '/usr/include/stab.h' (the values can be found in any BSD or Linux headers), doing a 'make distclean', reconfiguring and rebuilding is recommended. Your expected configure-command should be something like: ./configure --build=hppa1.1-hp-hpux10.20 --host=hppa1.1-hp-hpux10.20 \ --target=m68k-coff Ie. your build and host systems are the same, 'hppa1.1-hp-hpux10.20', and your target is 'm68k-coff'... Cheers, Kai ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* problem installing gcc 2.95.2 on m68k-coff 2000-09-25 12:31 ` Kai Ruottu @ 2000-10-26 6:24 ` Meir Hacohen 2000-10-27 4:13 ` Kai Ruottu 0 siblings, 1 reply; 9+ messages in thread From: Meir Hacohen @ 2000-10-26 6:24 UTC (permalink / raw) To: karuottu; +Cc: Meir Hacohen, crossgcc Hello, I tried to install gcc cross compilers, and that is what I got : 1. When I work with target m68k-aout I use the command : ../gcc-2.95.2/configure --build=hppa1.1-hp-hpux10.20 --host=hppa1.1-hp-hpux10.20 --target=m68k-aout --with-headers=/gnu/newlib-1.8.2 and the configure works fine. But after the make command I got at the end : _mulsi3 as: unrecognized option `-mc68020' make[1]: *** [libgcc1-asm.a] Error 1 make[1]: Leaving directory `/gnu/gcc-obj/gcc' make: *** [all-gcc] Error 2 2. When I work with target m68k-coff I use the command : ../gcc-2.95.2/configure --build=hppa1.1-hp-hpux10.20 --host=hppa1.1-hp-hpux10.2$ --target=m68k-coff --with-headers=/gnu/newlib-1.8.2 and the configure works fine. But after the make command I got at the end : make[2]: Entering directory `/gnu/gcc-obj/m68k-coff/libiberty' make[2]: *** No targets. Stop. make[2]: Leaving directory `/gnu/gcc-obj/m68k-coff/libiberty' make[1]: *** [../libiberty/libiberty.a] Error 2 make[1]: Leaving directory `/gnu/gcc-obj/m68k-coff/libstdc++' make: *** [all-target-libstdc++] Error 2 Can someone tell me what is the meaning of those errors , and how can I fix them ? Regards, Meir ******************************************************************************* Meir Hacohen Tel : +972-8-9133394 PLS-DMS-SIR team Fax : +972-8-9133264 Telrad, Israel e-mail: meir.hacohen@telrad.co.il P.O.B. 50, Lod ESN : 828-3394 ******************************************************************************* On Mon, 25 Sep 2000, Kai Ruottu wrote: > Meir Hacohen wrote: > > > > Thanks for your help. I checked some things and now I got another error : > > gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc > > -I../../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include > > Somehow the '-DCROSS_COMPILE' is not appearing here... So my bet is that > something went wrong with your configure... > > This was seen already in your original message, but this case revealed the > oddity... > > > ../../gcc-2.95.2/gcc/dbxout.c: In function `dbxout_start_new_source_file': > > ../../gcc-2.95.2/gcc/dbxout.c:511: `N_BINCL' undeclared (first use in this > > function) > > ../../gcc-2.95.2/gcc/dbxout.c:523: `N_EINCL' undeclared (first use in this > > function) > > > > I look for this variable - N_BINCL - in some places, and I really do not > > find it. Can you help again ? > > These seem to be related to the stabs debug format. All cross-compilers > should use the 'gstab.h' and the 'stab.def' it includes. The lines 170... > in the 'dbxout.c' say: > > --------------------------- clip ------------------------------------- > #if defined (USG) || !defined (HAVE_STAB_H) || defined (CROSS_COMPILE) > #include "gstab.h" /* If doing DBX on sysV, use our own stab.h. */ > #else > #include <stab.h> > --------------------------- clip ------------------------------------- > > If the 'CROSS_COMPILE' would be defined, as it should, there weren't > problems with the N_BINCL and N_EINCL, because the 'stab.def' has them. > So the configure command you used, seems to be the reason... > > Although you could just add the missing values (N_BINCL, N_EINCL) into > your native '/usr/include/stab.h' (the values can be found in any BSD > or Linux headers), doing a 'make distclean', reconfiguring and rebuilding > is recommended. > > Your expected configure-command should be something like: > > ./configure --build=hppa1.1-hp-hpux10.20 --host=hppa1.1-hp-hpux10.20 \ > --target=m68k-coff > > Ie. your build and host systems are the same, 'hppa1.1-hp-hpux10.20', > and your target is 'm68k-coff'... > > Cheers, Kai > > ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff 2000-10-26 6:24 ` problem installing gcc 2.95.2 on m68k-coff Meir Hacohen @ 2000-10-27 4:13 ` Kai Ruottu 2000-10-31 1:28 ` Meir Hacohen 0 siblings, 1 reply; 9+ messages in thread From: Kai Ruottu @ 2000-10-27 4:13 UTC (permalink / raw) To: Meir Hacohen; +Cc: crossgcc [-- Warning: decoded text below may be mangled, UTF-8 assumed --] [-- Attachment #1: Type: text/plain, Size: 4970 bytes --] Meir Hacohen wrote: > > Hello, > > I tried to install gcc cross compilers, and that is what I got : > > 1. When I work with target m68k-aout I use the command : > > ../gcc-2.95.2/configure --build=hppa1.1-hp-hpux10.20 --host=hppa1.1-hp-hpux10.20 --target=m68k-aout > --with-headers=/gnu/newlib-1.8.2 > > and the configure works fine. > > But after the make command I got at the end : > > _mulsi3 > as: unrecognized option `-mc68020' > make[1]: *** [libgcc1-asm.a] Error 1 > make[1]: Leaving directory `/gnu/gcc-obj/gcc' > make: *** [all-gcc] Error 2 It doesn't find your 'm68k-aout' target assembler. It should be in the default place, '$prefix/$target/bin', ie. in your case in '/usr/local/m68k-aout/bin'. The default $prefix is '/usr/local' (used when the '--prefix=something' wasn't used). If the right assembler is not found, the wrong 'as' from the $PATH will be used. Perhaps you forgot to build the GNU binutils for 'm68k-aout' and install them... > 2. When I work with target m68k-coff I use the command : > > ../gcc-2.95.2/configure --build=hppa1.1-hp-hpux10.20 > --host=hppa1.1-hp-hpux10.2$ --target=m68k-coff > --with-headers=/gnu/newlib-1.8.2 > > and the configure works fine. > > But after the make command I got at the end : > > make[2]: Entering directory `/gnu/gcc-obj/m68k-coff/libiberty' > make[2]: *** No targets. Stop. > make[2]: Leaving directory `/gnu/gcc-obj/m68k-coff/libiberty' > make[1]: *** [../libiberty/libiberty.a] Error 2 > make[1]: Leaving directory `/gnu/gcc-obj/m68k-coff/libstdc++' > make: *** [all-target-libstdc++] Error 2 > > Can someone tell me what is the meaning of those errors , and how can I fix > them ? All the m68k-* embedded 'targets' come without a default 'real target', ie. the target board and the startup and glue library for it are unknown. The Cygnus newlib comes with some Board Support Packages and you could select one of these as your temporary default target. The libiberty is a library which should include all those routines missing from your standard C-library (newlib here) and before trying to build it a simple program is tried to be compiled and linked. This must succeed otherwise there are no sanity in the libiberty build. So you must edit a real target into your 'gcc/specs' file while building, ie. define the linker script and the used libs there. The startup (crt0.o) is defined in the linker script and the spec for it must be empty. Something like: -------------- clip --------------------- *endfile: *link: -m m68kcoff -T bcc.ld%s *lib: -lc -lbcc -lc *libgcc: -lgcc *startfile: *switches_need_spaces: -------------- clip --------------------- must then be seen in your 'specs'. Here the BCC-board was selected as the 'real target'. In the phase you are, the GCC is already built (only the stuff in 'gcc' belongs to the compiler). When you haven't installed the 'newlib' sources into the same directory with 'gcc', 'libiberty', 'libstdc++' etc., it is now a good time to install GCC from the 'gcc' subdir. Use 'make all-install-gcc' or something... Then configure and build newlib for the m68k-coff and install it. After that you should have the BCC and other support libs among your standard C-library. The linker scripts too... -------------------- clip ----------------------- Directory of E:\usr\local\m68k-coff\lib 11.08.2000 11:40 <DIR> . 11.08.2000 11:40 <DIR> .. 30.10.1998 06:57 2Â 790 bcc.ld 24.03.1999 17:52 934 crt0.o 27.10.1999 20:03 3Â 054 idp.ld 24.03.1999 16:08 <DIR> ldscripts 24.03.1999 17:52 48Â 462 libbcc.a 24.03.1999 17:20 194Â 010 libc.a 24.03.1999 17:52 48Â 634 libdbug.a 24.03.1999 17:20 194Â 010 libg.a 24.03.1999 17:52 60Â 584 libidp.a 24.03.1999 17:18 257Â 456 libm.a 24.03.1999 17:52 48Â 462 libmvme135.a 24.03.1999 17:52 48Â 462 libmvme162.a 24.03.1999 16:08 <DIR> m5200 24.03.1999 16:08 <DIR> m68000 24.03.1999 16:08 <DIR> m68020 24.03.1999 16:08 <DIR> m68040 24.03.1999 16:08 <DIR> m68060 30.10.1998 06:57 2Â 776 mvme135.ld 30.10.1998 06:57 2Â 828 mvme162.ld 30.10.1998 06:57 2Â 784 sbc5204.ld 30.10.1998 06:57 2Â 782 sbc5206.ld -------------------- clip ----------------------- Then you can try the extra libs coming with gcc-2.95.2 again and the 'make' should now work (of course the 'specs' must be edited to use the linker script and the glue library). The 'config.log' in 'm68k-coff/libiberty' will show the reason if it still fails... Cheers, Kai ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff 2000-10-27 4:13 ` Kai Ruottu @ 2000-10-31 1:28 ` Meir Hacohen 0 siblings, 0 replies; 9+ messages in thread From: Meir Hacohen @ 2000-10-31 1:28 UTC (permalink / raw) To: karuottu; +Cc: crossgcc [-- Warning: decoded text below may be mangled, UTF-8 assumed --] [-- Attachment #1: Type: text/plain, Size: 6921 bytes --] Hello, Kai ! Thanks a lot for your help. I did what you suggest (about the m68k-coff) and now I got : ******************************************************************** /gnu/gcc-obj/gcc/xgcc -B/gnu/gcc-obj/gcc/ -B/usr/local/m68k-coff/bin/ -I/usr/loc al/m68k-coff/include -DCROSS_COMPILE -DIN_GCC -O2 -I./include -c ../../g cc-2.95.2/gcc/libgcc1-test.c ../../gcc-2.95.2/gcc/libgcc1-test.c:101: warning: conflicting types for built-in function `memcpy' Testing libgcc1. Ignore linker warning messages. /gnu/gcc-obj/gcc/xgcc -B/gnu/gcc-obj/gcc/ -B/usr/local/m68k-coff/bin/ -I/usr/loc al/m68k-coff/include -DCROSS_COMPILE -DIN_GCC -O2 -I./include libgcc1-test. o -o libgcc1-test \ -nostartfiles -nostdlib `/gnu/gcc-obj/gcc/xgcc -B/gnu/gcc-obj/gcc/ -B/usr/loca l/m68k-coff/bin/ -I/usr/local/m68k-coff/include --print-libgcc-file-name` /usr/local/m68k-coff/bin/ld: cannot open -lc: No such file or directory collect2: ld returned 1 exit status make[1]: *** [libgcc1-test] Error 1 make[1]: Leaving directory `/gnu/gcc-obj/gcc' make: *** [all-gcc] Error 2 ********************************************************************** What should I do now ? Regards, Meir ******************************************************************************* Meir Hacohen Tel : +972-8-9133394 PLS-DMS-SIR team Fax : +972-8-9133264 Telrad, Israel e-mail: meir.hacohen@telrad.co.il P.O.B. 50, Lod ESN : 828-3394 ******************************************************************************* On Fri, 27 Oct 2000, Kai Ruottu wrote: > Meir Hacohen wrote: > > > > Hello, > > > > I tried to install gcc cross compilers, and that is what I got : > > > > 1. When I work with target m68k-aout I use the command : > > > > ../gcc-2.95.2/configure --build=hppa1.1-hp-hpux10.20 --host=hppa1.1-hp-hpux10.20 --target=m68k-aout > > --with-headers=/gnu/newlib-1.8.2 > > > > and the configure works fine. > > > > But after the make command I got at the end : > > > > _mulsi3 > > as: unrecognized option `-mc68020' > > make[1]: *** [libgcc1-asm.a] Error 1 > > make[1]: Leaving directory `/gnu/gcc-obj/gcc' > > make: *** [all-gcc] Error 2 > > It doesn't find your 'm68k-aout' target assembler. It should be in the default place, > '$prefix/$target/bin', ie. in your case in '/usr/local/m68k-aout/bin'. The default > $prefix is '/usr/local' (used when the '--prefix=something' wasn't used). If the right > assembler is not found, the wrong 'as' from the $PATH will be used. Perhaps you forgot > to build the GNU binutils for 'm68k-aout' and install them... > > > 2. When I work with target m68k-coff I use the command : > > > > ../gcc-2.95.2/configure --build=hppa1.1-hp-hpux10.20 > > --host=hppa1.1-hp-hpux10.2$ --target=m68k-coff > > --with-headers=/gnu/newlib-1.8.2 > > > > and the configure works fine. > > > > But after the make command I got at the end : > > > > make[2]: Entering directory `/gnu/gcc-obj/m68k-coff/libiberty' > > make[2]: *** No targets. Stop. > > make[2]: Leaving directory `/gnu/gcc-obj/m68k-coff/libiberty' > > make[1]: *** [../libiberty/libiberty.a] Error 2 > > make[1]: Leaving directory `/gnu/gcc-obj/m68k-coff/libstdc++' > > make: *** [all-target-libstdc++] Error 2 > > > > Can someone tell me what is the meaning of those errors , and how can I fix > > them ? > > All the m68k-* embedded 'targets' come without a default 'real target', ie. > the target board and the startup and glue library for it are unknown. The > Cygnus newlib comes with some Board Support Packages and you could select > one of these as your temporary default target. The libiberty is a library > which should include all those routines missing from your standard C-library > (newlib here) and before trying to build it a simple program is tried to be > compiled and linked. This must succeed otherwise there are no sanity in the > libiberty build. > > So you must edit a real target into your 'gcc/specs' file while building, > ie. define the linker script and the used libs there. The startup (crt0.o) > is defined in the linker script and the spec for it must be empty. Something > like: > > -------------- clip --------------------- > *endfile: > > > *link: > -m m68kcoff -T bcc.ld%s > > *lib: > -lc -lbcc -lc > > *libgcc: > -lgcc > > *startfile: > > > *switches_need_spaces: > -------------- clip --------------------- > > must then be seen in your 'specs'. Here the BCC-board was selected as the > 'real target'. > > In the phase you are, the GCC is already built (only the stuff in 'gcc' > belongs to the compiler). When you haven't installed the 'newlib' sources > into the same directory with 'gcc', 'libiberty', 'libstdc++' etc., it is > now a good time to install GCC from the 'gcc' subdir. Use 'make all-install-gcc' > or something... Then configure and build newlib for the m68k-coff and install > it. After that you should have the BCC and other support libs among your > standard C-library. The linker scripts too... > > -------------------- clip ----------------------- > Directory of E:\usr\local\m68k-coff\lib > > 11.08.2000 11:40 <DIR> . > 11.08.2000 11:40 <DIR> .. > 30.10.1998 06:57 2Â 790 bcc.ld > 24.03.1999 17:52 934 crt0.o > 27.10.1999 20:03 3Â 054 idp.ld > 24.03.1999 16:08 <DIR> ldscripts > 24.03.1999 17:52 48Â 462 libbcc.a > 24.03.1999 17:20 194Â 010 libc.a > 24.03.1999 17:52 48Â 634 libdbug.a > 24.03.1999 17:20 194Â 010 libg.a > 24.03.1999 17:52 60Â 584 libidp.a > 24.03.1999 17:18 257Â 456 libm.a > 24.03.1999 17:52 48Â 462 libmvme135.a > 24.03.1999 17:52 48Â 462 libmvme162.a > 24.03.1999 16:08 <DIR> m5200 > 24.03.1999 16:08 <DIR> m68000 > 24.03.1999 16:08 <DIR> m68020 > 24.03.1999 16:08 <DIR> m68040 > 24.03.1999 16:08 <DIR> m68060 > 30.10.1998 06:57 2Â 776 mvme135.ld > 30.10.1998 06:57 2Â 828 mvme162.ld > 30.10.1998 06:57 2Â 784 sbc5204.ld > 30.10.1998 06:57 2Â 782 sbc5206.ld > -------------------- clip ----------------------- > > Then you can try the extra libs coming with gcc-2.95.2 again > and the 'make' should now work (of course the 'specs' must be > edited to use the linker script and the glue library). The > 'config.log' in 'm68k-coff/libiberty' will show the reason if > it still fails... > > Cheers, Kai > ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff (fwd) 2000-09-21 13:06 ` Kai Ruottu 2000-09-24 7:25 ` Meir Hacohen @ 2000-09-25 1:14 ` Meir Hacohen 2000-09-25 12:31 ` Kai Ruottu 1 sibling, 1 reply; 9+ messages in thread From: Meir Hacohen @ 2000-09-25 1:14 UTC (permalink / raw) To: karuottu; +Cc: Meir Hacohen, crossgcc, crossgcc Hello, Thanks for the help. I use the option 'make CFLAGS="-v -O2"' and I found, that there is no search list for the include "... " . How can I solve this problem. Following is the relevant output : ****************************************************************************** GNU CPP version 2.95.2 19991024 (release) (hppa) #include "..." search starts here: #include <...> search starts here: . ../../gcc-2.95.2/gcc ../../gcc-2.95.2/gcc/config ../../gcc-2.95.2/gcc/../include /usr/local/include /usr/local/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.95.2/../../../../hppa1.1-hp-hpux1 0.20/include /usr/local/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.95.2/include /usr/include End of search list. ****************************************************************************** Regards, Meir ******************************************************************************* Meir Hacohen Tel : +972-8-9133394 PLS-DMS-SIR team Fax : +972-8-9133264 Telrad, Israel e-mail: meir.hacohen@telrad.co.il P.O.B. 50, Lod ESN : 828-3394 ******************************************************************************* On Thu, 21 Sep 2000, Kai Ruottu wrote: > Meir Hacohen wrote: > > > > I tried to install a GCC cross compiler 2.95 and 2.95.2 for target > > m68k-coff and I failed. I tried it at HPUX after I succeeded to install > > the native compiler. I got some compilation errors while the "make" is > > running. > > > > Following are some of the compilation errors I got : > > > > ******************************************************************************* > > > > gcc -c -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc > > -I../../gcc-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include > > ../../gcc-2.95.2/gcc/config/m68k/m68k.c > > ../../gcc-2.95.2/gcc/config/m68k/m68k.c: In function > > `valid_dbcc_comparison_p': > > ../../gcc-2.95.2/gcc/config/m68k/m68k.c:869: `cc_prev_status' undeclared > > (first use in this function) > > ******************************************************************************* > > Is there a known problem in doing this ? > > No AFAIK, but your host can cause this somehow. The hint is : > > - the 'cc_prev_status' etc. are declared in the 'gcc-2.95.2/gcc/conditions.h'. > Could it be possible that your native headers have a file with the same name > and it will be used instead of the GCC-one? > > The 'gcc-2.95.2/gcc/config/m68k/m68k.c' includes it using '"'s instead of > '<' and '>', so taking it from the system headers isn't expected... > > Your build-dir seems to be `/gnu/gcc-obj', and your sources to be in > '/gnu/gcc-2.95.2'. Although the '-I../../gcc-2.95.2/gcc' is the second > in the list and GCC should search from it after the 'gcc' build dir, and > no environment settings for GCC could change this, not even C_INCLUDE_PATH : > > ------------------------------ clip ---------------------------------------- > C_INCLUDE_PATH > CPLUS_INCLUDE_PATH > OBJC_INCLUDE_PATH > > These environment variables pertain to particular languages. Each variable's > value is a colon-separated list of directories, much like PATH. When GCC > searches for header files, it tries the directories listed in the variable for > the language you are using, after the directories specified with -I but before > the standard header file directories. > ------------------------------ clip ---------------------------------------- > > somehow it takes the header somewhere else... > > Please try the '-v' option, using 'make CFLAGS="-v -O2"' and see what > your 'cpp' gives for the search order for the headers... There must be > something wrong now... If the search order is ok, perhaps your > 'gcc/conditions.h' has been removed by an accident... > > BTW, the file(s) where the 'cc_prev_status' is declared could be found > by a simple 'grep cc_prev_status *.h *.c' command, done in the 'gcc' > source directory. Using 'grep' in cases like this is very useful... > > Cheers, Kai > > > ------ > Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ > Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com > ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: problem installing gcc 2.95.2 on m68k-coff (fwd) 2000-09-25 1:14 ` problem installing gcc 2.95.2 on m68k-coff (fwd) Meir Hacohen @ 2000-09-25 12:31 ` Kai Ruottu 0 siblings, 0 replies; 9+ messages in thread From: Kai Ruottu @ 2000-09-25 12:31 UTC (permalink / raw) To: Meir Hacohen; +Cc: crossgcc Meir Hacohen wrote: > > Thanks for the help. I use the option 'make CFLAGS="-v -O2"' and I > found, that there is no search list for the include "... " . How can I > solve this problem. Another interpretation could be that the search lists are identical. If quoting the GNU cpp manual: ---------------------- clip ------------------------------------ #include "file" This variant is used for header files of your own program. It searches for a file named file first in the current directory, then in the same directories used for system header files. The current directory is the directory of the current input file. It is tried first because it is presumed to be the location of the files that the current input file refers to. (If the -I- option is used, the special treatment of the current directory is inhibited.) ---------------------- clip ------------------------------------ So my memory about using the '"'s around the header names was wrong, using them causes searching also from the system headers. > Following is the relevant output : > > GNU CPP version 2.95.2 19991024 (release) (hppa) > #include "..." search starts here: > #include <...> search starts here: > . > ../../gcc-2.95.2/gcc > ../../gcc-2.95.2/gcc/config > ../../gcc-2.95.2/gcc/../include > /usr/local/include > > /usr/local/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.95.2/../../../../hppa1.1-hp-hpux1 > 0.20/include > /usr/local/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.95.2/include > /usr/include The first four come from the '-I' options on the GCC command line... So these seem to be quite ok for a native compiler, although the one with the 'relative path' may cause some questions, e.g. quite many Linux-distributions may have it symlinked to the standard headers in '/usr/include', and not as an empty directory for some extra headers, as the purpose for it seems to be. A cross-compiler has the standard target headers there, so it will be included last. Cheers, Kai ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com ^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2000-10-31 1:28 UTC | newest] Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2000-09-21 7:26 problem installing gcc 2.95.2 on m68k-coff (fwd) Meir Hacohen 2000-09-21 13:06 ` Kai Ruottu 2000-09-24 7:25 ` Meir Hacohen 2000-09-25 12:31 ` Kai Ruottu 2000-10-26 6:24 ` problem installing gcc 2.95.2 on m68k-coff Meir Hacohen 2000-10-27 4:13 ` Kai Ruottu 2000-10-31 1:28 ` Meir Hacohen 2000-09-25 1:14 ` problem installing gcc 2.95.2 on m68k-coff (fwd) Meir Hacohen 2000-09-25 12:31 ` Kai Ruottu
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).