Sorry for answering myself, but I only realized this is the same kind of
bug as the previous one, but more convoluted.
the "replace" script does the right only if *both* the
SunOS5/i386-5.10/i86 and SunOS5/i386-5.10/amd64 directories are present.
If not, it assumes it can replace the binaries directly. This is
obviously an insufficient check.
The "buildsunos" script is the one supposed to create those directories
in the first place.
Again, the check is insufficient. If that one fails, it assumes 32 bit
only, and won't create architecture subdirectories:
if [ $solrev -ge 10 ] && /bin/optisa sparcv8plus amd64 > /dev/null
However, I'm running S10 on a 32 bit processor, so that check will fail
(no amd64), but the i86/ and amd64/ subdirectories are *still* both needed.
I think the check should only be on the version of Solaris and the
ability of the compiler to generate both i86 and amd64 binaries. And in
any case, even if amd64 binaries are not possible, it should always use
an i86/ subdirectory, not . , and the "replace" script should, too.
Of course, this issue does not exist on SPARC, since there's no 32 bit
SPARC supported by S10.
Laurent
Laurent Blume a écrit :
> I've tried replacing S10U4 IPF by Darren's (since I'm badly hit by the
> NAT issue). I used the SunOS5/replace script as advised in a previous post.
>
> Right afterwards, I got an interesting problem:
> # w
> no rules loaded
>
> A whole bunch of files were not what they used to be:
> [...]
> -r-xr-xr-x 73 root bin 2025368 sept. 25 18:16 nohup
> -r-xr-xr-x 73 root bin 2025368 sept. 25 18:16 newtask
> -r-xr-xr-x 73 root bin 2025368 sept. 25 18:16 mdb
> -r-xr-xr-x 73 root bin 2025368 sept. 25 18:16 gcore
> -r-xr-xr-x 73 root bin 2025368 sept. 25 18:16 cputrack
> -r-xr-xr-x 73 root bin 2025368 sept. 25 18:16 adb
>
>
> All those files are actually stubs, that run isaexec(3C) to choose
> between i86/<file> or amd64/<file>.
>
> # pkgchk -lp /usr/sbin/ipf
> Pathname: /usr/sbin/ipf
> Type: linked file
> Source of link: ../../usr/lib/isaexec
>
> The "replace" script use cp to overwrite that stub, and since it's also
> used by many other files, it has a very nasty side-effect of replacing
> many things by IP Filter:
>
> $ SunOS5/replace
> i386-5.10
> UPDATE /usr/sbin ipf:ipf.exe
> /bin/cp i386-5.10//ipf.exe /usr/sbin//ipf
> UPDATE /usr/sbin ipfs
> /bin/cp i386-5.10//ipfs /usr/sbin//ipfs
> UPDATE /usr/sbin ipfstat
> /bin/cp i386-5.10//ipfstat /usr/sbin//ipfstat
> UPDATE /usr/sbin ipmon
> /bin/cp i386-5.10//ipmon /usr/sbin//ipmon
> UPDATE /usr/sbin ipnat
> /bin/cp i386-5.10//ipnat /usr/sbin//ipnat
> UPDATE /usr/sbin ippool
> /bin/cp i386-5.10//ippool /usr/sbin//ippool
>
>
> The targets of that script should be the real executables, ie:
> -r-xr-xr-x 1 root bin 57288 août 14 19:24 /usr/sbin/amd64/ipmon
> -r-xr-xr-x 1 root bin 45048 août 14 19:24 /usr/sbin/i86/ipmon
>
> Luckily, it's not difficult to switch back to the stub if it's still
> available:
>
> # cp -p ~laurent/isaexec /usr/sbin/ipf
>
> # w
> 6:44pm actif 2:26, 3 utilisateurs, charge moyenne : 0,05, 0,05, 0,09
> User tty login@ idle JCPU PCPU what
> laurent pts/1 6:01pm 1 w
> laurent pts/2 4:30pm 1:46 8 8 -bash
> laurent pts/3 6:22pm 6 19 -bash
> laurent pts/4 6:38pm 5 -bash
>
> Laurent
--
/ Leader de Projet & Communauté | I'm working, but not speaking for
\ G11N http://fr.opensolaris.org | Bull Services http://www.bull.com
/ FOSUG http://guses.org |
