Re: [3.0 branch] libjava testsuite, i686-pc-linux-gnu --disable-shared

[Alexandre Petit-Bianco <apbianco at cygnus dot com>]

Andrew Haley writes:

> 2.  Making the dejagnu timeout much shorter.
> 
> More news as soon as we have it.

I don't have yet a clean solution reducing the dejagnu timeout value
when testing java executable, but if a quick and dirty temporary fix
is OK with you (that is, provided that it's convenient and/or
customary for you to isolate java testing from the rest of the
toolchain testing) you can search your sources of dejagnu (I'm
guessing they would be extracted from a sourceware gdb snapshot or
checked out tree) for the file dejagnu/config/unix.exp. In the proc
unix_load you'll find the following line:

   set status [remote_wait $dest 300];

300 is the hardcoded timeout value used in testing produced
executable. Change it to a more appropriate value -- I'm guessing 3 to
5 (seconds) for java executable would be OK on most platforms. Of
course this value most likely won't be suitable when testing other
frontends/runtimes. Here's a patch extracted from a current copy of
the gdb sources.

I'm looking into a more acceptable solution. Meanwhile, I hope this
helps.

./A

Index: dejagnu/config/unix.exp
===================================================================
RCS file: /cvs/src/src/dejagnu/config/unix.exp,v
retrieving revision 1.1.1.1
diff -u -p -r1.1.1.1 unix.exp
--- unix.exp    1999/11/09 01:28:42     1.1.1.1
+++ unix.exp    2001/05/02 19:50:45
@@ -73,7 +73,7 @@ proc unix_load { dest prog args } {
            set output "remote_spawn failed"
            set status -1;
        } else {
-           set status [remote_wait $dest 300];
+           set status [remote_wait $dest 5];
            set output [lindex $status 1];
            set status [lindex $status 0];
        }