Apache OpenOffice (AOO) Bugzilla – Issue 4468
Installation aborted in RedHat 7.1, Kernel 2.4.18, glibc-2.2.4, and redhat updates.
Last modified: 2013-08-07 15:26:39 UTC
When I run the install script (or even setup) with or without flags, all I get is the decompress progress bar (in X) showing the decompression - a progress bar with a little box aside, and then after it decompresses, it stops. In the console, it shows up the glibc version and then after the progress decompression is over, it returns to prompt. The "intall" script will return "Installation Finished" without having installed anything. This is a Red Hat 7.1 system with various core system packages updated to be compatible with Red Hat 7.2, such as kernel (2.4.18), glibc (2.2.4), libstdc++ (2.96-85), XFree86 (4.1.0). I also have gcc downgraded to 2.95.3 - that is installed in /usr/local. The original gcc 2.96 rpms were removed, being only left the libstdc++ run time libraries.
============================================== an additional note: Celeron 400Mhz 160MB RAM 2GB+ Free disk space Video 8MB at 16Bit color
---------------------------------- One more note: The OpenOffice I am trying to install is the binary version for Linux Intel. It is not compiled from source. Plus, the MD5SUM is perfect, matching the original.
============================== CHECK THIS OUT ============================== 1) Someone from users list suggested to look if I was running libsafe. Answer: No. Don't even have it installed. 2) Other person suggested to ldd the setup file - here is the result: [quantum@linux install]$ ldd setup libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x40024000) libdl.so.2 => /lib/libdl.so.2 (0x4011a000) libpthread.so.0 => /lib/libpthread.so.0 (0x4011e000) libm.so.6 => /lib/libm.so.6 (0x40135000) libc.so.6 => /lib/libc.so.6 (0x40157000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000) [quantum@linux install]$ 3) Then I straced the binary setup. And analysing the output, I confirmed the following ABNORMAL LINES that are causing the setup to exit without installing. Those lines are just before the cleaning up. fork() = 836 wait4(836, [WIFSIGNALED(s) && WTERMSIG(s) == SIGSEGV && COREDUMP(s)],0, NULL) = 836 --- SIGCHLD (Child exited) --- So this is the hint for the engineers to look at. Please respond ASAP.
========================= ANOTHER SIGCHLD FOUND ========================= This is the very beginning of strace output. As you can notice, youll find another SIGCHLD in the end of this code. But the process continues and there is another one in the end - See previous comment. execve("./setup", ["./setup"], [/* 30 vars */]) = 0 uname({sys="Linux", node="linux", ...}) = 0 brk(0) = 0x8065764 old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40016000 open("/etc/ld.so.preload", O_RDONLY) = -1 ENOENT (No such file or directory) open("/etc/ld.so.cache", O_RDONLY) = 3 fstat64(3, {st_mode=S_IFREG|0644, st_size=51081, ...}) = 0 old_mmap(NULL, 51081, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40017000 close(3) = 0 open("/usr/X11R6/lib/libX11.so.6", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\260E\1"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=1008216, ...}) = 0 old_mmap(NULL, 1004248, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40024000 mprotect(0x40115000, 17112, PROT_NONE) = 0 old_mmap(0x40115000, 16384, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0xf0000) = 0x40115000 old_mmap(0x40119000, 728, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40119000 close(3) = 0 open("/lib/libdl.so.2", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\200\35"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=65405, ...}) = 0 old_mmap(NULL, 12532, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4011a000 mprotect(0x4011c000, 4340, PROT_NONE) = 0 old_mmap(0x4011c000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x1000) = 0x4011c000 close(3) = 0 open("/lib/libpthread.so.0", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\360O\0"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=557615, ...}) = 0 old_mmap(NULL, 90460, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4011e000 mprotect(0x4012d000, 29020, PROT_NONE) = 0 old_mmap(0x4012d000, 28672, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0xe000) = 0x4012d000 old_mmap(0x40134000, 348, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40134000 close(3) = 0 open("/lib/libm.so.6", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\320H\0"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=615554, ...}) = 0 old_mmap(NULL, 138964, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40135000 mprotect(0x40156000, 3796, PROT_NONE) = 0 old_mmap(0x40156000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x20000) = 0x40156000 close(3) = 0 open("/lib/libc.so.6", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\360\303"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=5723311, ...}) = 0 old_mmap(NULL, 1265288, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40157000 mprotect(0x40283000, 36488, PROT_NONE) = 0 old_mmap(0x40283000, 20480, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x12b000) = 0x40283000 old_mmap(0x40288000, 16008, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40288000 close(3) = 0 old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x4028c000 munmap(0x40017000, 51081) = 0 getrlimit(0x3, 0xbffff6d4) = 0 setrlimit(RLIMIT_STACK, {rlim_cur=2044*1024, rlim_max=RLIM_INFINITY}) = 0 getpid() = 833 rt_sigaction(SIGRT_0, {0x40127ec0, [], 0x4000000}, NULL, 8) = 0 rt_sigaction(SIGRT_1, {0x401274bc, [], 0x4000000}, NULL, 8) = 0 rt_sigaction(SIGRT_2, {0x40127f4c, [], 0x4000000}, NULL, 8) = 0 rt_sigprocmask(SIG_BLOCK, [RT_0], NULL, 8) = 0 _sysctl({{CTL_KERN, KERN_VERSION}, 2, 0xbffff6dc, 31, (nil), 0}) = 0 brk(0) = 0x8065764 brk(0x8065794) = 0x8065794 brk(0x8066000) = 0x8066000 pipe([3, 4]) = 0 fork() = 834 close(4) = 0 fstat64(3, {st_mode=S_IFIFO|0600, st_size=0, ...}) = 0 old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40017000 read(3, "/lib:\n\tlibnss_ldap-2.2.2.so -> l"..., 4096) = 4096 read(3, "ibkickermain.so.1 -> libkickerma"..., 4096) = 4096 read(3, "2.so\n\tlibscrollkeeper.so.0 -> li"..., 4096) = 4096 read(3, "srpc.so.3 -> libgssrpc.so.3.0\n\tl"..., 4096) = 2843 read(3, 0x40017000, 4096) = ? ERESTARTSYS (To be restarted) --- SIGCHLD (Child exited) --- read(3, "", 4096) = 0 close(3)
======================================== FINAL PART OF THE STRACE OUTPUT OF SETUP ======================================== Initially I introduced 3 lines. Now I am introducing the whole set of code that is by the end of the process. Notice the [SIGCHLD Exit] just before the deletion/unlinking of files. munmap(0x40018000, 4096) = 0 chmod("/tmp/sv001.tmp/setup.bin", 0700) = 0 chmod("/tmp/sv001.tmp/sopatchlevel.sh", 0700) = 0 chmod("/tmp/sv001.tmp/sorev", 0700) = -1 ENOENT (No such file or directory) chmod("/tmp/sv001.tmp/setup", 0700) = 0 open("/tmp/sv001.tmp", O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY) = 6 fstat64(6, {st_mode=S_IFDIR|0700, st_size=4096, ...}) = 0 fcntl64(6, F_SETFD, FD_CLOEXEC) = 0 getdents64(6, /* 66 entries */, 4096) = 2464 symlink("/tmp/sv001.tmp/libstore.so.3.0.1", "/tmp/sv001.tmp/libstore.so.3") = 0 symlink("/tmp/sv001.tmp/libreg.so.3.0.1", "/tmp/sv001.tmp/libreg.so.3") = 0 symlink("/tmp/sv001.tmp/libsal.so.3.0.1", "/tmp/sv001.tmp/libsal.so.3") = 0 symlink("/tmp/sv001.tmp/libstdc++.so.3.0.4", "/tmp/sv001.tmp/libstdc++.so.3") = 0 symlink("/tmp/sv001.tmp/libcppu.so.3.0.1", "/tmp/sv001.tmp/libcppu.so.3") = 0 getdents64(6, /* 0 entries */, 4096) = 0 close(6) = 0 ioctl(3, FIONREAD, [0]) = 0 write(3, "\31\0\v\0\4\0 \3\0\0\0\0! \0\0\4\0 \3\3\0\0\0\3\0\0\0\0"..., 44) = 44 read(3, "\241 \340\10\4\0 \3\3\0\0\0\3\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 32) = 32 ioctl(3, FIONREAD, [0]) = 0 write(3, "\n\0\2\0\4\0 \3\31\0\v\0\4\0 \3\0\0\0\0! \0\0\4\0 \3\4"..., 52) = 52 read(3, "\241 \342\10\4\0 \3\4\0\0\0\4\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 32) = 32 chdir("/tmp/install") = 0 fork() = 836 wait4(836, [WIFSIGNALED(s) && WTERMSIG(s) == SIGSEGV && WCOREDUMP(s)], 0, NULL) = 836 --- SIGCHLD (Child exited) --- chdir("/tmp/sv001.tmp") = 0 ioctl(3, FIONREAD, [0]) = 0 write(3, "\31\0\v\0\4\0 \3\0\0\0\0! \0\0\4\0 \3\5\0\0\0\5\0\0\0\0"..., 44) = 44 read(3, 0xbffff050, 32) = -1 EAGAIN (Resource temporarily unavailable) select(4, [3], NULL, NULL, NULL) = 1 (in [3]) read(3, "\241 \343\10\4\0 \3\5\0\0\0\5\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 32) = 32 open("/tmp/sv001.tmp", O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY) = 6 fstat64(6, {st_mode=S_IFDIR|0700, st_size=4096, ...}) = 0 fcntl64(6, F_SETFD, FD_CLOEXEC) = 0 getdents64(6, /* 72 entries */, 4096) = 2664 unlink(".") = -1 EISDIR (Is a directory) unlink("..") = -1 EISDIR (Is a directory) unlink("setup_services.rdb") = 0 unlink("libimpreg.so") = 0 unlink("libgcc_s.so.1") = 0 unlink("libvos2gcc3.so") = 0 unlink("libsmgr.so") = 0 unlink("libstore.so.3.0.1") = 0 unlink("libmcnttype.so") = 0 unlink("libregmimetypes641li.so") = 0 unlink("draw_o.bmp") = 0 unlink("libsvl641li.so") = 0 unlink("setup") = 0 unlink("libtdmgr.so") = 0 unlink("open_office_ani.bmp") = 0 unlink("sopatchlevel.sh") = 0 unlink("calc_o.bmp") = 0 unlink("libdefreg.so") = 0 unlink("setup.bin") = 0 unlink("libreg.so.3.0.1") = 0 unlink("libsalhelper3gcc3.so") = 0 unlink("vcl64101.res") = 0 unlink("libtl641li.so") = 0 unlink("libjvm641li.so") = 0 unlink("set64101.res") = 0 unlink("libcpld.so") = 0 unlink("libsimreg.so") = 0 unlink("allgemein_o.bmp") = 0 unlink("writer_o.bmp") = 0 unlink("libsbl641li.so") = 0 unlink("ende_o.bmp") = 0 unlink("liblocaledata_ascii.so") = 0 unlink("libzip641li.so") = 0 unlink("libset641li.so") = 0 unlink("svs64101.res") = 0 unlink("libsal.so.3.0.1") = 0 unlink("libstdc++.so.3.0.4") = 0 unlink("libxcr641li.so") = 0 unlink("svt64101.res") = 0 unlink("libtplx641li.so") = 0 unlink("libsot641li.so") = 0 unlink("libj641li_g.so") = 0 unlink("libtk641li.so") = 0 unlink("impress_o.bmp") = 0 unlink("libsvt641li.so") = 0 unlink("libcppu.so.3.0.1") = 0 unlink("libi18n641li.so") = 0 unlink("libucbhelper1gcc3.so") = 0 unlink("libpsp641li.so") = 0 unlink("applicat.rdb") = 0 unlink("libcppuhelper3gcc3.so") = 0 unlink("libstlport_gcc.so") = 0 unlink("tplx64101.res") = 0 unlink("libi18npool641li.so") = 0 unlink("libcomphelp2.so") = 0 unlink("libutl641li.so") = 0 unlink("libuui641li.so") = 0 unlink("liblocaledata_CJK.so") = 0 unlink("jvm64101.res") = 0 unlink("libvcl641li.so") = 0 unlink("librdbtdp.so") = 0 unlink("open_office_org_logo.bmp") = 0 unlink("uui64101.res") = 0 unlink("libdtransX11641li.so") = 0 unlink("setup.ini") = 0 unlink("unorc") = 0 unlink("libstore.so.3") = 0 unlink("libreg.so.3") = 0 unlink("libsal.so.3") = 0 unlink("libstdc++.so.3") = 0 unlink("libcppu.so.3") = 0 unlink("core") = 0 getdents64(6, /* 0 entries */, 4096) = 0 chdir("/tmp/install") = 0 rmdir("/tmp/sv001.tmp") = 0 write(3, ".\0\2\0\6\0 \0036\0\2\0\1\0 \0036\0\2\0\2\0 \3<\0\2\0\5"..., 44) = 44 read(3, 0xbffff8b0, 32) = -1 EAGAIN (Resource temporarily unavailable) select(4, [3], NULL, NULL, NULL) = 1 (in [3]) read(3, "\1\1\351\10\0\0\0\0\22\0`\1\0\0\0\0\0\0\0\0\1\0\0\0\210"..., 32) = 32 shutdown(3, 2 /* send and receive */) = 0 close(3) = 0 write(1, "glibc version: 2.2.4\n", 21glibc version: 2.2.4 ) = 21 munmap(0x40017000, 4096) = 0 _exit(1) = ?
==================================== SOLUTION - SOLUTION - SOLUTION ==================================== This whole mistery was caused by a naughty truetype font I have in my system. By cancelling the process installation before it self-aborts and clean up, I got access to the /tmp/sv001.tmp directory with the second stage decompressed files. Then it was necessary for me to include libcomphelp2.so in the LD_LIBRARY_PATH coz setup.bin was complaining to have not found it. That was the only library complaint. Guess this is due to the fact that we don't have a on-going installation with the environments set on. If setup.bin or setup was executed, it segfaulted. By stracing setup.bin, my friend JonC concluded that when the binary looked for the fonts installed, would segfault because the font ather___.ttf was being allocated to use more than 16 MB RAM. Which fonts needs 16MB RAM? So, it would segfault because I had 15MB free from 164MB total. (Mozilla running ya know...) - This is not a memory problem, but how a font is bad, needing 16 MB to allocate. I wonder if it would try to fill more RAM if I had 60 MB free. So, deleting the font and restarting xfs services. I was able to get OpenOffice.org installation running. Based on this, if you are having a similiar installation problem, check carefully if the fonts are not hindering your install somewhat. Many thanks to JonC, evrAk, Vapour for their valuable help of which if I didn't have, I would never be able to solve this. TrueType font name: ather___.ttf You might get problems with other truetype fonts. An additional note is: A while ago, I had installed OpenOffice.org 641c, and after I installed it, I installed the TTF support in X and got them running up. Everytime I would scroll the font menu drop down to choose them, OpenOffice.org would crash. This is definitely because the same font. But as I didn't have TTF support before OO installation, it installed without problems. You can experience crashes when dropping down the font menu-drop-down when operating OpenOffice in Windows. That is because of a bad font too.
Strace output indicating bad font: munmap(0x418e6000, 39960) = 0 stat64("/opt/truetype/ather___.ttf", {st_mode=S_IFREG|0644, st_size=256012, ...}) = 0 open("/opt/truetype/ather___.ttf", O_RDONLY) = 11 fstat64(11, {st_mode=S_IFREG|0644, st_size=256012, ...}) = 0 old_mmap(NULL, 256012, PROT_READ, MAP_SHARED, 11, 0) = 0x418e6000 close(11) = 0 old_mmap(NULL, 65536, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41925000 old_mmap(NULL, 69632, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41935000 munmap(0x41925000, 65536) = 0 old_mmap(NULL, 135168, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41946000 munmap(0x41935000, 69632) = 0 old_mmap(NULL, 266240, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41967000 munmap(0x41946000, 135168) = 0 old_mmap(NULL, 528384, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x419a8000 munmap(0x41967000, 266240) = 0 old_mmap(NULL, 1052672, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41a29000 munmap(0x419a8000, 528384) = 0 old_mmap(NULL, 2101248, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41b2a000 munmap(0x41a29000, 1052672) = 0 old_mmap(NULL, 4198400, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x41d2b000 munmap(0x41b2a000, 2101248) = 0 old_mmap(NULL, 8392704, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x4212c000 munmap(0x41d2b000, 4198400) = 0 old_mmap(NULL, 16781312, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x4292d000 munmap(0x4212c000, 8392704) = 0 --- SIGSEGV (Segmentation fault) --- +++ killed by SIGSEGV +++
-reopening to change state
- changed status to WORKSFORME
*** Issue 3869 has been marked as a duplicate of this issue. ***
I opened issue 3869, which I agree is a duplicate of this issue, but I don't consider either one resolved. We do not know what set of fonts causes the crash, except that it includes at least 2Rebels-Athena-medium-r-normal. Unless we can clearly fix the blame on some external library like freetype or xlib it looks the bug is in OOo. This font works in some other software like yudit. It is used by the public and will cause problems for others trying out OOo. OOo crashing in the presence of some fonts is not acceptable behavior. A minimal WORKAROUND would be to mention in the release notes that certain fonts may need to be removed before installing or running OOo, but that is not a final resolution. Let's fix it so it never crashes.
Hello Herbert, is there any solution for this font problem?
Can you attach the font file ather__.ttf which is causing the problem? HDU->PL: PSPRINT opens this font file during startup indicates.
The font stimulating this bug can be found at http://www.jiffycomp.com/smr/unicode/athena.zip
i can reproduce the bug now.
pl->mh: please update psprint/source/fontsubset/gsub.cxx to rev 1.3 (toplevel)
done, now in OOO_STABLE_1 branch.
Just a data point: it happened to me, but the culprit was NOT a ttfont; it was a Type1. The problem was that the RH update left an old fonts.dir / fonts.alias file that, with the new fonts, was bogus. I simply did mv fonts.dir.rpmnew fonts.dir mv fonts.scale.rpm fonts.scale mkfontdir . and the problem disappeared...
verified.
close issue.