Drop might_sleep() calls from get_user() & co. This should fix the issue

in http://www.linux-mips.org/cgi-bin/mesg.cgi?a=linux-mips&i=200508171321.20094.Joshua.Wise%40sicortex.com and it's the right thing to do anyway because
it was inflating those functions way too much.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This commit is contained in:
Ralf Baechle 2005-08-17 17:57:34 +00:00
parent 340ee4b98c
commit 27c7c1657d

View file

@ -213,7 +213,6 @@ struct __large_struct { unsigned long buf[100]; };
__typeof(*(ptr)) __gu_val = (__typeof(*(ptr))) 0; \
long __gu_err = 0; \
\
might_sleep(); \
switch (size) { \
case 1: __get_user_asm("lb", ptr); break; \
case 2: __get_user_asm("lh", ptr); break; \
@ -300,7 +299,6 @@ extern void __get_user_unknown(void);
__typeof__(*(ptr)) __pu_val; \
long __pu_err = 0; \
\
might_sleep(); \
__pu_val = (x); \
switch (size) { \
case 1: __put_user_asm("sb", ptr); break; \