Messages in this thread | | | From | "Manfred" <> | Subject | Re: 64-bit shm: abi problems | Date | Mon, 13 Dec 1999 21:04:00 +0100 |
| |
From: Christoph Rohland <hans-christoph.rohland@sap.com> > "David S. Miller" <davem@redhat.com> writes: > > You cannot simply change the ABI and still use the old system call > > entry, this is not an option. > > On many platforms you probably do not need a new syscall. They use the > sys_ipc demultiplexer and you can simply add a new sys_ipc command. > should we really add another function to the sys_ipc demultiplexer? Alpha has normal syscalls, and all other architectures complain about the "horrible hack".
I would propose that we add a new sysctl "shmget2()" with a "size_t size" variable, and that we truncate the segment length for shmctl(IPC_GET) to MAX_INT. The message queue code does that as well, and we can fix it properly with the 32-bit UID patches.
Btw, both the Sparc64 and the PPC code still call "lock_kernel()" before calling the ipc functions. The new ipc functions are thread safe, someone could remove these calls.
-- Manfred
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |