> Yes, this case probably doesn't get enough testing with my current
> setup, so it is quite hard (for me) to prove removing lockmap is
> no-no. Problem is that I don't understand shm swapping very well
Launch some time this proggy to try out shm swapping:
/*
* Copyright (C) 1999 Andrea Arcangeli
* shm swapout test
*/
#include <sys/ipc.h>
#include <sys/shm.h>
#define SIZE 16000000
main()
{
int shmid;
char *addr, *p;
if ((shmid = shmget(IPC_PRIVATE, SIZE, IPC_CREAT | 0644)) < 0)
perror("shmget");
if ((addr = shmat(shmid, NULL, 0)) < 0)
perror("shmat");
for (p = addr; p < addr + SIZE; p+=4096)
*p = 0;
}
To know if the lockmap is needed you can also reinsert the code and add a
printk() in the test_and_set_bit() path.
Andrea Arcangeli
-
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/