New submission from Дилян Палаузов:

Providing that during ./configure Py_DEBUG can either be set or not, but there 
is no third state, which third state would justify calling ./configure 
--with-pydebug=lambda, --with-pydebug shall be renamed to --enable-pydebug .

Likewise for --with-pymalloc , but it is more tricky.  Currently ./configure 
--help emits "--with(out)-pymalloc    disable/enable specialized mallocs".  
Usually from the output of ./configure --help one shall be able to conclude, 
whether a switch is on or off by default, if it is not provided.  E.g. if it is 
by default off, then "--enable-X   enable feature X" is printed; and 
"--disable-Y disable feature-Y" is printed when by default the feature is on.

The code says, that if pymalloc is enabled implicitly, if neither 
--with-pymalloc nor --without-pymalloc is provided.  Let's update the 
AS_HELP_STRING to show the default and also shorten the code, which makes it 
also easier to understand (for me at least)
which makes it also easier to understand (for me at least):

diff --git a/configure.ac b/configure.ac
--- a/configure.ac
+++ b/configure.ac
@@ -3290,12 +3290,7 @@ AC_MSG_RESULT($with_doc_strings)
 # Check for Python-specific malloc support
 AC_MSG_CHECKING(for --with-pymalloc)
 AC_ARG_WITH(pymalloc,
-            AS_HELP_STRING([--with(out)-pymalloc], [disable/enable specialized 
mallocs]))
-
-if test -z "$with_pymalloc"
-then
-    with_pymalloc="yes"
-fi
+            AS_HELP_STRING([--without-pymalloc], [disable specialized 
mallocs]),,with_pymalloc=yes)
 if test "$with_pymalloc" != "no"
 then
     AC_DEFINE(WITH_PYMALLOC, 1,

configure.ac:4629  mentions "# * --with-wide-unicode (adds a 'u')", but since 
commit d63a3b8beb4a0841cb59fb3515347ccaab34b733 "Implement PEP 393", one cannot 
pass --with-wide-unicode to ./configure.

----------
components: Build
messages: 293182
nosy: dilyan.palauzov
priority: normal
severity: normal
status: open
title: ./configure, pydebug and pymalloc

_______________________________________
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/issue30294>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

Reply via email to