On 29.01.2007 [13:48:58 -0600], Adam Litke wrote:
> On Mon, 2007-01-29 at 11:31 -0800, Nishanth Aravamudan wrote:
> > It was noted that in some benchmarks (SPEC, at least), it is necessary
> > for a successful run to completely silence the library's output. That
> > includes no ERROR messages. Add an env var for the user to request this,
> > HUGETLB_QUIET, which if set at all will cause the library to produce no
> > output.
> 
> Can't they just set HUGETLB_VERBOSE=0 with the code as is?

Sigh, brain fart -- this should work just as well, I'm checking into it.

> Also, what ERROR() conditions can be triggered that would still result
> in a passing testcase?

It's a weird situation, where the benchmark saves both stderr and stdout
during a period when no hugepages are allocated (to leave as much free
memory on the system as possible). The lib complains but it's a
don't-care as far as the compile is concerned, apparently.

Thanks,
Nish

-- 
Nishanth Aravamudan <[EMAIL PROTECTED]>
IBM Linux Technology Center

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Libhugetlbfs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libhugetlbfs-devel

Reply via email to