2009-06-10 21:20:19 +08:00
|
|
|
/**************************************************************************
|
|
|
|
*
|
|
|
|
* Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
|
|
|
|
* All Rights Reserved.
|
|
|
|
*
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a
|
|
|
|
* copy of this software and associated documentation files (the
|
|
|
|
* "Software"), to deal in the Software without restriction, including
|
|
|
|
* without limitation the rights to use, copy, modify, merge, publish,
|
|
|
|
* distribute, sub license, and/or sell copies of the Software, and to
|
|
|
|
* permit persons to whom the Software is furnished to do so, subject to
|
|
|
|
* the following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice (including the
|
|
|
|
* next paragraph) shall be included in all copies or substantial portions
|
|
|
|
* of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
|
|
|
|
* THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
|
|
|
|
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
|
|
|
|
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
|
|
|
|
* USE OR OTHER DEALINGS IN THE SOFTWARE.
|
|
|
|
*
|
|
|
|
**************************************************************************/
|
|
|
|
|
2012-03-17 12:43:50 +08:00
|
|
|
#define pr_fmt(fmt) "[TTM] " fmt
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
#include "ttm/ttm_memory.h"
|
2009-08-17 22:28:39 +08:00
|
|
|
#include "ttm/ttm_module.h"
|
2010-04-01 20:44:57 +08:00
|
|
|
#include "ttm/ttm_page_alloc.h"
|
2009-06-10 21:20:19 +08:00
|
|
|
#include <linux/spinlock.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/wait.h>
|
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/module.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 16:04:11 +08:00
|
|
|
#include <linux/slab.h>
|
2009-06-10 21:20:19 +08:00
|
|
|
|
|
|
|
#define TTM_MEMORY_ALLOC_RETRIES 4
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
struct ttm_mem_zone {
|
|
|
|
struct kobject kobj;
|
|
|
|
struct ttm_mem_global *glob;
|
|
|
|
const char *name;
|
|
|
|
uint64_t zone_mem;
|
|
|
|
uint64_t emer_mem;
|
|
|
|
uint64_t max_mem;
|
|
|
|
uint64_t swap_limit;
|
|
|
|
uint64_t used_mem;
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct attribute ttm_mem_sys = {
|
|
|
|
.name = "zone_memory",
|
|
|
|
.mode = S_IRUGO
|
|
|
|
};
|
|
|
|
static struct attribute ttm_mem_emer = {
|
|
|
|
.name = "emergency_memory",
|
|
|
|
.mode = S_IRUGO | S_IWUSR
|
|
|
|
};
|
|
|
|
static struct attribute ttm_mem_max = {
|
|
|
|
.name = "available_memory",
|
|
|
|
.mode = S_IRUGO | S_IWUSR
|
|
|
|
};
|
|
|
|
static struct attribute ttm_mem_swap = {
|
|
|
|
.name = "swap_limit",
|
|
|
|
.mode = S_IRUGO | S_IWUSR
|
|
|
|
};
|
|
|
|
static struct attribute ttm_mem_used = {
|
|
|
|
.name = "used_memory",
|
|
|
|
.mode = S_IRUGO
|
|
|
|
};
|
|
|
|
|
|
|
|
static void ttm_mem_zone_kobj_release(struct kobject *kobj)
|
|
|
|
{
|
|
|
|
struct ttm_mem_zone *zone =
|
|
|
|
container_of(kobj, struct ttm_mem_zone, kobj);
|
|
|
|
|
2012-03-17 12:43:50 +08:00
|
|
|
pr_info("Zone %7s: Used memory at exit: %llu kiB\n",
|
|
|
|
zone->name, (unsigned long long)zone->used_mem >> 10);
|
2009-08-17 22:28:39 +08:00
|
|
|
kfree(zone);
|
|
|
|
}
|
|
|
|
|
|
|
|
static ssize_t ttm_mem_zone_show(struct kobject *kobj,
|
|
|
|
struct attribute *attr,
|
|
|
|
char *buffer)
|
|
|
|
{
|
|
|
|
struct ttm_mem_zone *zone =
|
|
|
|
container_of(kobj, struct ttm_mem_zone, kobj);
|
|
|
|
uint64_t val = 0;
|
|
|
|
|
|
|
|
spin_lock(&zone->glob->lock);
|
|
|
|
if (attr == &ttm_mem_sys)
|
|
|
|
val = zone->zone_mem;
|
|
|
|
else if (attr == &ttm_mem_emer)
|
|
|
|
val = zone->emer_mem;
|
|
|
|
else if (attr == &ttm_mem_max)
|
|
|
|
val = zone->max_mem;
|
|
|
|
else if (attr == &ttm_mem_swap)
|
|
|
|
val = zone->swap_limit;
|
|
|
|
else if (attr == &ttm_mem_used)
|
|
|
|
val = zone->used_mem;
|
|
|
|
spin_unlock(&zone->glob->lock);
|
|
|
|
|
|
|
|
return snprintf(buffer, PAGE_SIZE, "%llu\n",
|
|
|
|
(unsigned long long) val >> 10);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ttm_check_swapping(struct ttm_mem_global *glob);
|
|
|
|
|
|
|
|
static ssize_t ttm_mem_zone_store(struct kobject *kobj,
|
|
|
|
struct attribute *attr,
|
|
|
|
const char *buffer,
|
|
|
|
size_t size)
|
|
|
|
{
|
|
|
|
struct ttm_mem_zone *zone =
|
|
|
|
container_of(kobj, struct ttm_mem_zone, kobj);
|
|
|
|
int chars;
|
|
|
|
unsigned long val;
|
|
|
|
uint64_t val64;
|
|
|
|
|
|
|
|
chars = sscanf(buffer, "%lu", &val);
|
|
|
|
if (chars == 0)
|
|
|
|
return size;
|
|
|
|
|
|
|
|
val64 = val;
|
|
|
|
val64 <<= 10;
|
|
|
|
|
|
|
|
spin_lock(&zone->glob->lock);
|
|
|
|
if (val64 > zone->zone_mem)
|
|
|
|
val64 = zone->zone_mem;
|
|
|
|
if (attr == &ttm_mem_emer) {
|
|
|
|
zone->emer_mem = val64;
|
|
|
|
if (zone->max_mem > val64)
|
|
|
|
zone->max_mem = val64;
|
|
|
|
} else if (attr == &ttm_mem_max) {
|
|
|
|
zone->max_mem = val64;
|
|
|
|
if (zone->emer_mem < val64)
|
|
|
|
zone->emer_mem = val64;
|
|
|
|
} else if (attr == &ttm_mem_swap)
|
|
|
|
zone->swap_limit = val64;
|
|
|
|
spin_unlock(&zone->glob->lock);
|
|
|
|
|
|
|
|
ttm_check_swapping(zone->glob);
|
|
|
|
|
|
|
|
return size;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct attribute *ttm_mem_zone_attrs[] = {
|
|
|
|
&ttm_mem_sys,
|
|
|
|
&ttm_mem_emer,
|
|
|
|
&ttm_mem_max,
|
|
|
|
&ttm_mem_swap,
|
|
|
|
&ttm_mem_used,
|
|
|
|
NULL
|
|
|
|
};
|
|
|
|
|
2010-01-19 09:58:23 +08:00
|
|
|
static const struct sysfs_ops ttm_mem_zone_ops = {
|
2009-08-17 22:28:39 +08:00
|
|
|
.show = &ttm_mem_zone_show,
|
|
|
|
.store = &ttm_mem_zone_store
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct kobj_type ttm_mem_zone_kobj_type = {
|
|
|
|
.release = &ttm_mem_zone_kobj_release,
|
|
|
|
.sysfs_ops = &ttm_mem_zone_ops,
|
|
|
|
.default_attrs = ttm_mem_zone_attrs,
|
|
|
|
};
|
|
|
|
|
|
|
|
static void ttm_mem_global_kobj_release(struct kobject *kobj)
|
|
|
|
{
|
|
|
|
struct ttm_mem_global *glob =
|
|
|
|
container_of(kobj, struct ttm_mem_global, kobj);
|
|
|
|
|
|
|
|
kfree(glob);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct kobj_type ttm_mem_glob_kobj_type = {
|
|
|
|
.release = &ttm_mem_global_kobj_release,
|
|
|
|
};
|
|
|
|
|
|
|
|
static bool ttm_zones_above_swap_target(struct ttm_mem_global *glob,
|
|
|
|
bool from_wq, uint64_t extra)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
struct ttm_mem_zone *zone;
|
|
|
|
uint64_t target;
|
|
|
|
|
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
|
|
|
|
|
|
|
if (from_wq)
|
|
|
|
target = zone->swap_limit;
|
|
|
|
else if (capable(CAP_SYS_ADMIN))
|
|
|
|
target = zone->emer_mem;
|
|
|
|
else
|
|
|
|
target = zone->max_mem;
|
|
|
|
|
|
|
|
target = (extra > target) ? 0ULL : target;
|
|
|
|
|
|
|
|
if (zone->used_mem > target)
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
/**
|
|
|
|
* At this point we only support a single shrink callback.
|
|
|
|
* Extend this if needed, perhaps using a linked list of callbacks.
|
|
|
|
* Note that this function is reentrant:
|
|
|
|
* many threads may try to swap out at any given time.
|
|
|
|
*/
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
static void ttm_shrink(struct ttm_mem_global *glob, bool from_wq,
|
2009-06-10 21:20:19 +08:00
|
|
|
uint64_t extra)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
struct ttm_mem_shrink *shrink;
|
|
|
|
|
|
|
|
spin_lock(&glob->lock);
|
|
|
|
if (glob->shrink == NULL)
|
|
|
|
goto out;
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
while (ttm_zones_above_swap_target(glob, from_wq, extra)) {
|
2009-06-10 21:20:19 +08:00
|
|
|
shrink = glob->shrink;
|
|
|
|
spin_unlock(&glob->lock);
|
|
|
|
ret = shrink->do_shrink(shrink);
|
|
|
|
spin_lock(&glob->lock);
|
|
|
|
if (unlikely(ret != 0))
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
out:
|
|
|
|
spin_unlock(&glob->lock);
|
|
|
|
}
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
static void ttm_shrink_work(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct ttm_mem_global *glob =
|
|
|
|
container_of(work, struct ttm_mem_global, work);
|
|
|
|
|
|
|
|
ttm_shrink(glob, true, 0ULL);
|
|
|
|
}
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
static int ttm_mem_init_kernel_zone(struct ttm_mem_global *glob,
|
|
|
|
const struct sysinfo *si)
|
|
|
|
{
|
|
|
|
struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL);
|
|
|
|
uint64_t mem;
|
2009-08-20 16:29:09 +08:00
|
|
|
int ret;
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
if (unlikely(!zone))
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
mem = si->totalram - si->totalhigh;
|
|
|
|
mem *= si->mem_unit;
|
|
|
|
|
|
|
|
zone->name = "kernel";
|
|
|
|
zone->zone_mem = mem;
|
|
|
|
zone->max_mem = mem >> 1;
|
|
|
|
zone->emer_mem = (mem >> 1) + (mem >> 2);
|
|
|
|
zone->swap_limit = zone->max_mem - (mem >> 3);
|
|
|
|
zone->used_mem = 0;
|
|
|
|
zone->glob = glob;
|
|
|
|
glob->zone_kernel = zone;
|
2010-03-13 18:36:32 +08:00
|
|
|
ret = kobject_init_and_add(
|
|
|
|
&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, zone->name);
|
2009-08-20 16:29:09 +08:00
|
|
|
if (unlikely(ret != 0)) {
|
|
|
|
kobject_put(&zone->kobj);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
glob->zones[glob->num_zones++] = zone;
|
|
|
|
return 0;
|
2009-08-17 22:28:39 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_HIGHMEM
|
|
|
|
static int ttm_mem_init_highmem_zone(struct ttm_mem_global *glob,
|
|
|
|
const struct sysinfo *si)
|
|
|
|
{
|
2009-11-28 18:30:32 +08:00
|
|
|
struct ttm_mem_zone *zone;
|
2009-08-17 22:28:39 +08:00
|
|
|
uint64_t mem;
|
2009-08-20 16:29:09 +08:00
|
|
|
int ret;
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
if (si->totalhigh == 0)
|
|
|
|
return 0;
|
|
|
|
|
2009-11-28 18:30:32 +08:00
|
|
|
zone = kzalloc(sizeof(*zone), GFP_KERNEL);
|
|
|
|
if (unlikely(!zone))
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
mem = si->totalram;
|
|
|
|
mem *= si->mem_unit;
|
|
|
|
|
|
|
|
zone->name = "highmem";
|
|
|
|
zone->zone_mem = mem;
|
|
|
|
zone->max_mem = mem >> 1;
|
|
|
|
zone->emer_mem = (mem >> 1) + (mem >> 2);
|
|
|
|
zone->swap_limit = zone->max_mem - (mem >> 3);
|
|
|
|
zone->used_mem = 0;
|
|
|
|
zone->glob = glob;
|
|
|
|
glob->zone_highmem = zone;
|
2010-03-13 18:36:32 +08:00
|
|
|
ret = kobject_init_and_add(
|
|
|
|
&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, zone->name);
|
2009-08-20 16:29:09 +08:00
|
|
|
if (unlikely(ret != 0)) {
|
|
|
|
kobject_put(&zone->kobj);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
glob->zones[glob->num_zones++] = zone;
|
|
|
|
return 0;
|
2009-08-17 22:28:39 +08:00
|
|
|
}
|
|
|
|
#else
|
|
|
|
static int ttm_mem_init_dma32_zone(struct ttm_mem_global *glob,
|
|
|
|
const struct sysinfo *si)
|
|
|
|
{
|
|
|
|
struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL);
|
|
|
|
uint64_t mem;
|
2009-08-20 16:29:09 +08:00
|
|
|
int ret;
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
if (unlikely(!zone))
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
mem = si->totalram;
|
|
|
|
mem *= si->mem_unit;
|
|
|
|
|
|
|
|
/**
|
|
|
|
* No special dma32 zone needed.
|
|
|
|
*/
|
|
|
|
|
2009-12-08 13:58:08 +08:00
|
|
|
if (mem <= ((uint64_t) 1ULL << 32)) {
|
|
|
|
kfree(zone);
|
2009-08-17 22:28:39 +08:00
|
|
|
return 0;
|
2009-12-08 13:58:08 +08:00
|
|
|
}
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Limit max dma32 memory to 4GB for now
|
|
|
|
* until we can figure out how big this
|
|
|
|
* zone really is.
|
|
|
|
*/
|
|
|
|
|
|
|
|
mem = ((uint64_t) 1ULL << 32);
|
|
|
|
zone->name = "dma32";
|
|
|
|
zone->zone_mem = mem;
|
|
|
|
zone->max_mem = mem >> 1;
|
|
|
|
zone->emer_mem = (mem >> 1) + (mem >> 2);
|
|
|
|
zone->swap_limit = zone->max_mem - (mem >> 3);
|
|
|
|
zone->used_mem = 0;
|
|
|
|
zone->glob = glob;
|
|
|
|
glob->zone_dma32 = zone;
|
2010-03-13 18:36:32 +08:00
|
|
|
ret = kobject_init_and_add(
|
|
|
|
&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, zone->name);
|
2009-08-20 16:29:09 +08:00
|
|
|
if (unlikely(ret != 0)) {
|
|
|
|
kobject_put(&zone->kobj);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
glob->zones[glob->num_zones++] = zone;
|
|
|
|
return 0;
|
2009-08-17 22:28:39 +08:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
int ttm_mem_global_init(struct ttm_mem_global *glob)
|
|
|
|
{
|
|
|
|
struct sysinfo si;
|
2009-08-17 22:28:39 +08:00
|
|
|
int ret;
|
|
|
|
int i;
|
|
|
|
struct ttm_mem_zone *zone;
|
2009-06-10 21:20:19 +08:00
|
|
|
|
|
|
|
spin_lock_init(&glob->lock);
|
|
|
|
glob->swap_queue = create_singlethread_workqueue("ttm_swap");
|
|
|
|
INIT_WORK(&glob->work, ttm_shrink_work);
|
|
|
|
init_waitqueue_head(&glob->queue);
|
2010-03-13 18:36:32 +08:00
|
|
|
ret = kobject_init_and_add(
|
|
|
|
&glob->kobj, &ttm_mem_glob_kobj_type, ttm_get_kobj(), "memory_accounting");
|
2009-08-20 16:29:09 +08:00
|
|
|
if (unlikely(ret != 0)) {
|
|
|
|
kobject_put(&glob->kobj);
|
|
|
|
return ret;
|
|
|
|
}
|
2009-06-10 21:20:19 +08:00
|
|
|
|
|
|
|
si_meminfo(&si);
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
ret = ttm_mem_init_kernel_zone(glob, &si);
|
|
|
|
if (unlikely(ret != 0))
|
|
|
|
goto out_no_zone;
|
|
|
|
#ifdef CONFIG_HIGHMEM
|
|
|
|
ret = ttm_mem_init_highmem_zone(glob, &si);
|
|
|
|
if (unlikely(ret != 0))
|
|
|
|
goto out_no_zone;
|
|
|
|
#else
|
|
|
|
ret = ttm_mem_init_dma32_zone(glob, &si);
|
|
|
|
if (unlikely(ret != 0))
|
|
|
|
goto out_no_zone;
|
|
|
|
#endif
|
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
2012-03-17 12:43:50 +08:00
|
|
|
pr_info("Zone %7s: Available graphics memory: %llu kiB\n",
|
|
|
|
zone->name, (unsigned long long)zone->max_mem >> 10);
|
2009-08-17 22:28:39 +08:00
|
|
|
}
|
2010-04-01 20:45:03 +08:00
|
|
|
ttm_page_alloc_init(glob, glob->zone_kernel->max_mem/(2*PAGE_SIZE));
|
drm/ttm: provide dma aware ttm page pool code V9
In TTM world the pages for the graphic drivers are kept in three different
pools: write combined, uncached, and cached (write-back). When the pages
are used by the graphic driver the graphic adapter via its built in MMU
(or AGP) programs these pages in. The programming requires the virtual address
(from the graphic adapter perspective) and the physical address (either System RAM
or the memory on the card) which is obtained using the pci_map_* calls (which does the
virtual to physical - or bus address translation). During the graphic application's
"life" those pages can be shuffled around, swapped out to disk, moved from the
VRAM to System RAM or vice-versa. This all works with the existing TTM pool code
- except when we want to use the software IOTLB (SWIOTLB) code to "map" the physical
addresses to the graphic adapter MMU. We end up programming the bounce buffer's
physical address instead of the TTM pool memory's and get a non-worky driver.
There are two solutions:
1) using the DMA API to allocate pages that are screened by the DMA API, or
2) using the pci_sync_* calls to copy the pages from the bounce-buffer and back.
This patch fixes the issue by allocating pages using the DMA API. The second
is a viable option - but it has performance drawbacks and potential correctness
issues - think of the write cache page being bounced (SWIOTLB->TTM), the
WC is set on the TTM page and the copy from SWIOTLB not making it to the TTM
page until the page has been recycled in the pool (and used by another application).
The bounce buffer does not get activated often - only in cases where we have
a 32-bit capable card and we want to use a page that is allocated above the
4GB limit. The bounce buffer offers the solution of copying the contents
of that 4GB page to an location below 4GB and then back when the operation has been
completed (or vice-versa). This is done by using the 'pci_sync_*' calls.
Note: If you look carefully enough in the existing TTM page pool code you will
notice the GFP_DMA32 flag is used - which should guarantee that the provided page
is under 4GB. It certainly is the case, except this gets ignored in two cases:
- If user specifies 'swiotlb=force' which bounces _every_ page.
- If user is using a Xen's PV Linux guest (which uses the SWIOTLB and the
underlaying PFN's aren't necessarily under 4GB).
To not have this extra copying done the other option is to allocate the pages
using the DMA API so that there is not need to map the page and perform the
expensive 'pci_sync_*' calls.
This DMA API capable TTM pool requires for this the 'struct device' to
properly call the DMA API. It also has to track the virtual and bus address of
the page being handed out in case it ends up being swapped out or de-allocated -
to make sure it is de-allocated using the proper's 'struct device'.
Implementation wise the code keeps two lists: one that is attached to the
'struct device' (via the dev->dma_pools list) and a global one to be used when
the 'struct device' is unavailable (think shrinker code). The global list can
iterate over all of the 'struct device' and its associated dma_pool. The list
in dev->dma_pools can only iterate the device's dma_pool.
/[struct device_pool]\
/---------------------------------------------------| dev |
/ +-------| dma_pool |
/-----+------\ / \--------------------/
|struct device| /-->[struct dma_pool for WC]</ /[struct device_pool]\
| dma_pools +----+ /-| dev |
| ... | \--->[struct dma_pool for uncached]<-/--| dma_pool |
\-----+------/ / \--------------------/
\----------------------------------------------/
[Two pools associated with the device (WC and UC), and the parallel list
containing the 'struct dev' and 'struct dma_pool' entries]
The maximum amount of dma pools a device can have is six: write-combined,
uncached, and cached; then there are the DMA32 variants which are:
write-combined dma32, uncached dma32, and cached dma32.
Currently this code only gets activated when any variant of the SWIOTLB IOMMU
code is running (Intel without VT-d, AMD without GART, IBM Calgary and Xen PV
with PCI devices).
Tested-by: Michel Dänzer <michel@daenzer.net>
[v1: Using swiotlb_nr_tbl instead of swiotlb_enabled]
[v2: Major overhaul - added 'inuse_list' to seperate used from inuse and reorder
the order of lists to get better performance.]
[v3: Added comments/and some logic based on review, Added Jerome tag]
[v4: rebase on top of ttm_tt & ttm_backend merge]
[v5: rebase on top of ttm memory accounting overhaul]
[v6: New rebase on top of more memory accouting changes]
[v7: well rebase on top of no memory accounting changes]
[v8: make sure pages list is initialized empty]
[v9: calll ttm_mem_global_free_page in unpopulate for accurate accountg]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Acked-by: Thomas Hellstrom <thellstrom@vmware.com>
2011-11-04 04:46:34 +08:00
|
|
|
ttm_dma_page_alloc_init(glob, glob->zone_kernel->max_mem/(2*PAGE_SIZE));
|
2009-06-10 21:20:19 +08:00
|
|
|
return 0;
|
2009-08-17 22:28:39 +08:00
|
|
|
out_no_zone:
|
|
|
|
ttm_mem_global_release(glob);
|
|
|
|
return ret;
|
2009-06-10 21:20:19 +08:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(ttm_mem_global_init);
|
|
|
|
|
|
|
|
void ttm_mem_global_release(struct ttm_mem_global *glob)
|
|
|
|
{
|
2009-08-17 22:28:39 +08:00
|
|
|
unsigned int i;
|
|
|
|
struct ttm_mem_zone *zone;
|
|
|
|
|
2010-04-01 20:44:57 +08:00
|
|
|
/* let the page allocator first stop the shrink work. */
|
|
|
|
ttm_page_alloc_fini();
|
drm/ttm: provide dma aware ttm page pool code V9
In TTM world the pages for the graphic drivers are kept in three different
pools: write combined, uncached, and cached (write-back). When the pages
are used by the graphic driver the graphic adapter via its built in MMU
(or AGP) programs these pages in. The programming requires the virtual address
(from the graphic adapter perspective) and the physical address (either System RAM
or the memory on the card) which is obtained using the pci_map_* calls (which does the
virtual to physical - or bus address translation). During the graphic application's
"life" those pages can be shuffled around, swapped out to disk, moved from the
VRAM to System RAM or vice-versa. This all works with the existing TTM pool code
- except when we want to use the software IOTLB (SWIOTLB) code to "map" the physical
addresses to the graphic adapter MMU. We end up programming the bounce buffer's
physical address instead of the TTM pool memory's and get a non-worky driver.
There are two solutions:
1) using the DMA API to allocate pages that are screened by the DMA API, or
2) using the pci_sync_* calls to copy the pages from the bounce-buffer and back.
This patch fixes the issue by allocating pages using the DMA API. The second
is a viable option - but it has performance drawbacks and potential correctness
issues - think of the write cache page being bounced (SWIOTLB->TTM), the
WC is set on the TTM page and the copy from SWIOTLB not making it to the TTM
page until the page has been recycled in the pool (and used by another application).
The bounce buffer does not get activated often - only in cases where we have
a 32-bit capable card and we want to use a page that is allocated above the
4GB limit. The bounce buffer offers the solution of copying the contents
of that 4GB page to an location below 4GB and then back when the operation has been
completed (or vice-versa). This is done by using the 'pci_sync_*' calls.
Note: If you look carefully enough in the existing TTM page pool code you will
notice the GFP_DMA32 flag is used - which should guarantee that the provided page
is under 4GB. It certainly is the case, except this gets ignored in two cases:
- If user specifies 'swiotlb=force' which bounces _every_ page.
- If user is using a Xen's PV Linux guest (which uses the SWIOTLB and the
underlaying PFN's aren't necessarily under 4GB).
To not have this extra copying done the other option is to allocate the pages
using the DMA API so that there is not need to map the page and perform the
expensive 'pci_sync_*' calls.
This DMA API capable TTM pool requires for this the 'struct device' to
properly call the DMA API. It also has to track the virtual and bus address of
the page being handed out in case it ends up being swapped out or de-allocated -
to make sure it is de-allocated using the proper's 'struct device'.
Implementation wise the code keeps two lists: one that is attached to the
'struct device' (via the dev->dma_pools list) and a global one to be used when
the 'struct device' is unavailable (think shrinker code). The global list can
iterate over all of the 'struct device' and its associated dma_pool. The list
in dev->dma_pools can only iterate the device's dma_pool.
/[struct device_pool]\
/---------------------------------------------------| dev |
/ +-------| dma_pool |
/-----+------\ / \--------------------/
|struct device| /-->[struct dma_pool for WC]</ /[struct device_pool]\
| dma_pools +----+ /-| dev |
| ... | \--->[struct dma_pool for uncached]<-/--| dma_pool |
\-----+------/ / \--------------------/
\----------------------------------------------/
[Two pools associated with the device (WC and UC), and the parallel list
containing the 'struct dev' and 'struct dma_pool' entries]
The maximum amount of dma pools a device can have is six: write-combined,
uncached, and cached; then there are the DMA32 variants which are:
write-combined dma32, uncached dma32, and cached dma32.
Currently this code only gets activated when any variant of the SWIOTLB IOMMU
code is running (Intel without VT-d, AMD without GART, IBM Calgary and Xen PV
with PCI devices).
Tested-by: Michel Dänzer <michel@daenzer.net>
[v1: Using swiotlb_nr_tbl instead of swiotlb_enabled]
[v2: Major overhaul - added 'inuse_list' to seperate used from inuse and reorder
the order of lists to get better performance.]
[v3: Added comments/and some logic based on review, Added Jerome tag]
[v4: rebase on top of ttm_tt & ttm_backend merge]
[v5: rebase on top of ttm memory accounting overhaul]
[v6: New rebase on top of more memory accouting changes]
[v7: well rebase on top of no memory accounting changes]
[v8: make sure pages list is initialized empty]
[v9: calll ttm_mem_global_free_page in unpopulate for accurate accountg]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Acked-by: Thomas Hellstrom <thellstrom@vmware.com>
2011-11-04 04:46:34 +08:00
|
|
|
ttm_dma_page_alloc_fini();
|
2010-04-01 20:44:57 +08:00
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
flush_workqueue(glob->swap_queue);
|
|
|
|
destroy_workqueue(glob->swap_queue);
|
|
|
|
glob->swap_queue = NULL;
|
2009-08-17 22:28:39 +08:00
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
|
|
|
kobject_del(&zone->kobj);
|
|
|
|
kobject_put(&zone->kobj);
|
2010-04-01 20:44:57 +08:00
|
|
|
}
|
2009-08-17 22:28:39 +08:00
|
|
|
kobject_del(&glob->kobj);
|
|
|
|
kobject_put(&glob->kobj);
|
2009-06-10 21:20:19 +08:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(ttm_mem_global_release);
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
static void ttm_check_swapping(struct ttm_mem_global *glob)
|
2009-06-10 21:20:19 +08:00
|
|
|
{
|
2009-08-17 22:28:39 +08:00
|
|
|
bool needs_swapping = false;
|
|
|
|
unsigned int i;
|
|
|
|
struct ttm_mem_zone *zone;
|
2009-06-10 21:20:19 +08:00
|
|
|
|
|
|
|
spin_lock(&glob->lock);
|
2009-08-17 22:28:39 +08:00
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
|
|
|
if (zone->used_mem > zone->swap_limit) {
|
|
|
|
needs_swapping = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
spin_unlock(&glob->lock);
|
|
|
|
|
|
|
|
if (unlikely(needs_swapping))
|
|
|
|
(void)queue_work(glob->swap_queue, &glob->work);
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
static void ttm_mem_global_free_zone(struct ttm_mem_global *glob,
|
|
|
|
struct ttm_mem_zone *single_zone,
|
|
|
|
uint64_t amount)
|
2009-06-10 21:20:19 +08:00
|
|
|
{
|
2009-08-17 22:28:39 +08:00
|
|
|
unsigned int i;
|
|
|
|
struct ttm_mem_zone *zone;
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
spin_lock(&glob->lock);
|
2009-08-17 22:28:39 +08:00
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
|
|
|
if (single_zone && zone != single_zone)
|
|
|
|
continue;
|
|
|
|
zone->used_mem -= amount;
|
|
|
|
}
|
2009-06-10 21:20:19 +08:00
|
|
|
spin_unlock(&glob->lock);
|
|
|
|
}
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
void ttm_mem_global_free(struct ttm_mem_global *glob,
|
|
|
|
uint64_t amount)
|
|
|
|
{
|
|
|
|
return ttm_mem_global_free_zone(glob, NULL, amount);
|
|
|
|
}
|
2009-12-07 04:46:27 +08:00
|
|
|
EXPORT_SYMBOL(ttm_mem_global_free);
|
2009-08-17 22:28:39 +08:00
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
static int ttm_mem_global_reserve(struct ttm_mem_global *glob,
|
2009-08-17 22:28:39 +08:00
|
|
|
struct ttm_mem_zone *single_zone,
|
|
|
|
uint64_t amount, bool reserve)
|
2009-06-10 21:20:19 +08:00
|
|
|
{
|
|
|
|
uint64_t limit;
|
|
|
|
int ret = -ENOMEM;
|
2009-08-17 22:28:39 +08:00
|
|
|
unsigned int i;
|
|
|
|
struct ttm_mem_zone *zone;
|
2009-06-10 21:20:19 +08:00
|
|
|
|
|
|
|
spin_lock(&glob->lock);
|
2009-08-17 22:28:39 +08:00
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
|
|
|
if (single_zone && zone != single_zone)
|
|
|
|
continue;
|
2009-06-10 21:20:19 +08:00
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
limit = (capable(CAP_SYS_ADMIN)) ?
|
|
|
|
zone->emer_mem : zone->max_mem;
|
2009-06-10 21:20:19 +08:00
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
if (zone->used_mem > limit)
|
|
|
|
goto out_unlock;
|
|
|
|
}
|
2009-06-10 21:20:19 +08:00
|
|
|
|
|
|
|
if (reserve) {
|
2009-08-17 22:28:39 +08:00
|
|
|
for (i = 0; i < glob->num_zones; ++i) {
|
|
|
|
zone = glob->zones[i];
|
|
|
|
if (single_zone && zone != single_zone)
|
|
|
|
continue;
|
|
|
|
zone->used_mem += amount;
|
|
|
|
}
|
2009-06-10 21:20:19 +08:00
|
|
|
}
|
2009-08-17 22:28:39 +08:00
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
ret = 0;
|
|
|
|
out_unlock:
|
|
|
|
spin_unlock(&glob->lock);
|
|
|
|
ttm_check_swapping(glob);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
static int ttm_mem_global_alloc_zone(struct ttm_mem_global *glob,
|
|
|
|
struct ttm_mem_zone *single_zone,
|
|
|
|
uint64_t memory,
|
|
|
|
bool no_wait, bool interruptible)
|
2009-06-10 21:20:19 +08:00
|
|
|
{
|
|
|
|
int count = TTM_MEMORY_ALLOC_RETRIES;
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
while (unlikely(ttm_mem_global_reserve(glob,
|
|
|
|
single_zone,
|
|
|
|
memory, true)
|
2009-06-10 21:20:19 +08:00
|
|
|
!= 0)) {
|
|
|
|
if (no_wait)
|
|
|
|
return -ENOMEM;
|
|
|
|
if (unlikely(count-- == 0))
|
|
|
|
return -ENOMEM;
|
|
|
|
ttm_shrink(glob, false, memory + (memory >> 2) + 16);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-08-17 22:28:39 +08:00
|
|
|
int ttm_mem_global_alloc(struct ttm_mem_global *glob, uint64_t memory,
|
|
|
|
bool no_wait, bool interruptible)
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
* Normal allocations of kernel memory are registered in
|
|
|
|
* all zones.
|
|
|
|
*/
|
|
|
|
|
|
|
|
return ttm_mem_global_alloc_zone(glob, NULL, memory, no_wait,
|
|
|
|
interruptible);
|
|
|
|
}
|
2009-12-07 04:46:27 +08:00
|
|
|
EXPORT_SYMBOL(ttm_mem_global_alloc);
|
2009-08-17 22:28:39 +08:00
|
|
|
|
|
|
|
int ttm_mem_global_alloc_page(struct ttm_mem_global *glob,
|
|
|
|
struct page *page,
|
|
|
|
bool no_wait, bool interruptible)
|
|
|
|
{
|
|
|
|
|
|
|
|
struct ttm_mem_zone *zone = NULL;
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Page allocations may be registed in a single zone
|
|
|
|
* only if highmem or !dma32.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifdef CONFIG_HIGHMEM
|
|
|
|
if (PageHighMem(page) && glob->zone_highmem != NULL)
|
|
|
|
zone = glob->zone_highmem;
|
|
|
|
#else
|
|
|
|
if (glob->zone_dma32 && page_to_pfn(page) > 0x00100000UL)
|
|
|
|
zone = glob->zone_kernel;
|
|
|
|
#endif
|
|
|
|
return ttm_mem_global_alloc_zone(glob, zone, PAGE_SIZE, no_wait,
|
|
|
|
interruptible);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ttm_mem_global_free_page(struct ttm_mem_global *glob, struct page *page)
|
|
|
|
{
|
|
|
|
struct ttm_mem_zone *zone = NULL;
|
|
|
|
|
|
|
|
#ifdef CONFIG_HIGHMEM
|
|
|
|
if (PageHighMem(page) && glob->zone_highmem != NULL)
|
|
|
|
zone = glob->zone_highmem;
|
|
|
|
#else
|
|
|
|
if (glob->zone_dma32 && page_to_pfn(page) > 0x00100000UL)
|
|
|
|
zone = glob->zone_kernel;
|
|
|
|
#endif
|
|
|
|
ttm_mem_global_free_zone(glob, zone, PAGE_SIZE);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2009-06-10 21:20:19 +08:00
|
|
|
size_t ttm_round_pot(size_t size)
|
|
|
|
{
|
|
|
|
if ((size & (size - 1)) == 0)
|
|
|
|
return size;
|
|
|
|
else if (size > PAGE_SIZE)
|
|
|
|
return PAGE_ALIGN(size);
|
|
|
|
else {
|
|
|
|
size_t tmp_size = 4;
|
|
|
|
|
|
|
|
while (tmp_size < size)
|
|
|
|
tmp_size <<= 1;
|
|
|
|
|
|
|
|
return tmp_size;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
2009-12-07 04:46:27 +08:00
|
|
|
EXPORT_SYMBOL(ttm_round_pot);
|