9:14 a.m.
Are there any test s390x instances available for a peon like me to use
to diagnose a build failure? Failing that, does anybody with access
to an s390x instance have time to help me? The latest bigloo release
built successfully on all arches but s390x:
https://koji.fedoraproject.org/koji/taskinfo?taskID=20740401
The failure comes after the main binary has been built and linked
successfully, but then is executed to compile some Scheme files. The
binary is reporting that it was invoked on an empty input file.
Thank you,
--
Jerry James
http://www.jamezone.org/
4:10 a.m.
On Wed, Jul 26, 2017 at 08:14:14AM -0600, Jerry James wrote:
Failing that, does anybody with access
to an s390x instance have time to help me? The latest bigloo release
built successfully on all arches but s390x:
https://koji.fedoraproject.org/koji/taskinfo?taskID=20740401
The failure comes after the main binary has been built and linked
successfully, but then is executed to compile some Scheme files. The
binary is reporting that it was invoked on an empty input file.
Well it's certainly reproducible:
/home/nfs/rjones/d/fedora/bigloo/master/bigloo4.3a/bin/bigloo.sh -srfi enable-unistring
-srfi enable-resolv -srfi enable-gmp -srfi enable-pcre -unsafe -q -mkheap -mklib -s
-no-hello -L /home/nfs/rjones/d/fedora/bigloo/master/bigloo4.3a/lib/bigloo/4.3a
Llib/make_lib.scm -heap bigloo.heap
*** ERROR:top-level:Parse error:
Illegal source file -- #eof-object
I ran the same command under gdb to get a stack trace:
Thread 1 (Thread 0x3fffdff4720 (LWP 38254)):
#0 0x000003fffd843ea8 in exit () from /lib64/libc.so.6
#1 0x000003fffdc3893e in bigloo_exit (val=val@entry=0x9) at Clib/cmain.c:118
#2 0x000003fffddb8c80 in BGl_exitz00zz__errorz00 (BgL_nz00_13=<optimized out>)
at objs/obj_s/Llib/error.c:1530
#3 0x000003fffddb9066 in BGl_raisez00zz__errorz00 (BgL_valz00_16=0x16e5480)
at objs/obj_s/Llib/error.c:1823
#4 0x000003fffddb9516 in BGl_withzd2exceptionzd2handlerz00zz__errorz00 (
BgL_handlerz00_14=<optimized out>,
BgL_thunkz00_15=BgL_thunkz00_15@entry=0x178f640)
at objs/obj_s/Llib/error.c:1595
#5 0x00000000011a3bd6 in BGl_z62zc3anonymousza31407ze3ze1zztools_errorz00 (
BgL_envz00_1676=<optimized out>) at Tools/error.c:1342
#6 0x00000000011a3e56 in BGl_withzd2dumpzd2stackz00zztools_errorz00 (
BgL_thunkz00_20=0x179d9c3) at Tools/error.c:548
#7 0x00000000011a465e in BGl_userzd2errorzd2zztools_errorz00 (
BgL_procz00_31=BgL_procz00_31@entry=0x12f0340
<BgL_bgl_string1902za700za7za7e1953za7+8>,
BgL_mesz00_32=BgL_mesz00_32@entry=0x12f0308
<BgL_bgl_string1903za700za7za7e1954za7+8>, BgL_objz00_33=BgL_objz00_33@entry=0x802,
BgL_continuez00_34=BgL_continuez00_34@entry=0x2) at Tools/error.c:1041
#8 0x000000000100c84c in BGl_compilerz00zzengine_compilerz00 ()
at Engine/compiler.c:802
#9 0x0000000001233244 in BGl_mainz00zzinit_mainz00 (
BgL_argvz00_3=BgL_argvz00_3@entry=0x1517d43) at Init/main.c:274
#10 0x000000000123331e in bigloo_main (BgL_argvz00_46=0x1517d43)
at Init/main.c:144
#11 0x000003fffdc38bb4 in _bigloo_main (argc=<optimized out>,
argv=<optimized out>, env=<optimized out>,
bigloo_main=bigloo_main@entry=0x12332c8 <bigloo_main>,
libinit=libinit@entry=0x1232f28 <bigloo_libinit>, uheapsize=0)
at Clib/cmain.c:229
#12 0x000000000100befc in main (argc=<optimized out>, argv=<optimized out>,
env=<optimized out>) at Init/main.c:107
Unfortunately once I started to look at the code it seems to be some
kind of generated C which is almost impossible to follow, eg:
(gdb) frame 8
#8 0x000000000100c84c in BGl_compilerz00zzengine_compilerz00 ()
at Engine/compiler.c:802
802 BGl_userzd2errorzd2zztools_errorz00
(gdb) list
797 MAKE_YOUNG_PAIR(BINT(((long) 1)), BNIL);
798 BGl_exitz00zz__errorz00(BgL_list1426z00_1638);
799 }
800 else
801 {/* Engine/compiler.scm 123 */
802 BGl_userzd2errorzd2zztools_errorz00
803 (BGl_string1902z00zzengine_compilerz00,
804 BGl_string1903z00zzengine_compilerz00,
BgL_srcz00_1635,
805 BNIL);
806 }
Are there any test s390x instances available for a peon like me to
use
to diagnose a build failure?
You can ask for access to a machine on #fedora-s390x (Freenode).
Rich.
--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
virt-builder quickly builds VMs from scratch
http://libguestfs.org/virt-builder.1.html
9:29 a.m.
On Thu, Jul 27, 2017 at 3:10 AM, Richard W.M. Jones <rjones(a)redhat.com> wrote:
Unfortunately once I started to look at the code it seems to be some
kind of generated C which is almost impossible to follow, eg:
Yes, it is Scheme code compiled to C. Lots of fun to dig through. :-)
You can ask for access to a machine on #fedora-s390x (Freenode).
Dan helped me out, which is good because I have several other packages
that failed the mass rebuild on s390x only.
Regards,
--
Jerry James
http://www.jamezone.org/
2486
days inactive
2487
days old
2 comments
2 participants
participants (2)
-
Jerry James -
Richard W.M. Jones