[PATCH] uml: use GFP_ATOMIC for allocations under spinlocks.

setup_initial_poll is only called with sigio_lock() held, so use appropriate
allocation.

Also, parse_chan() can also be called when holding a spinlock (see line_open()
 -> parse_chan_pair()).

I have sporadic problems (spinlock taken twice, with spinlock debugging on UP)
which could be caused by a sequence like "take spinlock, alloc and go to
sleep, take again the spinlock in the other thread".

Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Cc: Jeff Dike <jdike@addtoit.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

diff --git a/arch/um/drivers/chan_kern.c b/arch/um/drivers/chan_kern.c
index 14a12d6..244e54a 100644 (file)
--- a/arch/um/drivers/chan_kern.c
+++ b/arch/um/drivers/chan_kern.c
@@ -465,7 +465,7 @@ static struct chan *parse_chan(char *str, int pri, int device,
        data = (*ops->init)(str, device, opts);
        if(data == NULL) return(NULL);
 
-       chan = kmalloc(sizeof(*chan), GFP_KERNEL);
+       chan = kmalloc(sizeof(*chan), GFP_ATOMIC);
        if(chan == NULL) return(NULL);
        *chan = ((struct chan) { .list          = LIST_HEAD_INIT(chan->list),
                                 .primary       = 1,

diff --git a/arch/um/kernel/sigio_user.c b/arch/um/kernel/sigio_user.c
index e892189..a527511 100644 (file)
--- a/arch/um/kernel/sigio_user.c
+++ b/arch/um/kernel/sigio_user.c
@@ -340,7 +340,7 @@ static int setup_initial_poll(int fd)
 {
        struct pollfd *p;
 
-       p = um_kmalloc(sizeof(struct pollfd));
+       p = um_kmalloc_atomic(sizeof(struct pollfd));
        if(p == NULL){
                printk("setup_initial_poll : failed to allocate poll\n");
                return(-1);