Lucky I saw this (I'm probably not going to be at Sun in July)...
>>>> "MB" == Martin Buchholz
<martin(a)xemacs.org> writes:
>>>> "ajc" == Andrew J Cosgriff
<ajc(a)bing.wattle.id.au> writes:
>> and it just hangs. nothing's been written to xemacs itself.
>>
>> Has anybody else been seeing this sort of thing on Solaris 2.5.1
>> ?
ajc> just an extra data point - it seems to be hanging on line 323 of
ajc> dynodump.c -
ajc> if ((oehdr = elf_newehdr(oelf)) == NULL)
ajc> (gdb) print oelf $1 = (Elf *) 0x545a00
MB> Perhaps dynodump is broken on Ultra 30s?
Doubt it. elf_newehdr() is in libelf which is common for all SPARC
arches...
It's more likely that you're doing something silly with malloc
(elf_newehdr() calls it)...
MB> dynodump is only used on Solaris 2.5.1, not 2.6.
Correct. In 2.6 and beyond, the functionality of dynodump is subsumed
by the dldump(3X) call...
MB> XEmacs successfully builds for me on Solaris 2.5.1 on a
MB> pre-Ultrasparc, and on 2.6 on an Ultra 30.
I, too, have an Ultra 30 here, albeit running 2.7...