2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* S390 version
|
2012-07-20 17:15:04 +08:00
|
|
|
* Copyright IBM Corp. 1999
|
2005-04-17 06:20:36 +08:00
|
|
|
* Author(s): Hartmut Penner (hp@de.ibm.com)
|
|
|
|
* Ulrich Weigand (uweigand@de.ibm.com)
|
|
|
|
*
|
|
|
|
* Derived from "arch/i386/mm/fault.c"
|
|
|
|
* Copyright (C) 1995 Linus Torvalds
|
|
|
|
*/
|
|
|
|
|
2011-01-05 19:47:28 +08:00
|
|
|
#include <linux/kernel_stat.h>
|
perf: Do the big rename: Performance Counters -> Performance Events
Bye-bye Performance Counters, welcome Performance Events!
In the past few months the perfcounters subsystem has grown out its
initial role of counting hardware events, and has become (and is
becoming) a much broader generic event enumeration, reporting, logging,
monitoring, analysis facility.
Naming its core object 'perf_counter' and naming the subsystem
'perfcounters' has become more and more of a misnomer. With pending
code like hw-breakpoints support the 'counter' name is less and
less appropriate.
All in one, we've decided to rename the subsystem to 'performance
events' and to propagate this rename through all fields, variables
and API names. (in an ABI compatible fashion)
The word 'event' is also a bit shorter than 'counter' - which makes
it slightly more convenient to write/handle as well.
Thanks goes to Stephane Eranian who first observed this misnomer and
suggested a rename.
User-space tooling and ABI compatibility is not affected - this patch
should be function-invariant. (Also, defconfigs were not touched to
keep the size down.)
This patch has been generated via the following script:
FILES=$(find * -type f | grep -vE 'oprofile|[^K]config')
sed -i \
-e 's/PERF_EVENT_/PERF_RECORD_/g' \
-e 's/PERF_COUNTER/PERF_EVENT/g' \
-e 's/perf_counter/perf_event/g' \
-e 's/nb_counters/nb_events/g' \
-e 's/swcounter/swevent/g' \
-e 's/tpcounter_event/tp_event/g' \
$FILES
for N in $(find . -name perf_counter.[ch]); do
M=$(echo $N | sed 's/perf_counter/perf_event/g')
mv $N $M
done
FILES=$(find . -name perf_event.*)
sed -i \
-e 's/COUNTER_MASK/REG_MASK/g' \
-e 's/COUNTER/EVENT/g' \
-e 's/\<event\>/event_id/g' \
-e 's/counter/event/g' \
-e 's/Counter/Event/g' \
$FILES
... to keep it as correct as possible. This script can also be
used by anyone who has pending perfcounters patches - it converts
a Linux kernel tree over to the new naming. We tried to time this
change to the point in time where the amount of pending patches
is the smallest: the end of the merge window.
Namespace clashes were fixed up in a preparatory patch - and some
stylistic fallout will be fixed up in a subsequent patch.
( NOTE: 'counters' are still the proper terminology when we deal
with hardware registers - and these sed scripts are a bit
over-eager in renaming them. I've undone some of that, but
in case there's something left where 'counter' would be
better than 'event' we can undo that on an individual basis
instead of touching an otherwise nicely automated patch. )
Suggested-by: Stephane Eranian <eranian@google.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Paul Mackerras <paulus@samba.org>
Reviewed-by: Arjan van de Ven <arjan@linux.intel.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: <linux-arch@vger.kernel.org>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 18:02:48 +08:00
|
|
|
#include <linux/perf_event.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <linux/signal.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/ptrace.h>
|
|
|
|
#include <linux/mman.h>
|
|
|
|
#include <linux/mm.h>
|
2009-06-12 16:26:25 +08:00
|
|
|
#include <linux/compat.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <linux/smp.h>
|
2007-05-08 15:27:03 +08:00
|
|
|
#include <linux/kdebug.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/console.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/hardirq.h>
|
2006-09-20 21:58:39 +08:00
|
|
|
#include <linux/kprobes.h>
|
2007-04-27 22:01:44 +08:00
|
|
|
#include <linux/uaccess.h>
|
2008-04-30 19:38:46 +08:00
|
|
|
#include <linux/hugetlb.h>
|
2010-02-27 05:37:43 +08:00
|
|
|
#include <asm/asm-offsets.h>
|
2015-08-20 23:28:44 +08:00
|
|
|
#include <asm/diag.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <asm/pgtable.h>
|
2016-03-08 18:49:57 +08:00
|
|
|
#include <asm/gmap.h>
|
2011-05-26 15:48:24 +08:00
|
|
|
#include <asm/irq.h>
|
2008-02-10 01:24:37 +08:00
|
|
|
#include <asm/mmu_context.h>
|
2012-03-29 01:30:02 +08:00
|
|
|
#include <asm/facility.h>
|
2008-04-17 13:46:26 +08:00
|
|
|
#include "../kernel/entry.h"
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
#define __FAIL_ADDR_MASK -4096L
|
|
|
|
#define __SUBCODE_MASK 0x0600
|
|
|
|
#define __PF_RES_FIELD 0x8000000000000000ULL
|
|
|
|
|
2009-12-07 19:51:45 +08:00
|
|
|
#define VM_FAULT_BADCONTEXT 0x010000
|
|
|
|
#define VM_FAULT_BADMAP 0x020000
|
|
|
|
#define VM_FAULT_BADACCESS 0x040000
|
2012-10-30 21:49:37 +08:00
|
|
|
#define VM_FAULT_SIGNAL 0x080000
|
2013-06-17 22:25:18 +08:00
|
|
|
#define VM_FAULT_PFAULT 0x100000
|
2009-12-07 19:51:45 +08:00
|
|
|
|
2012-10-30 21:49:37 +08:00
|
|
|
static unsigned long store_indication __read_mostly;
|
2010-10-25 22:10:13 +08:00
|
|
|
|
2012-10-30 21:49:37 +08:00
|
|
|
static int __init fault_init(void)
|
2010-10-25 22:10:13 +08:00
|
|
|
{
|
2012-10-30 21:49:37 +08:00
|
|
|
if (test_facility(75))
|
2010-10-25 22:10:13 +08:00
|
|
|
store_indication = 0xc00;
|
2012-10-30 21:49:37 +08:00
|
|
|
return 0;
|
2010-10-25 22:10:13 +08:00
|
|
|
}
|
2012-10-30 21:49:37 +08:00
|
|
|
early_initcall(fault_init);
|
2010-10-25 22:10:13 +08:00
|
|
|
|
2009-12-07 19:51:44 +08:00
|
|
|
static inline int notify_page_fault(struct pt_regs *regs)
|
2007-04-27 22:01:43 +08:00
|
|
|
{
|
2007-05-05 00:47:46 +08:00
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/* kprobe_running() needs smp_processor_id() */
|
2010-02-27 05:37:45 +08:00
|
|
|
if (kprobes_built_in() && !user_mode(regs)) {
|
2007-05-05 00:47:46 +08:00
|
|
|
preempt_disable();
|
|
|
|
if (kprobe_running() && kprobe_fault_handler(regs, 14))
|
|
|
|
ret = 1;
|
|
|
|
preempt_enable();
|
|
|
|
}
|
|
|
|
return ret;
|
2006-09-20 21:58:39 +08:00
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Unlock any spinlocks which will prevent us from getting the
|
2007-02-10 17:46:18 +08:00
|
|
|
* message out.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
|
|
|
void bust_spinlocks(int yes)
|
|
|
|
{
|
|
|
|
if (yes) {
|
|
|
|
oops_in_progress = 1;
|
|
|
|
} else {
|
|
|
|
int loglevel_save = console_loglevel;
|
|
|
|
console_unblank();
|
|
|
|
oops_in_progress = 0;
|
|
|
|
/*
|
|
|
|
* OK, the message is on the console. Now we call printk()
|
|
|
|
* without oops_in_progress set so that printk will give klogd
|
|
|
|
* a poke. Hold onto your hats...
|
|
|
|
*/
|
|
|
|
console_loglevel = 15;
|
|
|
|
printk(" ");
|
|
|
|
console_loglevel = loglevel_save;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2007-03-06 06:35:54 +08:00
|
|
|
* Returns the address space associated with the fault.
|
2009-12-07 19:51:42 +08:00
|
|
|
* Returns 0 for kernel space and 1 for user space.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
s390/uaccess: rework uaccess code - fix locking issues
The current uaccess code uses a page table walk in some circumstances,
e.g. in case of the in atomic futex operations or if running on old
hardware which doesn't support the mvcos instruction.
However it turned out that the page table walk code does not correctly
lock page tables when accessing page table entries.
In other words: a different cpu may invalidate a page table entry while
the current cpu inspects the pte. This may lead to random data corruption.
Adding correct locking however isn't trivial for all uaccess operations.
Especially copy_in_user() is problematic since that requires to hold at
least two locks, but must be protected against ABBA deadlock when a
different cpu also performs a copy_in_user() operation.
So the solution is a different approach where we change address spaces:
User space runs in primary address mode, or access register mode within
vdso code, like it currently already does.
The kernel usually also runs in home space mode, however when accessing
user space the kernel switches to primary or secondary address mode if
the mvcos instruction is not available or if a compare-and-swap (futex)
instruction on a user space address is performed.
KVM however is special, since that requires the kernel to run in home
address space while implicitly accessing user space with the sie
instruction.
So we end up with:
User space:
- runs in primary or access register mode
- cr1 contains the user asce
- cr7 contains the user asce
- cr13 contains the kernel asce
Kernel space:
- runs in home space mode
- cr1 contains the user or kernel asce
-> the kernel asce is loaded when a uaccess requires primary or
secondary address mode
- cr7 contains the user or kernel asce, (changed with set_fs())
- cr13 contains the kernel asce
In case of uaccess the kernel changes to:
- primary space mode in case of a uaccess (copy_to_user) and uses
e.g. the mvcp instruction to access user space. However the kernel
will stay in home space mode if the mvcos instruction is available
- secondary space mode in case of futex atomic operations, so that the
instructions come from primary address space and data from secondary
space
In case of kvm the kernel runs in home space mode, but cr1 gets switched
to contain the gmap asce before the sie instruction gets executed. When
the sie instruction is finished cr1 will be switched back to contain the
user asce.
A context switch between two processes will always load the kernel asce
for the next process in cr1. So the first exit to user space is a bit
more expensive (one extra load control register instruction) than before,
however keeps the code rather simple.
In sum this means there is no need to perform any error prone page table
walks anymore when accessing user space.
The patch seems to be rather large, however it mainly removes the
the page table walk code and restores the previously deleted "standard"
uaccess code, with a couple of changes.
The uaccess without mvcos mode can be enforced with the "uaccess_primary"
kernel parameter.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2014-03-21 17:42:25 +08:00
|
|
|
static inline int user_space_fault(struct pt_regs *regs)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
s390/uaccess: rework uaccess code - fix locking issues
The current uaccess code uses a page table walk in some circumstances,
e.g. in case of the in atomic futex operations or if running on old
hardware which doesn't support the mvcos instruction.
However it turned out that the page table walk code does not correctly
lock page tables when accessing page table entries.
In other words: a different cpu may invalidate a page table entry while
the current cpu inspects the pte. This may lead to random data corruption.
Adding correct locking however isn't trivial for all uaccess operations.
Especially copy_in_user() is problematic since that requires to hold at
least two locks, but must be protected against ABBA deadlock when a
different cpu also performs a copy_in_user() operation.
So the solution is a different approach where we change address spaces:
User space runs in primary address mode, or access register mode within
vdso code, like it currently already does.
The kernel usually also runs in home space mode, however when accessing
user space the kernel switches to primary or secondary address mode if
the mvcos instruction is not available or if a compare-and-swap (futex)
instruction on a user space address is performed.
KVM however is special, since that requires the kernel to run in home
address space while implicitly accessing user space with the sie
instruction.
So we end up with:
User space:
- runs in primary or access register mode
- cr1 contains the user asce
- cr7 contains the user asce
- cr13 contains the kernel asce
Kernel space:
- runs in home space mode
- cr1 contains the user or kernel asce
-> the kernel asce is loaded when a uaccess requires primary or
secondary address mode
- cr7 contains the user or kernel asce, (changed with set_fs())
- cr13 contains the kernel asce
In case of uaccess the kernel changes to:
- primary space mode in case of a uaccess (copy_to_user) and uses
e.g. the mvcp instruction to access user space. However the kernel
will stay in home space mode if the mvcos instruction is available
- secondary space mode in case of futex atomic operations, so that the
instructions come from primary address space and data from secondary
space
In case of kvm the kernel runs in home space mode, but cr1 gets switched
to contain the gmap asce before the sie instruction gets executed. When
the sie instruction is finished cr1 will be switched back to contain the
user asce.
A context switch between two processes will always load the kernel asce
for the next process in cr1. So the first exit to user space is a bit
more expensive (one extra load control register instruction) than before,
however keeps the code rather simple.
In sum this means there is no need to perform any error prone page table
walks anymore when accessing user space.
The patch seems to be rather large, however it mainly removes the
the page table walk code and restores the previously deleted "standard"
uaccess code, with a couple of changes.
The uaccess without mvcos mode can be enforced with the "uaccess_primary"
kernel parameter.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2014-03-21 17:42:25 +08:00
|
|
|
unsigned long trans_exc_code;
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
2009-12-07 19:51:42 +08:00
|
|
|
* The lowest two bits of the translation exception
|
|
|
|
* identification indicate which paging table was used.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
s390/uaccess: rework uaccess code - fix locking issues
The current uaccess code uses a page table walk in some circumstances,
e.g. in case of the in atomic futex operations or if running on old
hardware which doesn't support the mvcos instruction.
However it turned out that the page table walk code does not correctly
lock page tables when accessing page table entries.
In other words: a different cpu may invalidate a page table entry while
the current cpu inspects the pte. This may lead to random data corruption.
Adding correct locking however isn't trivial for all uaccess operations.
Especially copy_in_user() is problematic since that requires to hold at
least two locks, but must be protected against ABBA deadlock when a
different cpu also performs a copy_in_user() operation.
So the solution is a different approach where we change address spaces:
User space runs in primary address mode, or access register mode within
vdso code, like it currently already does.
The kernel usually also runs in home space mode, however when accessing
user space the kernel switches to primary or secondary address mode if
the mvcos instruction is not available or if a compare-and-swap (futex)
instruction on a user space address is performed.
KVM however is special, since that requires the kernel to run in home
address space while implicitly accessing user space with the sie
instruction.
So we end up with:
User space:
- runs in primary or access register mode
- cr1 contains the user asce
- cr7 contains the user asce
- cr13 contains the kernel asce
Kernel space:
- runs in home space mode
- cr1 contains the user or kernel asce
-> the kernel asce is loaded when a uaccess requires primary or
secondary address mode
- cr7 contains the user or kernel asce, (changed with set_fs())
- cr13 contains the kernel asce
In case of uaccess the kernel changes to:
- primary space mode in case of a uaccess (copy_to_user) and uses
e.g. the mvcp instruction to access user space. However the kernel
will stay in home space mode if the mvcos instruction is available
- secondary space mode in case of futex atomic operations, so that the
instructions come from primary address space and data from secondary
space
In case of kvm the kernel runs in home space mode, but cr1 gets switched
to contain the gmap asce before the sie instruction gets executed. When
the sie instruction is finished cr1 will be switched back to contain the
user asce.
A context switch between two processes will always load the kernel asce
for the next process in cr1. So the first exit to user space is a bit
more expensive (one extra load control register instruction) than before,
however keeps the code rather simple.
In sum this means there is no need to perform any error prone page table
walks anymore when accessing user space.
The patch seems to be rather large, however it mainly removes the
the page table walk code and restores the previously deleted "standard"
uaccess code, with a couple of changes.
The uaccess without mvcos mode can be enforced with the "uaccess_primary"
kernel parameter.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2014-03-21 17:42:25 +08:00
|
|
|
trans_exc_code = regs->int_parm_long & 3;
|
|
|
|
if (trans_exc_code == 3) /* home space -> kernel */
|
|
|
|
return 0;
|
|
|
|
if (user_mode(regs))
|
|
|
|
return 1;
|
|
|
|
if (trans_exc_code == 2) /* secondary space -> set_fs */
|
2009-12-07 19:51:42 +08:00
|
|
|
return current->thread.mm_segment.ar4;
|
s390/uaccess: rework uaccess code - fix locking issues
The current uaccess code uses a page table walk in some circumstances,
e.g. in case of the in atomic futex operations or if running on old
hardware which doesn't support the mvcos instruction.
However it turned out that the page table walk code does not correctly
lock page tables when accessing page table entries.
In other words: a different cpu may invalidate a page table entry while
the current cpu inspects the pte. This may lead to random data corruption.
Adding correct locking however isn't trivial for all uaccess operations.
Especially copy_in_user() is problematic since that requires to hold at
least two locks, but must be protected against ABBA deadlock when a
different cpu also performs a copy_in_user() operation.
So the solution is a different approach where we change address spaces:
User space runs in primary address mode, or access register mode within
vdso code, like it currently already does.
The kernel usually also runs in home space mode, however when accessing
user space the kernel switches to primary or secondary address mode if
the mvcos instruction is not available or if a compare-and-swap (futex)
instruction on a user space address is performed.
KVM however is special, since that requires the kernel to run in home
address space while implicitly accessing user space with the sie
instruction.
So we end up with:
User space:
- runs in primary or access register mode
- cr1 contains the user asce
- cr7 contains the user asce
- cr13 contains the kernel asce
Kernel space:
- runs in home space mode
- cr1 contains the user or kernel asce
-> the kernel asce is loaded when a uaccess requires primary or
secondary address mode
- cr7 contains the user or kernel asce, (changed with set_fs())
- cr13 contains the kernel asce
In case of uaccess the kernel changes to:
- primary space mode in case of a uaccess (copy_to_user) and uses
e.g. the mvcp instruction to access user space. However the kernel
will stay in home space mode if the mvcos instruction is available
- secondary space mode in case of futex atomic operations, so that the
instructions come from primary address space and data from secondary
space
In case of kvm the kernel runs in home space mode, but cr1 gets switched
to contain the gmap asce before the sie instruction gets executed. When
the sie instruction is finished cr1 will be switched back to contain the
user asce.
A context switch between two processes will always load the kernel asce
for the next process in cr1. So the first exit to user space is a bit
more expensive (one extra load control register instruction) than before,
however keeps the code rather simple.
In sum this means there is no need to perform any error prone page table
walks anymore when accessing user space.
The patch seems to be rather large, however it mainly removes the
the page table walk code and restores the previously deleted "standard"
uaccess code, with a couple of changes.
The uaccess without mvcos mode can be enforced with the "uaccess_primary"
kernel parameter.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2014-03-21 17:42:25 +08:00
|
|
|
if (current->flags & PF_VCPU)
|
|
|
|
return 1;
|
|
|
|
return 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2014-04-07 16:20:40 +08:00
|
|
|
static int bad_address(void *p)
|
|
|
|
{
|
|
|
|
unsigned long dummy;
|
|
|
|
|
|
|
|
return probe_kernel_address((unsigned long *)p, dummy);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void dump_pagetable(unsigned long asce, unsigned long address)
|
|
|
|
{
|
|
|
|
unsigned long *table = __va(asce & PAGE_MASK);
|
|
|
|
|
|
|
|
pr_alert("AS:%016lx ", asce);
|
|
|
|
switch (asce & _ASCE_TYPE_MASK) {
|
|
|
|
case _ASCE_TYPE_REGION1:
|
|
|
|
table = table + ((address >> 53) & 0x7ff);
|
|
|
|
if (bad_address(table))
|
|
|
|
goto bad;
|
|
|
|
pr_cont("R1:%016lx ", *table);
|
|
|
|
if (*table & _REGION_ENTRY_INVALID)
|
|
|
|
goto out;
|
|
|
|
table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
|
|
|
|
/* fallthrough */
|
|
|
|
case _ASCE_TYPE_REGION2:
|
|
|
|
table = table + ((address >> 42) & 0x7ff);
|
|
|
|
if (bad_address(table))
|
|
|
|
goto bad;
|
|
|
|
pr_cont("R2:%016lx ", *table);
|
|
|
|
if (*table & _REGION_ENTRY_INVALID)
|
|
|
|
goto out;
|
|
|
|
table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
|
|
|
|
/* fallthrough */
|
|
|
|
case _ASCE_TYPE_REGION3:
|
|
|
|
table = table + ((address >> 31) & 0x7ff);
|
|
|
|
if (bad_address(table))
|
|
|
|
goto bad;
|
|
|
|
pr_cont("R3:%016lx ", *table);
|
|
|
|
if (*table & (_REGION_ENTRY_INVALID | _REGION3_ENTRY_LARGE))
|
|
|
|
goto out;
|
|
|
|
table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
|
|
|
|
/* fallthrough */
|
|
|
|
case _ASCE_TYPE_SEGMENT:
|
|
|
|
table = table + ((address >> 20) & 0x7ff);
|
|
|
|
if (bad_address(table))
|
|
|
|
goto bad;
|
2015-01-05 20:29:18 +08:00
|
|
|
pr_cont("S:%016lx ", *table);
|
2014-04-07 16:20:40 +08:00
|
|
|
if (*table & (_SEGMENT_ENTRY_INVALID | _SEGMENT_ENTRY_LARGE))
|
|
|
|
goto out;
|
|
|
|
table = (unsigned long *)(*table & _SEGMENT_ENTRY_ORIGIN);
|
|
|
|
}
|
|
|
|
table = table + ((address >> 12) & 0xff);
|
|
|
|
if (bad_address(table))
|
|
|
|
goto bad;
|
|
|
|
pr_cont("P:%016lx ", *table);
|
|
|
|
out:
|
|
|
|
pr_cont("\n");
|
|
|
|
return;
|
|
|
|
bad:
|
|
|
|
pr_cont("BAD\n");
|
|
|
|
}
|
|
|
|
|
|
|
|
static void dump_fault_info(struct pt_regs *regs)
|
|
|
|
{
|
|
|
|
unsigned long asce;
|
|
|
|
|
2016-02-24 21:27:46 +08:00
|
|
|
pr_alert("Failing address: %016lx TEID: %016lx\n",
|
|
|
|
regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long);
|
2014-04-07 16:20:40 +08:00
|
|
|
pr_alert("Fault in ");
|
|
|
|
switch (regs->int_parm_long & 3) {
|
|
|
|
case 3:
|
|
|
|
pr_cont("home space ");
|
|
|
|
break;
|
|
|
|
case 2:
|
|
|
|
pr_cont("secondary space ");
|
|
|
|
break;
|
|
|
|
case 1:
|
|
|
|
pr_cont("access register ");
|
|
|
|
break;
|
|
|
|
case 0:
|
|
|
|
pr_cont("primary space ");
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
pr_cont("mode while using ");
|
|
|
|
if (!user_space_fault(regs)) {
|
|
|
|
asce = S390_lowcore.kernel_asce;
|
|
|
|
pr_cont("kernel ");
|
|
|
|
}
|
|
|
|
#ifdef CONFIG_PGSTE
|
|
|
|
else if ((current->flags & PF_VCPU) && S390_lowcore.gmap) {
|
|
|
|
struct gmap *gmap = (struct gmap *)S390_lowcore.gmap;
|
|
|
|
asce = gmap->asce;
|
|
|
|
pr_cont("gmap ");
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
else {
|
|
|
|
asce = S390_lowcore.user_asce;
|
|
|
|
pr_cont("user ");
|
|
|
|
}
|
|
|
|
pr_cont("ASCE.\n");
|
|
|
|
dump_pagetable(asce, regs->int_parm_long & __FAIL_ADDR_MASK);
|
|
|
|
}
|
|
|
|
|
2016-02-24 21:27:46 +08:00
|
|
|
int show_unhandled_signals = 1;
|
|
|
|
|
|
|
|
void report_user_fault(struct pt_regs *regs, long signr, int is_mm_fault)
|
2010-05-17 16:00:21 +08:00
|
|
|
{
|
|
|
|
if ((task_pid_nr(current) > 1) && !show_unhandled_signals)
|
|
|
|
return;
|
|
|
|
if (!unhandled_signal(current, signr))
|
|
|
|
return;
|
|
|
|
if (!printk_ratelimit())
|
|
|
|
return;
|
2015-01-29 21:38:38 +08:00
|
|
|
printk(KERN_ALERT "User process fault: interruption code %04x ilc:%d ",
|
2014-11-19 20:31:08 +08:00
|
|
|
regs->int_code & 0xffff, regs->int_code >> 17);
|
2016-01-18 20:12:19 +08:00
|
|
|
print_vma_addr(KERN_CONT "in ", regs->psw.addr);
|
2011-12-27 18:27:18 +08:00
|
|
|
printk(KERN_CONT "\n");
|
2016-02-24 21:27:46 +08:00
|
|
|
if (is_mm_fault)
|
|
|
|
dump_fault_info(regs);
|
2010-05-17 16:00:21 +08:00
|
|
|
show_regs(regs);
|
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* Send SIGSEGV to task. This is an external routine
|
|
|
|
* to keep the stack usage of do_page_fault small.
|
|
|
|
*/
|
2011-12-27 18:27:18 +08:00
|
|
|
static noinline void do_sigsegv(struct pt_regs *regs, int si_code)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
struct siginfo si;
|
|
|
|
|
2016-02-24 21:27:46 +08:00
|
|
|
report_user_fault(regs, SIGSEGV, 1);
|
2005-04-17 06:20:36 +08:00
|
|
|
si.si_signo = SIGSEGV;
|
2016-05-23 21:35:51 +08:00
|
|
|
si.si_errno = 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
si.si_code = si_code;
|
2011-12-27 18:27:18 +08:00
|
|
|
si.si_addr = (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK);
|
2005-04-17 06:20:36 +08:00
|
|
|
force_sig_info(SIGSEGV, &si, current);
|
|
|
|
}
|
|
|
|
|
2011-12-27 18:27:18 +08:00
|
|
|
static noinline void do_no_context(struct pt_regs *regs)
|
2007-04-27 22:01:43 +08:00
|
|
|
{
|
|
|
|
const struct exception_table_entry *fixup;
|
|
|
|
|
|
|
|
/* Are we prepared to handle this kernel fault? */
|
2016-01-18 20:12:19 +08:00
|
|
|
fixup = search_exception_tables(regs->psw.addr);
|
2007-04-27 22:01:43 +08:00
|
|
|
if (fixup) {
|
2016-01-18 19:49:44 +08:00
|
|
|
regs->psw.addr = extable_fixup(fixup);
|
2007-04-27 22:01:43 +08:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Oops. The kernel tried to access some bad page. We'll have to
|
|
|
|
* terminate things with extreme prejudice.
|
|
|
|
*/
|
s390/uaccess: rework uaccess code - fix locking issues
The current uaccess code uses a page table walk in some circumstances,
e.g. in case of the in atomic futex operations or if running on old
hardware which doesn't support the mvcos instruction.
However it turned out that the page table walk code does not correctly
lock page tables when accessing page table entries.
In other words: a different cpu may invalidate a page table entry while
the current cpu inspects the pte. This may lead to random data corruption.
Adding correct locking however isn't trivial for all uaccess operations.
Especially copy_in_user() is problematic since that requires to hold at
least two locks, but must be protected against ABBA deadlock when a
different cpu also performs a copy_in_user() operation.
So the solution is a different approach where we change address spaces:
User space runs in primary address mode, or access register mode within
vdso code, like it currently already does.
The kernel usually also runs in home space mode, however when accessing
user space the kernel switches to primary or secondary address mode if
the mvcos instruction is not available or if a compare-and-swap (futex)
instruction on a user space address is performed.
KVM however is special, since that requires the kernel to run in home
address space while implicitly accessing user space with the sie
instruction.
So we end up with:
User space:
- runs in primary or access register mode
- cr1 contains the user asce
- cr7 contains the user asce
- cr13 contains the kernel asce
Kernel space:
- runs in home space mode
- cr1 contains the user or kernel asce
-> the kernel asce is loaded when a uaccess requires primary or
secondary address mode
- cr7 contains the user or kernel asce, (changed with set_fs())
- cr13 contains the kernel asce
In case of uaccess the kernel changes to:
- primary space mode in case of a uaccess (copy_to_user) and uses
e.g. the mvcp instruction to access user space. However the kernel
will stay in home space mode if the mvcos instruction is available
- secondary space mode in case of futex atomic operations, so that the
instructions come from primary address space and data from secondary
space
In case of kvm the kernel runs in home space mode, but cr1 gets switched
to contain the gmap asce before the sie instruction gets executed. When
the sie instruction is finished cr1 will be switched back to contain the
user asce.
A context switch between two processes will always load the kernel asce
for the next process in cr1. So the first exit to user space is a bit
more expensive (one extra load control register instruction) than before,
however keeps the code rather simple.
In sum this means there is no need to perform any error prone page table
walks anymore when accessing user space.
The patch seems to be rather large, however it mainly removes the
the page table walk code and restores the previously deleted "standard"
uaccess code, with a couple of changes.
The uaccess without mvcos mode can be enforced with the "uaccess_primary"
kernel parameter.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2014-03-21 17:42:25 +08:00
|
|
|
if (!user_space_fault(regs))
|
2007-04-27 22:01:43 +08:00
|
|
|
printk(KERN_ALERT "Unable to handle kernel pointer dereference"
|
2014-04-07 16:20:40 +08:00
|
|
|
" in virtual kernel address space\n");
|
2007-04-27 22:01:43 +08:00
|
|
|
else
|
|
|
|
printk(KERN_ALERT "Unable to handle kernel paging request"
|
2014-04-07 16:20:40 +08:00
|
|
|
" in virtual user address space\n");
|
|
|
|
dump_fault_info(regs);
|
2011-12-27 18:27:18 +08:00
|
|
|
die(regs, "Oops");
|
2007-04-27 22:01:43 +08:00
|
|
|
do_exit(SIGKILL);
|
|
|
|
}
|
|
|
|
|
2011-12-27 18:27:18 +08:00
|
|
|
static noinline void do_low_address(struct pt_regs *regs)
|
2007-04-27 22:01:43 +08:00
|
|
|
{
|
|
|
|
/* Low-address protection hit in kernel mode means
|
|
|
|
NULL pointer write access in kernel mode. */
|
|
|
|
if (regs->psw.mask & PSW_MASK_PSTATE) {
|
|
|
|
/* Low-address protection hit in user mode 'cannot happen'. */
|
2011-12-27 18:27:18 +08:00
|
|
|
die (regs, "Low-address protection");
|
2007-04-27 22:01:43 +08:00
|
|
|
do_exit(SIGKILL);
|
|
|
|
}
|
|
|
|
|
2011-12-27 18:27:18 +08:00
|
|
|
do_no_context(regs);
|
2007-04-27 22:01:43 +08:00
|
|
|
}
|
|
|
|
|
2011-12-27 18:27:18 +08:00
|
|
|
static noinline void do_sigbus(struct pt_regs *regs)
|
2007-04-27 22:01:43 +08:00
|
|
|
{
|
|
|
|
struct task_struct *tsk = current;
|
2010-10-25 22:10:35 +08:00
|
|
|
struct siginfo si;
|
2007-04-27 22:01:43 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Send a sigbus, regardless of whether we were in kernel
|
|
|
|
* or user mode.
|
|
|
|
*/
|
2010-10-25 22:10:35 +08:00
|
|
|
si.si_signo = SIGBUS;
|
|
|
|
si.si_errno = 0;
|
|
|
|
si.si_code = BUS_ADRERR;
|
2011-12-27 18:27:18 +08:00
|
|
|
si.si_addr = (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK);
|
2010-10-25 22:10:35 +08:00
|
|
|
force_sig_info(SIGBUS, &si, tsk);
|
2007-04-27 22:01:43 +08:00
|
|
|
}
|
|
|
|
|
2011-12-27 18:27:18 +08:00
|
|
|
static noinline void do_fault_error(struct pt_regs *regs, int fault)
|
2009-12-07 19:51:45 +08:00
|
|
|
{
|
|
|
|
int si_code;
|
|
|
|
|
|
|
|
switch (fault) {
|
|
|
|
case VM_FAULT_BADACCESS:
|
|
|
|
case VM_FAULT_BADMAP:
|
|
|
|
/* Bad memory access. Check if it is kernel or user space. */
|
2012-07-27 16:31:12 +08:00
|
|
|
if (user_mode(regs)) {
|
2009-12-07 19:51:45 +08:00
|
|
|
/* User mode accesses just cause a SIGSEGV */
|
|
|
|
si_code = (fault == VM_FAULT_BADMAP) ?
|
|
|
|
SEGV_MAPERR : SEGV_ACCERR;
|
2011-12-27 18:27:18 +08:00
|
|
|
do_sigsegv(regs, si_code);
|
2009-12-07 19:51:45 +08:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
case VM_FAULT_BADCONTEXT:
|
2013-06-17 22:25:18 +08:00
|
|
|
case VM_FAULT_PFAULT:
|
2011-12-27 18:27:18 +08:00
|
|
|
do_no_context(regs);
|
2009-12-07 19:51:45 +08:00
|
|
|
break;
|
2012-07-27 14:54:20 +08:00
|
|
|
case VM_FAULT_SIGNAL:
|
|
|
|
if (!user_mode(regs))
|
|
|
|
do_no_context(regs);
|
|
|
|
break;
|
2009-12-07 19:51:45 +08:00
|
|
|
default: /* fault & VM_FAULT_ERROR */
|
2011-05-26 15:48:29 +08:00
|
|
|
if (fault & VM_FAULT_OOM) {
|
2012-07-27 16:31:12 +08:00
|
|
|
if (!user_mode(regs))
|
2011-12-27 18:27:18 +08:00
|
|
|
do_no_context(regs);
|
2011-05-26 15:48:29 +08:00
|
|
|
else
|
|
|
|
pagefault_out_of_memory();
|
vm: add VM_FAULT_SIGSEGV handling support
The core VM already knows about VM_FAULT_SIGBUS, but cannot return a
"you should SIGSEGV" error, because the SIGSEGV case was generally
handled by the caller - usually the architecture fault handler.
That results in lots of duplication - all the architecture fault
handlers end up doing very similar "look up vma, check permissions, do
retries etc" - but it generally works. However, there are cases where
the VM actually wants to SIGSEGV, and applications _expect_ SIGSEGV.
In particular, when accessing the stack guard page, libsigsegv expects a
SIGSEGV. And it usually got one, because the stack growth is handled by
that duplicated architecture fault handler.
However, when the generic VM layer started propagating the error return
from the stack expansion in commit fee7e49d4514 ("mm: propagate error
from stack expansion even for guard page"), that now exposed the
existing VM_FAULT_SIGBUS result to user space. And user space really
expected SIGSEGV, not SIGBUS.
To fix that case, we need to add a VM_FAULT_SIGSEGV, and teach all those
duplicate architecture fault handlers about it. They all already have
the code to handle SIGSEGV, so it's about just tying that new return
value to the existing code, but it's all a bit annoying.
This is the mindless minimal patch to do this. A more extensive patch
would be to try to gather up the mostly shared fault handling logic into
one generic helper routine, and long-term we really should do that
cleanup.
Just from this patch, you can generally see that most architectures just
copied (directly or indirectly) the old x86 way of doing things, but in
the meantime that original x86 model has been improved to hold the VM
semaphore for shorter times etc and to handle VM_FAULT_RETRY and other
"newer" things, so it would be a good idea to bring all those
improvements to the generic case and teach other architectures about
them too.
Reported-and-tested-by: Takashi Iwai <tiwai@suse.de>
Tested-by: Jan Engelhardt <jengelh@inai.de>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> # "s390 still compiles and boots"
Cc: linux-arch@vger.kernel.org
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-01-30 02:51:32 +08:00
|
|
|
} else if (fault & VM_FAULT_SIGSEGV) {
|
|
|
|
/* Kernel mode? Handle exceptions or die */
|
|
|
|
if (!user_mode(regs))
|
|
|
|
do_no_context(regs);
|
|
|
|
else
|
|
|
|
do_sigsegv(regs, SEGV_MAPERR);
|
2011-05-26 15:48:29 +08:00
|
|
|
} else if (fault & VM_FAULT_SIGBUS) {
|
2009-12-07 19:51:45 +08:00
|
|
|
/* Kernel mode? Handle exceptions or die */
|
2012-07-27 16:31:12 +08:00
|
|
|
if (!user_mode(regs))
|
2011-12-27 18:27:18 +08:00
|
|
|
do_no_context(regs);
|
2010-10-25 22:10:35 +08:00
|
|
|
else
|
2011-12-27 18:27:18 +08:00
|
|
|
do_sigbus(regs);
|
2009-12-07 19:51:45 +08:00
|
|
|
} else
|
|
|
|
BUG();
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* This routine handles page faults. It determines the address,
|
|
|
|
* and the problem, and then passes it off to one of the appropriate
|
|
|
|
* routines.
|
|
|
|
*
|
2009-12-07 19:51:45 +08:00
|
|
|
* interruption code (int_code):
|
2005-04-17 06:20:36 +08:00
|
|
|
* 04 Protection -> Write-Protection (suprression)
|
|
|
|
* 10 Segment translation -> Not present (nullification)
|
|
|
|
* 11 Page translation -> Not present (nullification)
|
|
|
|
* 3b Region third trans. -> Not present (nullification)
|
|
|
|
*/
|
2011-12-27 18:27:18 +08:00
|
|
|
static inline int do_exception(struct pt_regs *regs, int access)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2013-06-17 22:25:18 +08:00
|
|
|
#ifdef CONFIG_PGSTE
|
|
|
|
struct gmap *gmap;
|
|
|
|
#endif
|
2007-04-27 22:01:43 +08:00
|
|
|
struct task_struct *tsk;
|
|
|
|
struct mm_struct *mm;
|
|
|
|
struct vm_area_struct *vma;
|
2011-12-27 18:27:18 +08:00
|
|
|
unsigned long trans_exc_code;
|
2007-04-27 22:01:43 +08:00
|
|
|
unsigned long address;
|
2011-05-26 15:48:30 +08:00
|
|
|
unsigned int flags;
|
|
|
|
int fault;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2012-11-21 23:36:27 +08:00
|
|
|
tsk = current;
|
|
|
|
/*
|
|
|
|
* The instruction that caused the program check has
|
|
|
|
* been nullified. Don't signal single step via SIGTRAP.
|
|
|
|
*/
|
2014-04-15 18:55:07 +08:00
|
|
|
clear_pt_regs_flag(regs, PIF_PER_TRAP);
|
2012-11-21 23:36:27 +08:00
|
|
|
|
2009-12-07 19:51:44 +08:00
|
|
|
if (notify_page_fault(regs))
|
2009-12-07 19:51:45 +08:00
|
|
|
return 0;
|
2006-09-20 21:58:39 +08:00
|
|
|
|
2007-04-27 22:01:43 +08:00
|
|
|
mm = tsk->mm;
|
2011-12-27 18:27:18 +08:00
|
|
|
trans_exc_code = regs->int_parm_long;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Verify that the fault happened in user space, that
|
|
|
|
* we are not in an interrupt and that there is a
|
|
|
|
* user context.
|
|
|
|
*/
|
2009-12-07 19:51:45 +08:00
|
|
|
fault = VM_FAULT_BADCONTEXT;
|
2015-05-11 23:52:11 +08:00
|
|
|
if (unlikely(!user_space_fault(regs) || faulthandler_disabled() || !mm))
|
2009-12-07 19:51:45 +08:00
|
|
|
goto out;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-12-07 19:51:42 +08:00
|
|
|
address = trans_exc_code & __FAIL_ADDR_MASK;
|
2011-06-27 20:41:57 +08:00
|
|
|
perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
|
2012-07-27 14:54:20 +08:00
|
|
|
flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
|
2013-09-13 06:13:39 +08:00
|
|
|
if (user_mode(regs))
|
|
|
|
flags |= FAULT_FLAG_USER;
|
2011-05-26 15:48:30 +08:00
|
|
|
if (access == VM_WRITE || (trans_exc_code & store_indication) == 0x400)
|
|
|
|
flags |= FAULT_FLAG_WRITE;
|
2007-04-27 22:01:43 +08:00
|
|
|
down_read(&mm->mmap_sem);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2011-07-24 16:48:20 +08:00
|
|
|
#ifdef CONFIG_PGSTE
|
2014-04-30 22:04:25 +08:00
|
|
|
gmap = (current->flags & PF_VCPU) ?
|
|
|
|
(struct gmap *) S390_lowcore.gmap : NULL;
|
2013-06-17 22:25:18 +08:00
|
|
|
if (gmap) {
|
2014-04-30 22:04:25 +08:00
|
|
|
current->thread.gmap_addr = address;
|
2016-03-08 19:12:18 +08:00
|
|
|
current->thread.gmap_write_flag = !!(flags & FAULT_FLAG_WRITE);
|
2016-03-08 19:31:52 +08:00
|
|
|
current->thread.gmap_int_code = regs->int_code & 0xffff;
|
2014-04-30 22:04:25 +08:00
|
|
|
address = __gmap_translate(gmap, address);
|
2011-07-24 16:48:20 +08:00
|
|
|
if (address == -EFAULT) {
|
|
|
|
fault = VM_FAULT_BADMAP;
|
|
|
|
goto out_up;
|
|
|
|
}
|
2013-06-17 22:25:18 +08:00
|
|
|
if (gmap->pfault_enabled)
|
|
|
|
flags |= FAULT_FLAG_RETRY_NOWAIT;
|
2011-07-24 16:48:20 +08:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
retry:
|
2009-12-07 19:51:45 +08:00
|
|
|
fault = VM_FAULT_BADMAP;
|
2007-03-06 06:35:54 +08:00
|
|
|
vma = find_vma(mm, address);
|
|
|
|
if (!vma)
|
2009-12-07 19:51:45 +08:00
|
|
|
goto out_up;
|
2007-02-06 04:18:17 +08:00
|
|
|
|
2009-12-07 19:51:45 +08:00
|
|
|
if (unlikely(vma->vm_start > address)) {
|
|
|
|
if (!(vma->vm_flags & VM_GROWSDOWN))
|
|
|
|
goto out_up;
|
|
|
|
if (expand_stack(vma, address))
|
|
|
|
goto out_up;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Ok, we have a good vm_area for this memory access, so
|
|
|
|
* we can handle it..
|
|
|
|
*/
|
|
|
|
fault = VM_FAULT_BADACCESS;
|
2009-12-07 19:51:46 +08:00
|
|
|
if (unlikely(!(vma->vm_flags & access)))
|
2009-12-07 19:51:45 +08:00
|
|
|
goto out_up;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-04-30 19:38:46 +08:00
|
|
|
if (is_vm_hugetlb_page(vma))
|
|
|
|
address &= HPAGE_MASK;
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* If for any reason at all we couldn't handle the fault,
|
|
|
|
* make sure we exit gracefully rather than endlessly redo
|
|
|
|
* the fault.
|
|
|
|
*/
|
2011-05-26 15:48:30 +08:00
|
|
|
fault = handle_mm_fault(mm, vma, address, flags);
|
2012-07-27 14:54:20 +08:00
|
|
|
/* No reason to continue if interrupted by SIGKILL. */
|
|
|
|
if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current)) {
|
|
|
|
fault = VM_FAULT_SIGNAL;
|
|
|
|
goto out;
|
|
|
|
}
|
2009-12-07 19:51:45 +08:00
|
|
|
if (unlikely(fault & VM_FAULT_ERROR))
|
|
|
|
goto out_up;
|
|
|
|
|
2011-05-26 15:48:30 +08:00
|
|
|
/*
|
|
|
|
* Major/minor page fault accounting is only done on the
|
|
|
|
* initial attempt. If we go through a retry, it is extremely
|
|
|
|
* likely that the page will be found in page cache at that point.
|
|
|
|
*/
|
|
|
|
if (flags & FAULT_FLAG_ALLOW_RETRY) {
|
|
|
|
if (fault & VM_FAULT_MAJOR) {
|
|
|
|
tsk->maj_flt++;
|
2011-06-27 20:41:57 +08:00
|
|
|
perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1,
|
2011-05-26 15:48:30 +08:00
|
|
|
regs, address);
|
|
|
|
} else {
|
|
|
|
tsk->min_flt++;
|
2011-06-27 20:41:57 +08:00
|
|
|
perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1,
|
2011-05-26 15:48:30 +08:00
|
|
|
regs, address);
|
|
|
|
}
|
|
|
|
if (fault & VM_FAULT_RETRY) {
|
2013-06-17 22:25:18 +08:00
|
|
|
#ifdef CONFIG_PGSTE
|
|
|
|
if (gmap && (flags & FAULT_FLAG_RETRY_NOWAIT)) {
|
|
|
|
/* FAULT_FLAG_RETRY_NOWAIT has been set,
|
|
|
|
* mmap_sem has not been released */
|
|
|
|
current->thread.gmap_pfault = 1;
|
|
|
|
fault = VM_FAULT_PFAULT;
|
|
|
|
goto out_up;
|
|
|
|
}
|
|
|
|
#endif
|
2011-05-26 15:48:30 +08:00
|
|
|
/* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk
|
|
|
|
* of starvation. */
|
2013-06-17 22:25:18 +08:00
|
|
|
flags &= ~(FAULT_FLAG_ALLOW_RETRY |
|
|
|
|
FAULT_FLAG_RETRY_NOWAIT);
|
2012-10-09 07:32:19 +08:00
|
|
|
flags |= FAULT_FLAG_TRIED;
|
2011-07-24 16:48:20 +08:00
|
|
|
down_read(&mm->mmap_sem);
|
2011-05-26 15:48:30 +08:00
|
|
|
goto retry;
|
|
|
|
}
|
2009-09-11 16:29:06 +08:00
|
|
|
}
|
2014-04-30 22:04:25 +08:00
|
|
|
#ifdef CONFIG_PGSTE
|
|
|
|
if (gmap) {
|
|
|
|
address = __gmap_link(gmap, current->thread.gmap_addr,
|
|
|
|
address);
|
|
|
|
if (address == -EFAULT) {
|
|
|
|
fault = VM_FAULT_BADMAP;
|
|
|
|
goto out_up;
|
|
|
|
}
|
|
|
|
if (address == -ENOMEM) {
|
|
|
|
fault = VM_FAULT_OOM;
|
|
|
|
goto out_up;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
2009-12-07 19:51:45 +08:00
|
|
|
fault = 0;
|
|
|
|
out_up:
|
2007-04-27 22:01:43 +08:00
|
|
|
up_read(&mm->mmap_sem);
|
2009-12-07 19:51:45 +08:00
|
|
|
out:
|
|
|
|
return fault;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2014-10-22 18:42:38 +08:00
|
|
|
void do_protection_exception(struct pt_regs *regs)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2011-12-27 18:27:18 +08:00
|
|
|
unsigned long trans_exc_code;
|
2009-12-07 19:51:45 +08:00
|
|
|
int fault;
|
2009-12-07 19:51:42 +08:00
|
|
|
|
2011-12-27 18:27:18 +08:00
|
|
|
trans_exc_code = regs->int_parm_long;
|
2013-04-16 19:25:06 +08:00
|
|
|
/*
|
|
|
|
* Protection exceptions are suppressing, decrement psw address.
|
|
|
|
* The exception to this rule are aborted transactions, for these
|
|
|
|
* the PSW already points to the correct location.
|
|
|
|
*/
|
|
|
|
if (!(regs->int_code & 0x200))
|
|
|
|
regs->psw.addr = __rewind_psw(regs->psw, regs->int_code >> 16);
|
2007-04-27 22:01:43 +08:00
|
|
|
/*
|
|
|
|
* Check for low-address protection. This needs to be treated
|
|
|
|
* as a special case because the translation exception code
|
|
|
|
* field is not guaranteed to contain valid data in this case.
|
|
|
|
*/
|
2009-12-07 19:51:42 +08:00
|
|
|
if (unlikely(!(trans_exc_code & 4))) {
|
2011-12-27 18:27:18 +08:00
|
|
|
do_low_address(regs);
|
2007-04-27 22:01:43 +08:00
|
|
|
return;
|
|
|
|
}
|
2011-12-27 18:27:18 +08:00
|
|
|
fault = do_exception(regs, VM_WRITE);
|
2009-12-07 19:51:45 +08:00
|
|
|
if (unlikely(fault))
|
2011-12-27 18:27:18 +08:00
|
|
|
do_fault_error(regs, fault);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2014-10-22 18:42:38 +08:00
|
|
|
NOKPROBE_SYMBOL(do_protection_exception);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2014-10-22 18:42:38 +08:00
|
|
|
void do_dat_exception(struct pt_regs *regs)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2009-12-07 19:51:46 +08:00
|
|
|
int access, fault;
|
2009-12-07 19:51:45 +08:00
|
|
|
|
2009-12-07 19:51:46 +08:00
|
|
|
access = VM_READ | VM_EXEC | VM_WRITE;
|
2011-12-27 18:27:18 +08:00
|
|
|
fault = do_exception(regs, access);
|
2009-12-07 19:51:45 +08:00
|
|
|
if (unlikely(fault))
|
2011-12-27 18:27:18 +08:00
|
|
|
do_fault_error(regs, fault);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2014-10-22 18:42:38 +08:00
|
|
|
NOKPROBE_SYMBOL(do_dat_exception);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
#ifdef CONFIG_PFAULT
|
|
|
|
/*
|
|
|
|
* 'pfault' pseudo page faults routines.
|
|
|
|
*/
|
2011-01-05 19:47:39 +08:00
|
|
|
static int pfault_disable;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
static int __init nopfault(char *str)
|
|
|
|
{
|
|
|
|
pfault_disable = 1;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
__setup("nopfault", nopfault);
|
|
|
|
|
2011-05-23 16:24:35 +08:00
|
|
|
struct pfault_refbk {
|
|
|
|
u16 refdiagc;
|
|
|
|
u16 reffcode;
|
|
|
|
u16 refdwlen;
|
|
|
|
u16 refversn;
|
|
|
|
u64 refgaddr;
|
|
|
|
u64 refselmk;
|
|
|
|
u64 refcmpmk;
|
|
|
|
u64 reserved;
|
|
|
|
} __attribute__ ((packed, aligned(8)));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
int pfault_init(void)
|
|
|
|
{
|
2011-05-23 16:24:35 +08:00
|
|
|
struct pfault_refbk refbk = {
|
|
|
|
.refdiagc = 0x258,
|
|
|
|
.reffcode = 0,
|
|
|
|
.refdwlen = 5,
|
|
|
|
.refversn = 2,
|
2015-10-07 00:06:15 +08:00
|
|
|
.refgaddr = __LC_LPP,
|
2011-05-23 16:24:35 +08:00
|
|
|
.refselmk = 1ULL << 48,
|
|
|
|
.refcmpmk = 1ULL << 48,
|
|
|
|
.reserved = __PF_RES_FIELD };
|
2005-04-17 06:20:36 +08:00
|
|
|
int rc;
|
|
|
|
|
2011-12-27 18:27:11 +08:00
|
|
|
if (pfault_disable)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -1;
|
2015-08-20 23:28:44 +08:00
|
|
|
diag_stat_inc(DIAG_STAT_X258);
|
2006-09-28 22:56:43 +08:00
|
|
|
asm volatile(
|
|
|
|
" diag %1,%0,0x258\n"
|
|
|
|
"0: j 2f\n"
|
|
|
|
"1: la %0,8\n"
|
2005-04-17 06:20:36 +08:00
|
|
|
"2:\n"
|
2006-09-28 22:56:43 +08:00
|
|
|
EX_TABLE(0b,1b)
|
|
|
|
: "=d" (rc) : "a" (&refbk), "m" (refbk) : "cc");
|
2005-04-17 06:20:36 +08:00
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
|
|
|
void pfault_fini(void)
|
|
|
|
{
|
2011-05-23 16:24:35 +08:00
|
|
|
struct pfault_refbk refbk = {
|
|
|
|
.refdiagc = 0x258,
|
|
|
|
.reffcode = 1,
|
|
|
|
.refdwlen = 5,
|
|
|
|
.refversn = 2,
|
|
|
|
};
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2011-12-27 18:27:11 +08:00
|
|
|
if (pfault_disable)
|
2005-04-17 06:20:36 +08:00
|
|
|
return;
|
2015-08-20 23:28:44 +08:00
|
|
|
diag_stat_inc(DIAG_STAT_X258);
|
2006-09-28 22:56:43 +08:00
|
|
|
asm volatile(
|
|
|
|
" diag %0,0,0x258\n"
|
2005-04-17 06:20:36 +08:00
|
|
|
"0:\n"
|
2006-09-28 22:56:43 +08:00
|
|
|
EX_TABLE(0b,0b)
|
|
|
|
: : "a" (&refbk), "m" (refbk) : "cc");
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2011-05-23 16:24:34 +08:00
|
|
|
static DEFINE_SPINLOCK(pfault_lock);
|
|
|
|
static LIST_HEAD(pfault_list);
|
|
|
|
|
2016-03-23 04:42:53 +08:00
|
|
|
#define PF_COMPLETE 0x0080
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The mechanism of our pfault code: if Linux is running as guest, runs a user
|
|
|
|
* space process and the user space process accesses a page that the host has
|
|
|
|
* paged out we get a pfault interrupt.
|
|
|
|
*
|
|
|
|
* This allows us, within the guest, to schedule a different process. Without
|
|
|
|
* this mechanism the host would have to suspend the whole virtual cpu until
|
|
|
|
* the page has been paged in.
|
|
|
|
*
|
|
|
|
* So when we get such an interrupt then we set the state of the current task
|
|
|
|
* to uninterruptible and also set the need_resched flag. Both happens within
|
|
|
|
* interrupt context(!). If we later on want to return to user space we
|
|
|
|
* recognize the need_resched flag and then call schedule(). It's not very
|
|
|
|
* obvious how this works...
|
|
|
|
*
|
|
|
|
* Of course we have a lot of additional fun with the completion interrupt (->
|
|
|
|
* host signals that a page of a process has been paged in and the process can
|
|
|
|
* continue to run). This interrupt can arrive on any cpu and, since we have
|
|
|
|
* virtual cpus, actually appear before the interrupt that signals that a page
|
|
|
|
* is missing.
|
|
|
|
*/
|
2012-03-11 23:59:31 +08:00
|
|
|
static void pfault_interrupt(struct ext_code ext_code,
|
2010-10-25 22:10:38 +08:00
|
|
|
unsigned int param32, unsigned long param64)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
struct task_struct *tsk;
|
|
|
|
__u16 subcode;
|
2011-05-23 16:24:34 +08:00
|
|
|
pid_t pid;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/*
|
2016-03-23 04:42:53 +08:00
|
|
|
* Get the external interruption subcode & pfault initial/completion
|
|
|
|
* signal bit. VM stores this in the 'cpu address' field associated
|
|
|
|
* with the external interrupt.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2012-03-11 23:59:31 +08:00
|
|
|
subcode = ext_code.subcode;
|
2005-04-17 06:20:36 +08:00
|
|
|
if ((subcode & 0xff00) != __SUBCODE_MASK)
|
|
|
|
return;
|
2013-01-02 22:18:18 +08:00
|
|
|
inc_irq_stat(IRQEXT_PFL);
|
2012-05-10 15:44:35 +08:00
|
|
|
/* Get the token (= pid of the affected task). */
|
2015-10-07 00:06:15 +08:00
|
|
|
pid = param64 & LPP_PFAULT_PID_MASK;
|
2012-05-10 15:44:35 +08:00
|
|
|
rcu_read_lock();
|
|
|
|
tsk = find_task_by_pid_ns(pid, &init_pid_ns);
|
|
|
|
if (tsk)
|
|
|
|
get_task_struct(tsk);
|
|
|
|
rcu_read_unlock();
|
|
|
|
if (!tsk)
|
|
|
|
return;
|
2011-05-23 16:24:34 +08:00
|
|
|
spin_lock(&pfault_lock);
|
2016-03-23 04:42:53 +08:00
|
|
|
if (subcode & PF_COMPLETE) {
|
2005-04-17 06:20:36 +08:00
|
|
|
/* signal bit is set -> a page has been swapped in by VM */
|
2011-05-23 16:24:34 +08:00
|
|
|
if (tsk->thread.pfault_wait == 1) {
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Initial interrupt was faster than the completion
|
|
|
|
* interrupt. pfault_wait is valid. Set pfault_wait
|
|
|
|
* back to zero and wake up the process. This can
|
|
|
|
* safely be done because the task is still sleeping
|
2005-09-04 06:58:02 +08:00
|
|
|
* and can't produce new pfaults. */
|
2005-04-17 06:20:36 +08:00
|
|
|
tsk->thread.pfault_wait = 0;
|
2011-05-23 16:24:34 +08:00
|
|
|
list_del(&tsk->thread.list);
|
2005-04-17 06:20:36 +08:00
|
|
|
wake_up_process(tsk);
|
2012-05-09 15:37:30 +08:00
|
|
|
put_task_struct(tsk);
|
2011-05-23 16:24:34 +08:00
|
|
|
} else {
|
|
|
|
/* Completion interrupt was faster than initial
|
|
|
|
* interrupt. Set pfault_wait to -1 so the initial
|
2011-11-14 18:19:01 +08:00
|
|
|
* interrupt doesn't put the task to sleep.
|
|
|
|
* If the task is not running, ignore the completion
|
|
|
|
* interrupt since it must be a leftover of a PFAULT
|
|
|
|
* CANCEL operation which didn't remove all pending
|
|
|
|
* completion interrupts. */
|
|
|
|
if (tsk->state == TASK_RUNNING)
|
|
|
|
tsk->thread.pfault_wait = -1;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
/* signal bit not set -> a real page is missing. */
|
2012-05-10 16:47:21 +08:00
|
|
|
if (WARN_ON_ONCE(tsk != current))
|
|
|
|
goto out;
|
2012-05-09 15:37:30 +08:00
|
|
|
if (tsk->thread.pfault_wait == 1) {
|
|
|
|
/* Already on the list with a reference: put to sleep */
|
2016-03-23 04:42:53 +08:00
|
|
|
goto block;
|
2012-05-09 15:37:30 +08:00
|
|
|
} else if (tsk->thread.pfault_wait == -1) {
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Completion interrupt was faster than the initial
|
2011-05-23 16:24:34 +08:00
|
|
|
* interrupt (pfault_wait == -1). Set pfault_wait
|
|
|
|
* back to zero and exit. */
|
2005-04-17 06:20:36 +08:00
|
|
|
tsk->thread.pfault_wait = 0;
|
2011-05-23 16:24:34 +08:00
|
|
|
} else {
|
|
|
|
/* Initial interrupt arrived before completion
|
2012-05-09 15:37:30 +08:00
|
|
|
* interrupt. Let the task sleep.
|
|
|
|
* An extra task reference is needed since a different
|
|
|
|
* cpu may set the task state to TASK_RUNNING again
|
|
|
|
* before the scheduler is reached. */
|
|
|
|
get_task_struct(tsk);
|
2011-05-23 16:24:34 +08:00
|
|
|
tsk->thread.pfault_wait = 1;
|
|
|
|
list_add(&tsk->thread.list, &pfault_list);
|
2016-03-23 04:42:53 +08:00
|
|
|
block:
|
|
|
|
/* Since this must be a userspace fault, there
|
|
|
|
* is no kernel task state to trample. Rely on the
|
|
|
|
* return to userspace schedule() to block. */
|
|
|
|
__set_current_state(TASK_UNINTERRUPTIBLE);
|
2005-04-17 06:20:36 +08:00
|
|
|
set_tsk_need_resched(tsk);
|
2011-05-23 16:24:34 +08:00
|
|
|
}
|
|
|
|
}
|
2012-05-10 16:47:21 +08:00
|
|
|
out:
|
2011-05-23 16:24:34 +08:00
|
|
|
spin_unlock(&pfault_lock);
|
2012-05-10 15:44:35 +08:00
|
|
|
put_task_struct(tsk);
|
2011-05-23 16:24:34 +08:00
|
|
|
}
|
|
|
|
|
2013-06-19 05:04:52 +08:00
|
|
|
static int pfault_cpu_notify(struct notifier_block *self, unsigned long action,
|
|
|
|
void *hcpu)
|
2011-05-23 16:24:34 +08:00
|
|
|
{
|
|
|
|
struct thread_struct *thread, *next;
|
|
|
|
struct task_struct *tsk;
|
|
|
|
|
2012-08-27 21:43:49 +08:00
|
|
|
switch (action & ~CPU_TASKS_FROZEN) {
|
2011-05-23 16:24:34 +08:00
|
|
|
case CPU_DEAD:
|
|
|
|
spin_lock_irq(&pfault_lock);
|
|
|
|
list_for_each_entry_safe(thread, next, &pfault_list, list) {
|
|
|
|
thread->pfault_wait = 0;
|
|
|
|
list_del(&thread->list);
|
|
|
|
tsk = container_of(thread, struct task_struct, thread);
|
|
|
|
wake_up_process(tsk);
|
2012-05-09 15:37:30 +08:00
|
|
|
put_task_struct(tsk);
|
2011-05-23 16:24:34 +08:00
|
|
|
}
|
|
|
|
spin_unlock_irq(&pfault_lock);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
break;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2011-05-23 16:24:34 +08:00
|
|
|
return NOTIFY_OK;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2011-01-05 19:47:39 +08:00
|
|
|
static int __init pfault_irq_init(void)
|
2006-12-04 22:40:40 +08:00
|
|
|
{
|
2011-01-05 19:47:39 +08:00
|
|
|
int rc;
|
2006-12-04 22:40:40 +08:00
|
|
|
|
2014-03-31 21:24:08 +08:00
|
|
|
rc = register_external_irq(EXT_IRQ_CP_SERVICE, pfault_interrupt);
|
2011-05-23 16:24:35 +08:00
|
|
|
if (rc)
|
|
|
|
goto out_extint;
|
|
|
|
rc = pfault_init() == 0 ? 0 : -EOPNOTSUPP;
|
|
|
|
if (rc)
|
|
|
|
goto out_pfault;
|
2013-09-04 19:35:45 +08:00
|
|
|
irq_subclass_register(IRQ_SUBCLASS_SERVICE_SIGNAL);
|
2011-05-23 16:24:35 +08:00
|
|
|
hotcpu_notifier(pfault_cpu_notify, 0);
|
|
|
|
return 0;
|
2006-12-04 22:40:40 +08:00
|
|
|
|
2011-05-23 16:24:35 +08:00
|
|
|
out_pfault:
|
2014-03-31 21:24:08 +08:00
|
|
|
unregister_external_irq(EXT_IRQ_CP_SERVICE, pfault_interrupt);
|
2011-05-23 16:24:35 +08:00
|
|
|
out_extint:
|
|
|
|
pfault_disable = 1;
|
|
|
|
return rc;
|
2006-12-04 22:40:40 +08:00
|
|
|
}
|
2011-01-05 19:47:39 +08:00
|
|
|
early_initcall(pfault_irq_init);
|
|
|
|
|
2011-05-23 16:24:35 +08:00
|
|
|
#endif /* CONFIG_PFAULT */
|