?
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
On Sep 16, 2005, at 4:31 PM, David Landgren wrote:
This Week on perl5-porters (5-11 September 2005)
The Return of the perl5-porters Summaries
Hooray! Thanks for doing!
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
I'm almost done with all my NN/NULLOK excitement in embed.fnc. Here's a
patch of accumulated goodness. Nothing exciting.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/embed.fnc bw/trunk/embed.fnc
--- bleadperl/embed.fnc 2005-09-01 10:14
don't see a patch, though.
How did you make sure that it works *exactly* like the original? Did you
write tests for it? Can we turn those tests in .t files?
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
- rename $makefile_rename, $makefile
+ rename $makefile_rename, $makefile . $makefile_ext
or die Can't rename '$makefile_rename' to '$makefile': $!;
Shouldn't that be
or die Can't rename '$makefile_rename' to '$makefile$makefile_ext': $!
xoa
--
Andy Lester = [EMAIL PROTECTED
of this change? I need to make it in Test::Harness
directly, but would like to know why.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Not a lot of excitement here. More NN/NULLOK in embed.fnc. Some
consting in op.c.
oxox,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/embed.fnc bw/trunk/embed.fnc
--- bleadperl/embed.fnc 2005-07-29 03:59:52.0 -0500
+++ bw/trunk/embed.fnc
/errors/affect.html
It is perfectly fine, although less common, to say to effect a sleep of
250 milliseconds.
I also agree that it would be preferable to s/effect/cause/ in the
example above.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
. Added a couple of consts.
* pp_ctl.c: Consting.
* regcomp.c: Consting.
* regexec.c: Consting.
* sv.c: Fixed indenting.
* util.c: Consting.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/doio.c bw/trunk/doio.c
--- bleadperl/doio.c2005
In Simon's Advanced Perl Programming, 2e there's a quote from
overload.pm that warrants a [sic]. This fixed that.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/lib/overload.pm bw/trunk/lib/overload.pm
--- bleadperl/lib/overload.pm 2004-11
instead of
Xoperators.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
websites.
Yes, great. Please start.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
?
How about we phrase it Do not use capitals unless it would be incorrect
to not have them.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Yes, this is correct. The loop variable is an alias.
will squawk at
compile if a null is passed in, because the code refers to *args, which
would be a NULL deref if args is NULL. However, util.c has code that
passes in hardcoded NULLs. :-(
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/doio.c bw/trunk
derefs are
protected by an if(args).
But not svargs. See the line with sv_catsv(sv, *svargs);
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
that combination?
It'd be nice to have the compiler watch these things.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
. Removed an #if 0 section for static
functions that are hard-coded in toke.c.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/doio.c bw/trunk/doio.c
--- bleadperl/doio.c2005-07-18 05:22:23.0 -0500
+++ bw/trunk/doio.c 2005-07-21
On Jul 20, 2005, at 6:21 AM, Stas Bekman wrote:
someone has broken croak(Nullch) in blead:
That would be me.
It's a trivial change in embed.fnc, if anyone wants to get to it today.
I can't.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
On Mon, Jul 18, 2005 at 11:19:45AM +0100, Nicholas Clark ([EMAIL PROTECTED])
wrote:
If this is true then we can re-order functions in embed.fnc to our heart's
content in maint.
Good, 'cause I already have a couple times. :-)
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
repeated calls
to GvHVn, a complicated macro. Saves about 100 bytes of object code.
* regcomp.c: One consting.
* utf8.c: One consting.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
in pp_chdir, and removed repeated calls
to GvHVn, a complicated macro. Saves about 100 bytes of object code.
* regcomp.c: One consting.
* utf8.c: One consting.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/dump.c bw/trunk/dump.c
--- bleadperl/dump.c2005-07-08 12:03:01.0 -0500
+++ bw/trunk/dump.c 2005-07-14 15:11:45.0 -0500
@@ -154,7 +154,7 @@ char *
Perl_sv_peek(pTHX_ SV *sv)
{
dVAR
()
* op.c: Consting and PERL_UNUSED_ARG().
* pad.c: Consting.
* perlio.c: Lots of consting.
* pp_ctl.c: Consting.
* pp_hot.c: Consting. Removed an unnecessary setting of a null pointer.
* toke.c: Consting.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur
: Consting.
* util.c: Consting and localizing.
I do realize that there are some warnings on get_av() that pop up.
Those need to be fixed, but by something other than allowing get_av() to
be ignored.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl
Here's a patch that brings the hi-res timings now in Test::Harness to
make test. As with T::H, you'll have to set HARNESS_TIMER=1 to get 'em.
Without that, you'll see no difference.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/t/TEST bw
On Wed, Jul 06, 2005 at 09:22:32AM -0700, Michael G Schwern ([EMAIL PROTECTED])
wrote:
* Simplify the what's case-insensitive docs. Its just all non-Unix.
What about Mac OS X running HFS?
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
just figured 1.5 million compares eliminated would buy us something.
:-/
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
and to make the connection even more obvious you can refer to $type
in
the docs.
If the $type contains one of the substrings...
Or even better,
If I$type contains one of...
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
http://www.ccl4.org/~nick/P/25060-g-gcov/
It's am-I-hot-or-not meets perlapi.pod. And about as exciting as the
harmonic
mean.
Oh, I'm excited by it! Stuff to pick at! Look at all those lines of
to eradicate!
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
have to
constify the pointer, since it's not really a pointer.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: hv.c
===
--- hv.c(revision 725)
+++ hv.c(working copy)
@@ -33,7
, even though we check it 1,560,856 times? That's because
the function is only called in two places, and in both of those, the
parent has already guaranteed we've not passing a null pointer.
The attached patch cleans that up.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
On Jul 4, 2005, at 3:03 PM, Nicholas Clark wrote:
On Mon, Jul 04, 2005 at 02:12:00PM -0500, Andy Lester wrote:
The attached patch cleans that up.
Thanks, applied (25079)
now, can someone with better measurement-fu tell us if it sped anything
up?
xoxo,
Andy
--
Andy Lester = [EMAIL
and localizing.
Next on my list: Making a version of SvREFCNT_inc and _dec that don't
check their parm for NULLness.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
.
* toke.c: Consting and localizing.
Next on my list: Making a version of SvREFCNT_inc and _dec that don't
check their parm for NULLness.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/deb.c bw/trunk/deb.c
--- bleadperl/deb.c 2005-05-11
On Mon, Jun 27, 2005 at 09:21:18AM +0100, Steve Hay ([EMAIL PROTECTED]) wrote:
Andy Lester wrote:
I'm not seeing what you're referring to. Our numbers may be out of
sync.
Perhaps you need to resync?
Oh whoops, yes, I need to send patch.
--
Andy Lester = [EMAIL PROTECTED
On Mon, Jun 27, 2005 at 12:52:14PM -0500, Andy Lester ([EMAIL PROTECTED]) wrote:
Perhaps you need to resync?
Oh whoops, yes, I need to send patch.
* embed.fnc: New NN/NULLOK. Downgraded the parms on sv_len and
sv_len_utf8 to NULLOK.
* gv.c: New consting. Converting to PERL_UNUSED_ARG
On Mon, Jun 27, 2005 at 09:25:02AM +0100, Steve Hay ([EMAIL PROTECTED]) wrote:
Aww - I was beginning to like the timer output now the Abe graciously
fixed Test-Smoke to understand it.
Set HARNESS_TIMER in your environment.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
I've just uploaded Test::Harness 2.51_02. It turns off the timer by
default, and adds a --timer switch to prove. Please try it out and see
if all is well because I'm going to make it 2.52 tomorrow.
And now, I must go to bed so I can drive to Toronto...
xoxo,
Andy
--
Andy Lester = [EMAIL
On Jun 24, 2005, at 2:48 AM, Steve Hay wrote:
24953 added some initializations to silence bcc32 compiler warnings
24963 corrected the initialization (sorry!)
24965 removed the initializations!
Do the PERL_UNUSED_VAR() calls not make your BCC happy?
xoa
--
Andy Lester = [EMAIL PROTECTED
this output with some environment variable. Andy, advice ?
I'd rather that smoke use Test::Harness directly, rather than trying to
parse output. THAT would be the best solution, I think.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
#define PERL_UNUSED_VAR(x) (x)=1
# else
#define PERL_UNUSED_VAR(x) ((void)x)
# endif
#endif
Does that make BCC happy?
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
there was no setting of newsp at all.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
*|newXS |NN const char* name|XSUBADDR_t f|NN const
char* filename
+Apd|CV*|newXS |NULLOK const char* name|XSUBADDR_t f|NN const
char* filename
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
may be out of
sync.
At the tail end of pp_last, there are two PERL_UNUSED_VAR(), but one is
optype and the other is gimme.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
patches before submitting them.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
: Switched to PERL_UNUSED_ARG() rather than
unnecessarily setting a var, which doesn't get us past lint anyway,
since lint will say set but not used.
* pp_pack.c, scope.c: PERL_UNUSED_ARG()
* taint.c: consting
* util.c: Consting.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
warning.
universal.c: Lots of localizing. Switching to PERL_UNUSED_ARG() rather
than naked (void)x.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/doop.c bw/trunk/doop.c
--- bleadperl/doop.c2005-06-20 06:06:18.0 -0500
+++ bw/trunk
can't remove it from the outer scope
due to other scratch uses?
Yes, because it makes explicitly clear that the outer and inner ones are
unrelated.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
these.
I don't think that many people will be able to do that - especially if
I don't imagine that there's anyone on p5p who's savvy enough to patch
the Perl core but doesn't know how to make changes to a read-only file
in his editor.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
: Localizing and consting.
scope.c: Localizing and consting.
This patch will need to have make regen run.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -pur bleadperl/embed.fnc bw/trunk/embed.fnc
--- bleadperl/embed.fnc 2005-06-15 04:33:08.0 -0500
+++ bw/trunk
?
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
I'm betting that nobody is using the lint that all the /*SUPPRESS xxx*/
messages are there for, so here's a patch to remove them.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: av.c
Here's a patch for quieting some lint warnings on Solaris lint. Also
introduces a little bit of infrastructure.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/perl.h /home/alester/bw/trunk/perl.h
--- /home/alester/bleadperl
On Thu, Jun 16, 2005 at 09:20:05AM -0700, Michael G Schwern ([EMAIL PROTECTED])
wrote:
+ * ex: set ts=8 sts=4 sw=4 noet:
What are these lines for?
Looks like VIM formatting codes. *HI!!*
Yes, and they've been added in regularly for at least the past month.
xoa
--
Andy Lester
, Gray Fool, the quote from ill-fated steward
of Gondor Denethor went.
Inexplicably? Did this guy notice it's in the debugging code?
I've always found Larry's quotes to be marvelously appropriate, and I
love reading them every time.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
to be putting editor specific jiggery-pokery formatting commands
into our source files?
Clearly we do, just like we're fine with putting lint-specific
directives in there as well.
I'm not seeing a problem here.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Here's a version of the CopSTASH_eq patch that puts the guts in a
function, but only for the threads.
Needs a make regen afterwards.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: util.c
I've moved the CopSTASH_eq macro into a function. This quiets many
compiler warnings, and reduces a lot of redundant rework.
I also consolidated some groups of functions in embed.fnc.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester
debug-only code and moved it to the
bottom.
* sv.c: Consting.
* util.c: Passing in string lengths to sv_catpvn() instead of calling
sv_catpv().
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/embed.fnc /home/alester/bw/trunk
Here's some consting and localizing and reindenting of pp_pack.c
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: pp_pack.c
===
--- pp_pack.c (revision 522)
+++ pp_pack.c (working copy)
@@ -2319,7
This patch refactors some repeated code into macros, and throws in an
assertion to make sure that the UTF8 classes have been loaded succesfully,
all in the name of cleaning up a whole bunch of lint and GCC warnings.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
I am in danger of pointing to obvious, but am I right that in
particular
You mean it should have been
if (*x == *y++)
break;
x++;
Yes yes yes in the patch y is irrelevant.
Let's look at the patch before picking it apart.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com
: Simplified Perl_savepv() to not use conditional compilation,
which apparently existed only to save an autovar that the compiler will
get rid of anyway.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/embed.fnc /home/alester/bw/trunk
SvUPGRADE never returns anything meaningful. We (void) it everywhere.
Now it's (void)ed inside the macro itself.
Later, we can worry about making sv_upgrade void instead of returning an
unused boolean.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index
which
undoubtedly gets compiled away anyway.
* Changed sv_catpv( sv, literal ) to sv_catpvn( sv, literal, 7 )
throughout.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/util.c /home/alester/bw/trunk/util.c
--- /home/alester/bleadperl
Some recent benchmarks indicated that removing all register keywords
from
the perl core made things slower.
Fine, then let's leave 'em in apply the rest.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
On Sun, Jun 05, 2005 at 08:32:17PM +0100, Dave Mitchell ([EMAIL PROTECTED])
wrote:
Some recent benchmarks indicated that removing all register keywords from
the perl core made things slower.
So here's the patch without the removal of register.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED
Minor call fix for 24706
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: utf8.c
===
--- utf8.c (revision 496)
+++ utf8.c (working copy)
@@ -246,7 +246,7 @@
if (len = 4)
return
The UNI* macros at the top of toke.c make lint unhappy, specifically
about about undefined order of operations in the big chain of
comma-separated expressions.
This patch turns them into a block that doesn't have any such problems.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED
.
* util.c: Quieted an unused var warning.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/doio.c /home/alester/bw/trunk/doio.c
--- /home/alester/bleadperl/doio.c 2005-06-02 03:07:41.0 -0500
+++ /home/alester/bw/trunk/doio.c 2005-06
Here's some more quote cleaning in pod/*.pod.
I also removed Velocigen from the FAQ since it no longer exists, at
least that I can find on the web.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/pod/perlfaq3.pod
/home/alester/bw/trunk
Here are some cases where we know the length of the SvPV getting passed
in to sv_setpv(), so let's call sv_setpvn() instead.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: gv.c
===
--- gv.c
Here's quote fixing on *.c and *.h.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: av.c
===
--- av.c(revision 461)
+++ av.c(working copy)
@@ -383,7 +383,7 @@
av = (AV
10.
My processor is very relieved.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/lib/unicore/mktables
/home/alester/bw/trunk/lib/unicore/mktables
--- /home/alester/bleadperl/lib/unicore/mktables2005-04-25
07:09
I'm making my way through the ext/ directory.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: ext/Devel/Peek/Peek.xs
===
--- ext/Devel/Peek/Peek.xs (revision 455)
+++ ext/Devel/Peek
* Squashed two warnings
* Removed an unnecessary SvCUR_set() call
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: ext/ByteLoader/ByteLoader.xs
===
--- ext/ByteLoader/ByteLoader.xs(revision 443
I'm moving on to modules in my quest for const correctness. Here's
Opcode.xs.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: ext/Opcode/Opcode.xs
===
--- ext/Opcode/Opcode.xs
I've explained at least three times in IRC why consting is a good thing.
I figure I should write up something explaining it. Porting/pumpkin.pod
seems like the best choice, but it's not exactly a fit.
Suggestions?
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
A little patch for ext/attrs/attrs.xs to be consted.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: ext/attrs/attrs.xs
===
--- ext/attrs/attrs.xs (revision 440)
+++ ext/attrs/attrs.xs (working copy
This is a little more than a standard consting.
* Consted all input string parms
* Localized a couple of temporary buffers
* Changed some sv_catpv( xx ) to sv_catpvn( xx, 2 ), since we always
know the length.
* Called SvPV_len() where we were throwing away the length.
xoxo,
Andy
--
Andy
Too late for -C option at xmltv-0.5.39/grab/jp/tv_grab_jp line 1.
$
Even syntax check is impossible. :(
Syntax check is -c, not -C.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Bigger two:
(1) Surrogates such as 0xD800 and noncharacters such as 0x10
are subjected to test.
- It requires no warnings 'utf8' in addition to -w or use warnings
Either that or we use Test::Warnings to make sure that the warnings are
being emitted as expected.
--
Andy Lester
I finally ran blead on Tiger under gcc 4.0 and a bunch of different
errors than under 4.1.
Here are some patches to clean 'em up.
xoxo,
Andy
cleanup.patch
Description: Binary data
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
tweaks and patches to come as I find the nits.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: regcomp.h
===
--- regcomp.h (revision 401)
+++ regcomp.h (working copy)
@@ -154,15 +154,9
in util.c expects to see them. And in turn,
I would have thought that that could be cured by making it call a function
in op.c
I'm perfectly fine with that. I see no reason for them to be Perl_
whatever.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
This patch lets us exit out of S_hv_magic_check once we've set all the
flags that there are to set, rather than continuing down the linked list
of magic.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/hv.c /home/alester/bw/trunk/hv.c
Safefree() doesn't care what type its pointer parm is because it casts
it for us anyway. Remove these unused casts that make lint complain.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/scope.c /home/alester/bw/trunk/scope.c
--- /home
to remove the return value in the first call and
localize the U8 *e down below.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/embed.fnc /home/alester/bw/trunk/embed.fnc
--- /home/alester/bleadperl/embed.fnc 2005-05-24 09:18
.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: perlio.h
===
--- perlio.h(revision 377)
+++ perlio.h(working copy)
@@ -210,7 +210,11 @@
START_EXTERN_C
#ifndef __attribute__format__
Attached is a patch to quiet some universal.c warnings. I also rewrote
XS_universal_noop to remove the extraneous code it goes through just to
do a Perl_croak.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Index: universal.c
On Mon, May 23, 2005 at 10:33:03AM -0700, Yitzchak Scott-Thoennes ([EMAIL
PROTECTED]) wrote:
Does the main attribute part of perl.h need to move more to the top?
It might.
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
Is it worthwhile to try to slim down this struct?
On the contrary, I wonder if it would be a speedup to fatten it up by
giving each bit in the flags its own field.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
If ctags doesn't have the --langmap parm, it won't handle .h as C.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/Makefile.SH /home/alester/bw/trunk/Makefile.SH
--- /home/alester/bleadperl/Makefile.SH 2005-05-20 08:38
will need to be rerun after this.
xoxo,
Andy
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
diff -ur /home/alester/bleadperl/embed.fnc /home/alester/bw/trunk/embed.fnc
--- /home/alester/bleadperl/embed.fnc 2005-05-16 12:23:31.0 -0500
+++ /home/alester/bw/trunk/embed.fnc
to call,
NORETURN_FUNCTION_END that behaves appropriately to not cause warnings,
depending on compiler capabilities.
* Includes Jarkko's guessing at attributes based on GCC version.
* Removed a definition for NextStep 3.0 to force attributes off.
I await your feedback.
xoa
--
Andy Lester = [EMAIL
into my
code and then re-send the patch, including Porting/Glossary.
I'm pulling out HASATTRIBUTE in favor of the more specific
HASATTRIBUTE_xxx.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
On Wed, May 18, 2005 at 12:07:36PM -0500, Andy Lester ([EMAIL PROTECTED]) wrote:
Let's do this. I'll take Jarkko's patches and incorporate them into my
code and then re-send the patch, including Porting/Glossary.
I'm pulling out HASATTRIBUTE in favor of the more specific
HASATTRIBUTE_xxx
.
I'll include that in my megapatch.
xoa
--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance
1 - 100 of 205 matches
Mail list logo