NetBSD 10.0 - configure test process hang

Riccardo Mottola <[email protected]>
Newsgroups gmane.os.netbsd.ports.sparc
Message-ID <[email protected]>
Hi!

I have swapped CPU's between my SS10 with 9.4 and SS20 with 10.0 
(HyperSPARC vs SuperSPARC) and testing a couple of compilations to see.

On the SS20 (now with 2x SuperSPARC) I started compiling basic packages 
I need for 10.0 - it is just a little more stable than with HyperSPARC 
so things progress more.
Now I got something never happened to me - process hang:

Configure of bash hangs this way:
checking whether pthread_rwlock_rdlock prefers a writer to a reader...

hangs forever, I left it overnight, but nothing. no CPU usage and no 
processes in top.

I suppose it is this:

11361 root      85    0    24M 1336K nanosl/0    0:02  0.00%  0.00% conftest

deadlock? starved?

I see no errors in dmesg and serial console I left attached show snothing.

legolas: {31} ps -ax | grep conf
  7329 pts/1 S+   0:00.07 grep conf
10191 pts/2 I+   0:04.19 /bin/sh ./configure LOCAL_LDFLAGS=-Wl,-E 
--enable-alias --enable-array-variables --enable-
11361 pts/2 Sl+  0:02.49 ./conftest
14133 pts/2 I+   0:00.02 /bin/sh ./configure LOCAL_LDFLAGS=-Wl,-E 
--enable-alias --enable-array-variables --enable-

I wonder why there are two configures? I understand parallelism of 2, 
but maybe parallel tests not configures.

Anyway, I try to attach to it with gdb:
  gdb --pid=11361

but this hangs, nothing happens! I was able to "ctrl-c" configure but 
gdb then still hangs :)

I relaunched make... and it passed through, so it is a "temporary" hang, 
which is of course worse because hard to debug. Bash then completed 
compilation & installation.
Continuing with other stuff to "stress" more

Riccardo
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.