Commited the following to ams-branch (with changes noted by Thomas).
Index: serverboot/ChangeLog
2005-08-13 Alfred M. Szmidt <[EMAIL PROTECTED]>
* kalloc.c: #include
(init_hook, malloc_hook, free_hook): New functions.
(__malloc_initialize_hook): New variable.
(m
My question was rather if we can be sure that realloc() or
memalign() aren't used by e.g. printf() or other library functions
that mach-defpager is calling.
As I told you, you need to check that in the source code for realloc
and memalign.
__
On Thu, Sep 01, 2005 at 01:25:29PM +0200, Alfred M. Szmidt wrote:
>Can we be sure that we don't have to also hook realloc() and
>memalign()?
>
> I don't think so, check the code for realloc/memalign to be sure, I
> don't have access to it here.
I don't think they're used in the Hurd code;
Uhm, did you did you read that message of mine?
http://lists.gnu.org/archive/html/bug-hurd/2005-08/msg00253.html>
Must have missed it, or great chuncks of it, sorry.
For convenience, I attached the message to this mail.
Thanks.
Can we be sure that we don't have to also hook realloc(
On Wed, Aug 31, 2005 at 04:54:36PM +0200, Alfred M. Szmidt wrote:
>Ideally, fix the declaration of init_hook. Failing that, at least
>give a correct cast instead of kludging with void *.
>
>void (*__malloc_initialize_hook) (void) = (void (*)(void)) init_hook;
>
> This is nicer, indee
Ideally, fix the declaration of init_hook. Failing that, at least
give a correct cast instead of kludging with void *.
void (*__malloc_initialize_hook) (void) = (void (*)(void)) init_hook;
This is nicer, indeed. Could someone apply the patch with this
modification?
__
I think the proposed code should give a compiler error (which doesn't mean
that it does, of course).
The code is correct C99 code.
___
Bug-hurd mailing list
Bug-hurd@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-hurd
Shall I do any further testing (and how to do that)?
I told you, attach gdb to defpager, and put a breakpoint in malloc.
Since this is a hook, any usual code from malloc shouldn't be run.
> > +#include /* for malloc_hook/free_hook */
> > +
> > +void *init_hook (void);
Note that this is not my patch we're discussing, but as Alfred is
currently on vacations, ...
On Sun, Aug 28, 2005 at 02:51:37PM -0700, Thomas Bushnell BSG wrote:
> Thomas Schwinge <[EMAIL PROTECTED]> writes:
> >> > +void (*__malloc_initialize_hook) (void) = (void *) init_hook;
> >>
> >> Do you r
Bas Wijnen <[EMAIL PROTECTED]> writes:
> I think the proposed code should give a compiler error (which doesn't mean
> that it does, of course). Function pointers and data pointers are really
> different types, and cannot be mixed. They need not even be the same size.
Except that (void *) is com
Thomas Schwinge <[EMAIL PROTECTED]> writes:
>> > +void (*__malloc_initialize_hook) (void) = (void *) init_hook;
>>
>> Do you really need the void * cast?
> The cast avoids the following warning:
> #v+
> [...]
> ../../hurd-0/mach-defpager/../serverboot/kalloc.c:43: warning: initialization
> from
On Sun, Aug 28, 2005 at 11:03:31PM +0200, Thomas Schwinge wrote:
> > > +
> > > +void (*__malloc_initialize_hook) (void) = (void *) init_hook;
> >
> > Do you really need the void * cast?
>
> The cast avoids the following warning:
> #v+
> [...]
> ../../hurd-0/mach-defpager/../serverboot/kalloc.c:4
On Sun, Aug 28, 2005 at 09:00:41PM +0100, Neal H. Walfield wrote:
> > What about the following fix? This should fix things so that we don't
> > use vanilla malloc/free, and only use our own versions.
>
> This looks okay. Has it been tested?
Michael Banck and I have been running the fixed binary
> What about the following fix? This should fix things so that we don't
> use vanilla malloc/free, and only use our own versions.
This looks okay. Has it been tested?
> +#include /* for malloc_hook/free_hook */
> +
> +void *init_hook (void);
> +void *malloc_hook (size_t size, const vo
I reported this a while back (it happens with gcc 4.x too):
/usr/local/lib/gcc/i386-unknown-gnu0.3/3.4.0/../../../../i386-unknown-gnu0.3/li
b/libcrt.a(malloc.o)(.text+0x10d8):
In function `__libc_free':
/export/ams/libc/malloc/malloc.c:3355: multiple definition of `free'
kalloc.o(.text+0x334):../
15 matches
Mail list logo
