code review 1792042: tests/fixedbugs/bug243.go: instead of closing stdout,

tests/fixedbugs/bug243.go: instead of closing stdout,
remove the print statements.

This change is because of the port of gccgo to RTEMS.
These tests use the GCC DejaGNU framework. In some cases,
the tests need to be run on qemu where the status code
cannot be sent back to DejaGNU, so it prints the exit status
by putting a wrapper around the exit and abort calls.

This testcase closes the stdout, and hence prohibits DejaGNU
from knowing the status in such cases, and causes this test
to be wrongly declared as a failure.

[Raj_, 2010-07-12 12:06:11]

Hello rsc, iant (cc: golang-dev@googlegroups.com),

I'd like you to review this change.

[rsc, 2010-07-12 20:16:55]

LGTM

that was weird code

[rsc, 2010-07-12 20:17:37]

*** Submitted as http://code.google.com/p/go/source/detail?r=97f6a942d613 ***

test/fixedbugs/bug243.go: instead of closing stdout,
remove the print statements.

This change is because of the port of gccgo to RTEMS.
These tests use the GCC DejaGNU framework. In some cases,
the tests need to be run on qemu where the status code
cannot be sent back to DejaGNU, so it prints the exit status
by putting a wrapper around the exit and abort calls.

This testcase closes the stdout, and hence prohibits DejaGNU
from knowing the status in such cases, and causes this test
to be wrongly declared as a failure.

R=rsc, iant
CC=golang-dev
http://codereview.appspot.com/1792042

Committer: Russ Cox <rsc@golang.org>

[Raj_, 2010-07-12 21:25:33]

On Tue, Jul 13, 2010 at 1:46 AM, Russ Cox <rsc@golang.org> wrote:

> LGTM
>
> that was weird code
>

Sorry, the code I submitted had compiler errors. Didn't do a check
in haste.

And now I understand why you kept in the printlns, it was to prevent
the compiler from showing an error that the variable conn is unused.

Will submit a new CL.

Again, sorry.