develooper Front page | perl.perl5.porters | Postings from March 2003

Re: [PATCH] sig_num, etc.

Thread Previous | Thread Next
Craig A. Berry
March 13, 2003 23:08
Re: [PATCH] sig_num, etc.
Message ID:
At 9:44 PM -0800 3/13/03, Gurusamy Sarathy wrote:
>On Thu, 13 Mar 2003 21:38:19 CST, "Craig A. Berry" wrote:
>>The attached patch corrects three minor issues for VMS in
>>It's relevant (and should apply cleanly to) both 5.8.1-to-be and bleadperl.
>>-- fixes $Config{sig_num} to be space-separated and not zero-terminated
>>so we don't fail lib/Config.t. Does the same for $Config{sig_name}.
>That's not right according to the documentation in config_h.SH and
>Porting/Glossary.  I've fixed Config.t instead, since the same issue
>shows up on windows.  If the test passed on UNIXes, it sounds like
>Configure is not doing the right thing here?

You may be right, but after an hour or two I gave up trying to figure
out what the comments in config_h.SH and the docs in Porting/Glossary
were trying to tell me and went with what I thought Configure was
actually doing.  One of the confusing things is that the SIG_NUM
macro in config_h.SH derives from the sig_num_init variable in, not from sig_num.  So when the comments in config_h.SH say
that SIG_NUM should be comma-separated, that means
$Config{sig_num_init} should be comma-separated, but says nothing at
all about $Config{sig_num}.  $Config{sig_num} and
$Config{sig_num_init} are apparently two different views of the same
data, the former to be viewed from Perl, and the latter for
constructing the SIG_NUM macro in config.h for C and XS.

Whoever's been working on lib/Config.t will likely have an opinion on
this.  I'm happy to do the right thing if only I knew what it was.
Craig A. Berry

"... getting out of a sonnet is much more
 difficult than getting in."
                 Brad Leithauser

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About