Re: [Fault-injection-developer] Kernel module's implementation in 2.5.48/2.5.49
Status: Alpha
Brought to you by:
rustyl
From: Rusty L. <ru...@li...> - 2002-11-26 19:58:12
|
I disagree. We need to be working on the front lines with the rest of the kernel hackers. If in our development we find problems with other subsystems, then we report them and if possible supply a patch that fixes the problem. -rustyl ----- Original Message ----- From: "Zhuang, Louis" <lou...@in...> To: "Wang, Stanley" <sta...@in...>; "SourceforgeFI mail list (E-mail)" <fau...@li...> Sent: Tuesday, November 26, 2002 1:58 AM Subject: RE: [Fault-injection-developer] Kernel module's implementation in 2.5.48/2.5.49 > According to Stanley's investigation, I'd like to suggest delay our porting > against new module API until they are stable and compatible enough. -Louis > > > -----Original Message----- > > From: Wang, Stanley [mailto:sta...@in...] > > Sent: Tuesday, November 26, 2002 4:40 PM > > To: SourceforgeFI mail list (E-mail) > > Subject: [Fault-injection-developer] Kernel module's > > implementaiton in > > 2.5.48/2.5.49 > > > > > > Hello folks, > > The kernel module's implementation has been changed a lot > > since 2.5.48. And > > some of them would > > influence our development work: > > 1. kernel module's name should be specified by defining > > KBUILD_MODNAME > > 2. The in-kernel module loader had been rewrited, the > > modversions.h had been > > eliminated > > 3. The kernel implemented a new mechanism for get/put > > symbols between > > different modules: > > symbol_get(const char*)/symbol_put(const char > > *)/symbol_put_addr(void *). > > These function > > could get/put a symbol that had been exported by another > > kernel module. > > But there is a little bug with them, the following patch > > could fix this bug: > > > > > > diff -Naur -X dontdiff linux-2.5.49/include/linux/module.h > > linux-2.5.49-bugfix/include/linux/module.h > > --- linux-2.5.49/include/linux/module.h 2002-11-26 > > 16:06:36.000000000 +0800 > > +++ linux-2.5.49-bugfix/include/linux/module.h 2002-11-26 > > 16:01:52.000000000 +0800 > > @@ -86,7 +86,7 @@ > > /* Get/put a kernel symbol (calls must be symmetric) */ > > void *__symbol_get(const char *symbol); > > void *__symbol_get_gpl(const char *symbol); > > -#define symbol_get(x) ((typeof(&x))(__symbol_get(#x))) > > +#define symbol_get(x) ((typeof(&x))(__symbol_get(x))) > > > > /* For every exported symbol, place a struct in the > > __ksymtab section */ > > #define EXPORT_SYMBOL(sym) \ > > @@ -166,7 +166,7 @@ > > #ifdef CONFIG_MODULE_UNLOAD > > > > void __symbol_put(const char *symbol); > > -#define symbol_put(x) __symbol_put(#x) > > +#define symbol_put(x) __symbol_put(x) > > void symbol_put_addr(void *addr); > > > > /* We only need protection against local interrupts. */ > > > > > > > > 4. You should get a new module utiliy to cooperate with > > 2.5.48/2.5.49 > > (from http://ozlabs.org/~rusty/module-init-tools-0.7.tar.gz) > > > ------------------------------------------------------- > This SF.net email is sponsored by: Get the new Palm Tungsten T > handheld. Power & Color in a compact size! > http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en > _______________________________________________ > Fault-injection-developer mailing list > Fau...@li... > https://lists.sourceforge.net/lists/listinfo/fault-injection-developer |