Re: Lock module improvement

2008-08-18 Thread Yoann Vandoorselaere
Hi Bruno, Le lundi 18 août 2008 à 12:55 +0200, Bruno Haible a écrit : > Another fix: Here is another patch on top of current GnuLib HEAD: - Make use of the thread module in lock-test - Include missing cond unit-test. - Isolate thread specific CPP flags into THREADCPPFLAGS (this is useful for a li

Re: Lock module improvement

2008-08-18 Thread Bruno Haible
Another fix: 2008-08-18 Bruno Haible <[EMAIL PROTECTED]> * lib/glthread/lock.h [USE_SOLARIS_THREADS]: Fix glthread_recursive_lock_* macros. * lib/glthread/lock.c (glthread_recursive_lock_destroy_multithreaded): Fix syntax error. *** lib/glthread/lock.h.orig2

Re: Lock module improvement

2008-08-17 Thread Bruno Haible
Another addendum: Since the macros now use abort() on all platforms, must be included always. 2008-08-17 Bruno Haible <[EMAIL PROTECTED]> * lib/glthread/lock.h: Include always. * lib/glthread/tls.h: Likewise. * lib/glthread/cond.h: Likewise. --- lib/glthread/lock.h.or

Re: Lock module improvement

2008-08-14 Thread Bruno Haible
Addendum: For the gettext macros, it's better to distinguish macros and functions, i.e. not to have symbols which are macros on one platform and functions on others. I'm applying this: 2008-08-14 Bruno Haible <[EMAIL PROTECTED]> * lib/glthread/lock.h (glthread_lock_lock, glthread_lock_

[PATCH] Re: Lock module improvement

2008-08-07 Thread Yoann Vandoorselaere
Le mardi 22 juillet 2008 à 01:44 +0200, Bruno Haible a écrit : > > > > - Implement support for condition variable. > > > > > > You are welcome to contribute a module for this. > > > > I started working on condition variable support. To me, it make sense to > > do that directly within the lock mod

Re: Lock module improvement

2008-08-04 Thread Yoann Vandoorselaere
Le dimanche 03 août 2008 à 20:46 +0200, Bruno Haible a écrit : > Yoann Vandoorselaere wrote: > > > Declaration: gl_lock_define(extern, name) > > > Initializer: gl_lock_define_initialized(, name) > > > Initialization: gl_lock_init (name); > > > Taking the loc

Re: Lock module improvement

2008-08-02 Thread Yoann Vandoorselaere
Le samedi 02 août 2008 à 12:15 +0200, Bruno Haible a écrit : > Yoann Vandoorselaere wrote: > > > Can you propose a reasonable compromise? > > > > Move the macro code into inline function, use the macro to call the > > appropriate function. The function return the error, the macro abort() > > in ca

Re: Lock module improvement

2008-08-02 Thread Bruno Haible
Yoann Vandoorselaere wrote: > > Can you propose a reasonable compromise? > > Move the macro code into inline function, use the macro to call the > appropriate function. The function return the error, the macro abort() > in case an error is returned. OK, sounds acceptable. What naming convention

Re: Lock module improvement

2008-08-01 Thread Yoann Vandoorselaere
Le mercredi 30 juillet 2008 à 00:00 +0200, Bruno Haible a écrit : > Yoann Vandoorselaere wrote: > > Additionally, your current lock.h code still make use of abort() which > > I'm reluctant to see in library code. Would you agree to propagate the > > error return in case of problem? > > We started

Re: Lock module improvement

2008-07-29 Thread Bruno Haible
Yoann Vandoorselaere wrote: > Would it be okay if contrary to the current lock.h module (which define > the various gl_lock_ functions as macros), I use real function? Sure. For code that you write, you decide how it's written. I have no requirements coming from gettext or anywhere else. > Addit

Re: Lock module improvement

2008-07-29 Thread Yoann Vandoorselaere
Hi Bruno, Le mardi 22 juillet 2008 à 01:44 +0200, Bruno Haible a écrit : > > > > - Implement support for condition variable. > > > > > > You are welcome to contribute a module for this. > > > > I started working on condition variable support. To me, it make sense to > > do that directly within

Re: Lock module improvement

2008-07-21 Thread Bruno Haible
Hi Yoann, Yoann Vandoorselaere wrote: > > I think you can effectively already do it. Like this: > > > > gl_LOCK > > if test $gl_threads_api = pth; then > > AC_FATAL([The pth threads implementation is not supported by this > > package.]) > > fi > > Thanks! Wouldn't it be better to dire

Re: Lock module improvement

2008-07-21 Thread Yoann Vandoorselaere
Hi Bruno, Le lundi 21 juillet 2008 à 12:24 +0200, Bruno Haible a écrit : > > - Ability to disable specific backend on demand: some implementation use > > certain functionality that are incompatible with libpth, or depend on > > library that provide pthread support but no libpth support. As a resul

Re: Lock module improvement

2008-07-21 Thread Bruno Haible
Hi Yoann, > - Ability to disable specific backend on demand: some implementation use > certain functionality that are incompatible with libpth, or depend on > library that provide pthread support but no libpth support. As a result, > it would be great if it was possible to explicitly disable a giv

Lock module improvement

2008-07-21 Thread Yoann Vandoorselaere
Hi, I would be interested in performing the following lock module modification: - Ability to disable specific backend on demand: some implementation use certain functionality that are incompatible with libpth, or depend on library that provide pthread support but no libpth support. As a result, i