2006-12-25 03:51:47 +03:00
/*
* arch / sh / kernel / process . c
2005-04-17 02:20:36 +04:00
*
2006-12-25 03:51:47 +03:00
* This file handles the architecture - dependent parts of process handling . .
2005-04-17 02:20:36 +04:00
*
* Copyright ( C ) 1995 Linus Torvalds
*
* SuperH version : Copyright ( C ) 1999 , 2000 Niibe Yutaka & Kaz Kojima
2006-10-12 07:16:13 +04:00
* Copyright ( C ) 2006 Lineo Solutions Inc . support SH4A UBC
2008-09-21 14:04:55 +04:00
* Copyright ( C ) 2002 - 2008 Paul Mundt
*
* This file is subject to the terms and conditions of the GNU General Public
* License . See the file " COPYING " in the main directory of this archive
* for more details .
2005-04-17 02:20:36 +04:00
*/
# include <linux/module.h>
# include <linux/mm.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 11:04:11 +03:00
# include <linux/slab.h>
2005-04-17 02:20:36 +04:00
# include <linux/elfcore.h>
# include <linux/kallsyms.h>
2007-07-31 08:01:43 +04:00
# include <linux/fs.h>
2009-07-11 04:29:04 +04:00
# include <linux/ftrace.h>
2009-11-09 10:27:40 +03:00
# include <linux/hw_breakpoint.h>
2011-05-24 12:25:23 +04:00
# include <linux/prefetch.h>
2012-04-19 10:45:57 +04:00
# include <linux/stackprotector.h>
2005-04-17 02:20:36 +04:00
# include <asm/uaccess.h>
# include <asm/mmu_context.h>
2008-03-26 13:02:47 +03:00
# include <asm/fpu.h>
2008-09-04 13:53:58 +04:00
# include <asm/syscalls.h>
2012-03-30 14:29:57 +04:00
# include <asm/switch_to.h>
2005-04-17 02:20:36 +04:00
void show_regs ( struct pt_regs * regs )
{
printk ( " \n " ) ;
2008-11-26 09:20:35 +03:00
printk ( " Pid : %d, Comm: \t \t %s \n " , task_pid_nr ( current ) , current - > comm ) ;
printk ( " CPU : %d \t \t %s (%s %.*s) \n \n " ,
2008-08-07 20:23:34 +04:00
smp_processor_id ( ) , print_tainted ( ) , init_utsname ( ) - > release ,
( int ) strcspn ( init_utsname ( ) - > version , " " ) ,
init_utsname ( ) - > version ) ;
2006-10-12 12:07:45 +04:00
print_symbol ( " PC is at %s \n " , instruction_pointer ( regs ) ) ;
2008-08-07 20:23:34 +04:00
print_symbol ( " PR is at %s \n " , regs - > pr ) ;
2005-04-17 02:20:36 +04:00
printk ( " PC : %08lx SP : %08lx SR : %08lx " ,
regs - > pc , regs - > regs [ 15 ] , regs - > sr ) ;
# ifdef CONFIG_MMU
2010-01-26 06:58:40 +03:00
printk ( " TEA : %08x \n " , __raw_readl ( MMU_TEA ) ) ;
2005-04-17 02:20:36 +04:00
# else
2008-08-07 20:23:34 +04:00
printk ( " \n " ) ;
2005-04-17 02:20:36 +04:00
# endif
printk ( " R0 : %08lx R1 : %08lx R2 : %08lx R3 : %08lx \n " ,
regs - > regs [ 0 ] , regs - > regs [ 1 ] ,
regs - > regs [ 2 ] , regs - > regs [ 3 ] ) ;
printk ( " R4 : %08lx R5 : %08lx R6 : %08lx R7 : %08lx \n " ,
regs - > regs [ 4 ] , regs - > regs [ 5 ] ,
regs - > regs [ 6 ] , regs - > regs [ 7 ] ) ;
printk ( " R8 : %08lx R9 : %08lx R10 : %08lx R11 : %08lx \n " ,
regs - > regs [ 8 ] , regs - > regs [ 9 ] ,
regs - > regs [ 10 ] , regs - > regs [ 11 ] ) ;
printk ( " R12 : %08lx R13 : %08lx R14 : %08lx \n " ,
regs - > regs [ 12 ] , regs - > regs [ 13 ] ,
regs - > regs [ 14 ] ) ;
printk ( " MACH: %08lx MACL: %08lx GBR : %08lx PR : %08lx \n " ,
regs - > mach , regs - > macl , regs - > gbr , regs - > pr ) ;
2006-10-12 12:07:45 +04:00
show_trace ( NULL , ( unsigned long * ) regs - > regs [ 15 ] , regs ) ;
2008-11-26 08:31:03 +03:00
show_code ( regs ) ;
2005-04-17 02:20:36 +04:00
}
/*
* Create a kernel thread
*/
2012-01-13 05:17:21 +04:00
__noreturn void kernel_thread_helper ( void * arg , int ( * fn ) ( void * ) )
2008-12-17 06:20:15 +03:00
{
do_exit ( fn ( arg ) ) ;
}
2005-04-17 02:20:36 +04:00
2006-12-25 03:51:47 +03:00
/* Don't use this in BL=1(cli). Or else, CPU resets! */
2005-04-17 02:20:36 +04:00
int kernel_thread ( int ( * fn ) ( void * ) , void * arg , unsigned long flags )
2006-12-25 03:51:47 +03:00
{
2005-04-17 02:20:36 +04:00
struct pt_regs regs ;
2008-09-21 08:56:39 +04:00
int pid ;
2005-04-17 02:20:36 +04:00
memset ( & regs , 0 , sizeof ( regs ) ) ;
2006-12-25 03:51:47 +03:00
regs . regs [ 4 ] = ( unsigned long ) arg ;
regs . regs [ 5 ] = ( unsigned long ) fn ;
2005-04-17 02:20:36 +04:00
2006-12-25 03:51:47 +03:00
regs . pc = ( unsigned long ) kernel_thread_helper ;
sh: Minor optimisations to FPU handling
A number of small optimisations to FPU handling, in particular:
- move the task USEDFPU flag from the thread_info flags field (which
is accessed asynchronously to the thread) to a new status field,
which is only accessed by the thread itself. This allows locking to
be removed in most cases, or can be reduced to a preempt_lock().
This mimics the i386 behaviour.
- move the modification of regs->sr and thread_info->status flags out
of save_fpu() to __unlazy_fpu(). This gives the compiler a better
chance to optimise things, as well as making save_fpu() symmetrical
with restore_fpu() and init_fpu().
- implement prepare_to_copy(), so that when creating a thread, we can
unlazy the FPU prior to copying the thread data structures.
Also make sure that the FPU is disabled while in the kernel, in
particular while booting, and for newly created kernel threads,
In a very artificial benchmark, the execution time for 2500000
context switches was reduced from 50 to 45 seconds.
Signed-off-by: Stuart Menefy <stuart.menefy@st.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2009-09-25 21:25:10 +04:00
regs . sr = SR_MD ;
# if defined(CONFIG_SH_FPU)
regs . sr | = SR_FD ;
# endif
2005-04-17 02:20:36 +04:00
/* Ok, create the new process.. */
2008-09-21 08:56:39 +04:00
pid = do_fork ( flags | CLONE_VM | CLONE_UNTRACED , 0 ,
& regs , 0 , NULL , NULL ) ;
return pid ;
2005-04-17 02:20:36 +04:00
}
2009-10-27 05:51:19 +03:00
EXPORT_SYMBOL ( kernel_thread ) ;
2005-04-17 02:20:36 +04:00
2010-01-12 12:52:00 +03:00
void start_thread ( struct pt_regs * regs , unsigned long new_pc ,
unsigned long new_sp )
{
regs - > pr = 0 ;
regs - > sr = SR_FD ;
regs - > pc = new_pc ;
regs - > regs [ 15 ] = new_sp ;
2010-01-13 06:51:40 +03:00
free_thread_xstate ( current ) ;
2010-01-12 12:52:00 +03:00
}
EXPORT_SYMBOL ( start_thread ) ;
2005-04-17 02:20:36 +04:00
/*
* Free current thread data structures etc . .
*/
void exit_thread ( void )
{
}
void flush_thread ( void )
{
struct task_struct * tsk = current ;
2009-11-09 10:27:40 +03:00
flush_ptrace_hw_breakpoint ( tsk ) ;
# if defined(CONFIG_SH_FPU)
2005-04-17 02:20:36 +04:00
/* Forget lazy FPU state */
2006-01-12 12:05:44 +03:00
clear_fpu ( tsk , task_pt_regs ( tsk ) ) ;
2005-04-17 02:20:36 +04:00
clear_used_math ( ) ;
# endif
}
void release_thread ( struct task_struct * dead_task )
{
/* do nothing */
}
/* Fill in the fpu structure for a core dump.. */
int dump_fpu ( struct pt_regs * regs , elf_fpregset_t * fpu )
{
int fpvalid = 0 ;
# if defined(CONFIG_SH_FPU)
struct task_struct * tsk = current ;
fpvalid = ! ! tsk_used_math ( tsk ) ;
2008-09-21 14:04:55 +04:00
if ( fpvalid )
fpvalid = ! fpregs_get ( tsk , NULL , 0 ,
sizeof ( struct user_fpu_struct ) ,
fpu , NULL ) ;
2005-04-17 02:20:36 +04:00
# endif
return fpvalid ;
}
2009-10-27 05:51:19 +03:00
EXPORT_SYMBOL ( dump_fpu ) ;
2005-04-17 02:20:36 +04:00
asmlinkage void ret_from_fork ( void ) ;
2009-04-03 03:56:59 +04:00
int copy_thread ( unsigned long clone_flags , unsigned long usp ,
2005-04-17 02:20:36 +04:00
unsigned long unused ,
struct task_struct * p , struct pt_regs * regs )
{
2006-09-27 12:07:07 +04:00
struct thread_info * ti = task_thread_info ( p ) ;
2005-04-17 02:20:36 +04:00
struct pt_regs * childregs ;
sh: Minor optimisations to FPU handling
A number of small optimisations to FPU handling, in particular:
- move the task USEDFPU flag from the thread_info flags field (which
is accessed asynchronously to the thread) to a new status field,
which is only accessed by the thread itself. This allows locking to
be removed in most cases, or can be reduced to a preempt_lock().
This mimics the i386 behaviour.
- move the modification of regs->sr and thread_info->status flags out
of save_fpu() to __unlazy_fpu(). This gives the compiler a better
chance to optimise things, as well as making save_fpu() symmetrical
with restore_fpu() and init_fpu().
- implement prepare_to_copy(), so that when creating a thread, we can
unlazy the FPU prior to copying the thread data structures.
Also make sure that the FPU is disabled while in the kernel, in
particular while booting, and for newly created kernel threads,
In a very artificial benchmark, the execution time for 2500000
context switches was reduced from 50 to 45 seconds.
Signed-off-by: Stuart Menefy <stuart.menefy@st.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2009-09-25 21:25:10 +04:00
# if defined(CONFIG_SH_DSP)
2005-04-17 02:20:36 +04:00
struct task_struct * tsk = current ;
2009-04-03 21:32:33 +04:00
if ( is_dsp_enabled ( tsk ) ) {
/* We can use the __save_dsp or just copy the struct:
* __save_dsp ( p ) ;
* p - > thread . dsp_status . status | = SR_DSP
*/
p - > thread . dsp_status = tsk - > thread . dsp_status ;
}
# endif
2006-01-12 12:05:44 +03:00
childregs = task_pt_regs ( p ) ;
2005-04-17 02:20:36 +04:00
* childregs = * regs ;
if ( user_mode ( regs ) ) {
childregs - > regs [ 15 ] = usp ;
2006-09-27 12:07:07 +04:00
ti - > addr_limit = USER_DS ;
2005-04-17 02:20:36 +04:00
} else {
2007-02-28 12:35:42 +03:00
childregs - > regs [ 15 ] = ( unsigned long ) childregs ;
2006-09-27 12:07:07 +04:00
ti - > addr_limit = KERNEL_DS ;
sh: Minor optimisations to FPU handling
A number of small optimisations to FPU handling, in particular:
- move the task USEDFPU flag from the thread_info flags field (which
is accessed asynchronously to the thread) to a new status field,
which is only accessed by the thread itself. This allows locking to
be removed in most cases, or can be reduced to a preempt_lock().
This mimics the i386 behaviour.
- move the modification of regs->sr and thread_info->status flags out
of save_fpu() to __unlazy_fpu(). This gives the compiler a better
chance to optimise things, as well as making save_fpu() symmetrical
with restore_fpu() and init_fpu().
- implement prepare_to_copy(), so that when creating a thread, we can
unlazy the FPU prior to copying the thread data structures.
Also make sure that the FPU is disabled while in the kernel, in
particular while booting, and for newly created kernel threads,
In a very artificial benchmark, the execution time for 2500000
context switches was reduced from 50 to 45 seconds.
Signed-off-by: Stuart Menefy <stuart.menefy@st.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2009-09-25 21:25:10 +04:00
ti - > status & = ~ TS_USEDFPU ;
p - > fpu_counter = 0 ;
2005-04-17 02:20:36 +04:00
}
2006-12-25 03:51:47 +03:00
2007-02-28 12:35:42 +03:00
if ( clone_flags & CLONE_SETTLS )
2005-04-17 02:20:36 +04:00
childregs - > gbr = childregs - > regs [ 0 ] ;
2006-12-25 03:51:47 +03:00
2005-04-17 02:20:36 +04:00
childregs - > regs [ 0 ] = 0 ; /* Set return value for child */
p - > thread . sp = ( unsigned long ) childregs ;
p - > thread . pc = ( unsigned long ) ret_from_fork ;
2009-11-09 10:27:40 +03:00
memset ( p - > thread . ptrace_bps , 0 , sizeof ( p - > thread . ptrace_bps ) ) ;
2005-04-17 02:20:36 +04:00
return 0 ;
}
/*
* switch_to ( x , y ) should switch tasks from x to y .
*
*/
2009-07-11 04:29:04 +04:00
__notrace_funcgraph struct task_struct *
__switch_to ( struct task_struct * prev , struct task_struct * next )
2005-04-17 02:20:36 +04:00
{
2009-07-07 18:25:10 +04:00
struct thread_struct * next_t = & next - > thread ;
2012-04-19 10:45:57 +04:00
# if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP)
__stack_chk_guard = next - > stack_canary ;
# endif
2006-01-12 12:05:44 +03:00
unlazy_fpu ( prev , task_pt_regs ( prev ) ) ;
2009-07-07 18:25:10 +04:00
/* we're going to use this soon, after a few expensive things */
if ( next - > fpu_counter > 5 )
2010-01-13 06:51:40 +03:00
prefetch ( next_t - > xstate ) ;
2005-04-17 02:20:36 +04:00
2006-09-27 12:25:07 +04:00
# ifdef CONFIG_MMU
2005-04-17 02:20:36 +04:00
/*
* Restore the kernel mode register
2006-12-25 03:51:47 +03:00
* k7 ( r7_bank1 )
2005-04-17 02:20:36 +04:00
*/
asm volatile ( " ldc %0, r7_bank "
: /* no output */
2006-01-12 12:05:45 +03:00
: " r " ( task_thread_info ( next ) ) ) ;
2006-09-27 12:25:07 +04:00
# endif
2005-04-17 02:20:36 +04:00
2009-11-25 06:07:31 +03:00
/*
* If the task has used fpu the last 5 timeslices , just do a full
2009-07-07 18:25:10 +04:00
* restore of the math state immediately to avoid the trap ; the
* chances of needing FPU soon are obviously high now
*/
2009-11-25 06:07:31 +03:00
if ( next - > fpu_counter > 5 )
2010-01-13 06:51:40 +03:00
__fpu_state_restore ( ) ;
2009-07-07 18:25:10 +04:00
2005-04-17 02:20:36 +04:00
return prev ;
}
asmlinkage int sys_fork ( unsigned long r4 , unsigned long r5 ,
unsigned long r6 , unsigned long r7 ,
2006-11-21 05:16:57 +03:00
struct pt_regs __regs )
2005-04-17 02:20:36 +04:00
{
# ifdef CONFIG_MMU
2007-05-14 12:26:34 +04:00
struct pt_regs * regs = RELOC_HIDE ( & __regs , 0 ) ;
2006-11-21 05:16:57 +03:00
return do_fork ( SIGCHLD , regs - > regs [ 15 ] , regs , 0 , NULL , NULL ) ;
2005-04-17 02:20:36 +04:00
# else
/* fork almost works, enough to trick you into looking elsewhere :-( */
return - EINVAL ;
# endif
}
asmlinkage int sys_clone ( unsigned long clone_flags , unsigned long newsp ,
unsigned long parent_tidptr ,
unsigned long child_tidptr ,
2006-11-21 05:16:57 +03:00
struct pt_regs __regs )
2005-04-17 02:20:36 +04:00
{
2006-11-21 05:16:57 +03:00
struct pt_regs * regs = RELOC_HIDE ( & __regs , 0 ) ;
2005-04-17 02:20:36 +04:00
if ( ! newsp )
2006-11-21 05:16:57 +03:00
newsp = regs - > regs [ 15 ] ;
return do_fork ( clone_flags , newsp , regs , 0 ,
2006-12-25 03:51:47 +03:00
( int __user * ) parent_tidptr ,
( int __user * ) child_tidptr ) ;
2005-04-17 02:20:36 +04:00
}
/*
* This is trivial , and on the face of it looks like it
* could equally well be done in user mode .
*
* Not so , for quite unobvious reasons - register pressure .
* In user mode vfork ( ) cannot have a stack frame , and if
* done by calling the " clone() " system call directly , you
* do not have enough call - clobbered registers to hold all
* the information you need .
*/
asmlinkage int sys_vfork ( unsigned long r4 , unsigned long r5 ,
unsigned long r6 , unsigned long r7 ,
2006-11-21 05:16:57 +03:00
struct pt_regs __regs )
2005-04-17 02:20:36 +04:00
{
2006-11-21 05:16:57 +03:00
struct pt_regs * regs = RELOC_HIDE ( & __regs , 0 ) ;
return do_fork ( CLONE_VFORK | CLONE_VM | SIGCHLD , regs - > regs [ 15 ] , regs ,
2005-04-17 02:20:36 +04:00
0 , NULL , NULL ) ;
}
/*
* sys_execve ( ) executes a new program .
*/
2010-08-18 02:52:56 +04:00
asmlinkage int sys_execve ( const char __user * ufilename ,
const char __user * const __user * uargv ,
const char __user * const __user * uenvp ,
unsigned long r7 , struct pt_regs __regs )
2005-04-17 02:20:36 +04:00
{
2006-11-21 05:16:57 +03:00
struct pt_regs * regs = RELOC_HIDE ( & __regs , 0 ) ;
2005-04-17 02:20:36 +04:00
int error ;
char * filename ;
2007-05-14 07:52:56 +04:00
filename = getname ( ufilename ) ;
2005-04-17 02:20:36 +04:00
error = PTR_ERR ( filename ) ;
if ( IS_ERR ( filename ) )
goto out ;
2007-05-14 07:52:56 +04:00
error = do_execve ( filename , uargv , uenvp , regs ) ;
2005-04-17 02:20:36 +04:00
putname ( filename ) ;
out :
return error ;
}
unsigned long get_wchan ( struct task_struct * p )
{
unsigned long pc ;
if ( ! p | | p = = current | | p - > state = = TASK_RUNNING )
return 0 ;
/*
* The same comment as on the Alpha applies here , too . . .
*/
pc = thread_saved_pc ( p ) ;
2007-07-26 12:46:07 +04:00
# ifdef CONFIG_FRAME_POINTER
2005-04-17 02:20:36 +04:00
if ( in_sched_functions ( pc ) ) {
2007-07-26 12:46:07 +04:00
unsigned long schedule_frame = ( unsigned long ) p - > thread . sp ;
2006-12-08 11:46:29 +03:00
return ( ( unsigned long * ) schedule_frame ) [ 21 ] ;
2005-04-17 02:20:36 +04:00
}
2007-07-26 12:46:07 +04:00
# endif
2006-12-08 11:46:29 +03:00
2005-04-17 02:20:36 +04:00
return pc ;
}