Re: [PATCH] fix shmat

["Randy.Dunlap" <rddunlap@osdl.org>]

On Mon, 23 Feb 2004 19:15:04 +0100 Manfred Spraul <manfred@colorfullife.com> wrote:

|  From the bitkeeper commit message queue:
| 
| >	
| >	sys_shmat() need to be declared asmlinkage.  This causes breakage when we
| >	actually get the proper prototypes into caller's scope.
| >  
| >
| Why? sys_shmat is not a system call. Or at least there is a comment just 
| before the implementation that this is not a syscall.
| I think either the asmlinkage or the comment are wrong:
| /*
|  * Fix shmaddr, allocate descriptor, map shm, add attach descriptor to 
| lists.
|  *
|  * NOTE! Despite the name, this is NOT a direct system call entrypoint. The
| 
| >  * "raddr" thing points to kernel space, and there has to be a wrapper around
| >  * this.
| >  */
| >-long sys_shmat(int shmid, char __user *shmaddr, int shmflg, ulong *raddr)
| >+asmlinkage long sys_shmat(int shmid, char __user *shmaddr, int shmflg, ulong *raddr)
| > {
| > 	struct shmid_kernel *shp;
| > 	unsigned long addr;
| >
| 
| I'd propose to remove the asmlinkage and to move the prototype (without 
| asmlinkage) back from syscalls.h to shm.h - what do you think?

It's not a syscall AFAICT.
It's not listed in any .S files, like most syscalls are.
However, it is listed in kernel/sys.c as a "cond_syscall",
which I'm guessing is incorrect.

I'd like to rename it so that it doesn't begin with "sys_".

--
~Randy