From: Michal Hocko <mhocko@suse.com>
Date: Wed, 13 Sep 2017 16:28:29 -0700
Subject: mm: treewide: remove GFP_TEMPORARY allocation flag
Git-commit: 0ee931c4e31a5efb134c76440405e9219f896e33 (partial)
Patch-mainline: v4.14-rc1
References: FATE#326289 FATE#326079 FATE#326049 FATE#322398 FATE#326166
GFP_TEMPORARY was introduced by commit e12ba74d8ff3 ("Group short-lived
and reclaimable kernel allocations") along with __GFP_RECLAIMABLE. It's
primary motivation was to allow users to tell that an allocation is
short lived and so the allocator can try to place such allocations close
together and prevent long term fragmentation. As much as this sounds
like a reasonable semantic it becomes much less clear when to use the
highlevel GFP_TEMPORARY allocation flag. How long is temporary? Can the
context holding that memory sleep? Can it take locks? It seems there is
no good answer for those questions.
The current implementation of GFP_TEMPORARY is basically GFP_KERNEL |
__GFP_RECLAIMABLE which in itself is tricky because basically none of
the existing caller provide a way to reclaim the allocated memory. So
this is rather misleading and hard to evaluate for any benefits.
I have checked some random users and none of them has added the flag
with a specific justification. I suspect most of them just copied from
other existing users and others just thought it might be a good idea to
use without any measuring. This suggests that GFP_TEMPORARY just
motivates for cargo cult usage without any reasoning.
I believe that our gfp flags are quite complex already and especially
those with highlevel semantic should be clearly defined to prevent from
confusion and abuse. Therefore I propose dropping GFP_TEMPORARY and
replace all existing users to simply use GFP_KERNEL. Please note that
SLAB users with shrinkers will still get __GFP_RECLAIMABLE heuristic and
so they will be placed properly for memory fragmentation prevention.
I can see reasons we might want some gfp flag to reflect shorterm
allocations but I propose starting from a clear semantic definition and
only then add users with proper justification.
This was been brought up before LSF this year by Matthew [1] and it
turned out that GFP_TEMPORARY really doesn't have a clear semantic. It
seems to be a heuristic without any measured advantage for most (if not
all) its current users. The follow up discussion has revealed that
opinions on what might be temporary allocation differ a lot between
developers. So rather than trying to tweak existing users into a
semantic which they haven't expected I propose to simply remove the flag
and start from scratch if we really need a semantic for short term
allocations.
[1] http://lkml.kernel.org/r/20170118054945.GD18349@bombadil.infradead.org
[akpm@linux-foundation.org: fix typo]
[akpm@linux-foundation.org: coding-style fixes]
[sfr@canb.auug.org.au: drm/i915: fix up]
Link: http://lkml.kernel.org/r/20170816144703.378d4f4d@canb.auug.org.au
Link: http://lkml.kernel.org/r/20170728091904.14627-1-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Neil Brown <neilb@suse.de>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[ ptesarik: only change GFP_TEMPORARY to GFP_KERNEL under drm ]
Signed-off-by: Petr Tesarik <ptesarik@suse.com>
---
drivers/gpu/drm/drm_blend.c | 2 +-
drivers/gpu/drm/drm_dp_dual_mode_helper.c | 2 +-
drivers/gpu/drm/drm_scdc_helper.c | 2 +-
drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c | 2 +-
drivers/gpu/drm/i915/i915_gem.c | 2 +-
drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++++++------
drivers/gpu/drm/i915/i915_gem_gtt.c | 2 +-
drivers/gpu/drm/i915/i915_gem_userptr.c | 4 ++--
drivers/gpu/drm/i915/i915_gpu_error.c | 6 +++---
drivers/gpu/drm/i915/selftests/i915_random.c | 2 +-
drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c | 10 +++++-----
drivers/gpu/drm/i915/selftests/intel_uncore.c | 2 +-
drivers/gpu/drm/lib/drm_random.c | 2 +-
drivers/gpu/drm/msm/msm_gem_submit.c | 2 +-
drivers/gpu/drm/selftests/test-drm_mm.c | 4 ++--
15 files changed, 28 insertions(+), 28 deletions(-)
--- a/drivers/gpu/drm/drm_blend.c
+++ b/drivers/gpu/drm/drm_blend.c
@@ -319,7 +319,7 @@ static int drm_atomic_helper_crtc_normal
DRM_DEBUG_ATOMIC("[CRTC:%d:%s] calculating normalized zpos values\n",
crtc->base.id, crtc->name);
- states = kmalloc_array(total_planes, sizeof(*states), GFP_TEMPORARY);
+ states = kmalloc_array(total_planes, sizeof(*states), GFP_KERNEL);
if (!states)
return -ENOMEM;
--- a/drivers/gpu/drm/drm_dp_dual_mode_helper.c
+++ b/drivers/gpu/drm/drm_dp_dual_mode_helper.c
@@ -111,7 +111,7 @@ ssize_t drm_dp_dual_mode_write(struct i2
void *data;
int ret;
- data = kmalloc(msg.len, GFP_TEMPORARY);
+ data = kmalloc(msg.len, GFP_KERNEL);
if (!data)
return -ENOMEM;
--- a/drivers/gpu/drm/drm_scdc_helper.c
+++ b/drivers/gpu/drm/drm_scdc_helper.c
@@ -102,7 +102,7 @@ ssize_t drm_scdc_write(struct i2c_adapte
void *data;
int err;
- data = kmalloc(1 + size, GFP_TEMPORARY);
+ data = kmalloc(1 + size, GFP_KERNEL);
if (!data)
return -ENOMEM;
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c
@@ -37,7 +37,7 @@ static struct etnaviv_gem_submit *submit
struct etnaviv_gem_submit *submit;
size_t sz = size_vstruct(nr, sizeof(submit->bos[0]), sizeof(*submit));
- submit = kmalloc(sz, GFP_TEMPORARY | __GFP_NOWARN | __GFP_NORETRY);
+ submit = kmalloc(sz, GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY);
if (submit) {
submit->dev = dev;
submit->gpu = gpu;
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2540,7 +2540,7 @@ static void *i915_gem_object_map(const s
if (n_pages > ARRAY_SIZE(stack_pages)) {
/* Too big for stack -- allocate temporary array instead */
- pages = kvmalloc_array(n_pages, sizeof(*pages), GFP_TEMPORARY);
+ pages = kvmalloc_array(n_pages, sizeof(*pages), GFP_KERNEL);
if (!pages)
return NULL;
}
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -293,7 +293,7 @@ static int eb_create(struct i915_execbuf
* as possible to perform the allocation and warn
* if it fails.
*/
- flags = GFP_TEMPORARY;
+ flags = GFP_KERNEL;
if (size > 1)
flags |= __GFP_NORETRY | __GFP_NOWARN;
@@ -1515,7 +1515,7 @@ static int eb_copy_relocations(const str
urelocs = u64_to_user_ptr(eb->exec[i].relocs_ptr);
size = nreloc * sizeof(*relocs);
- relocs = kvmalloc_array(size, 1, GFP_TEMPORARY);
+ relocs = kvmalloc_array(size, 1, GFP_KERNEL);
if (!relocs) {
kvfree(relocs);
err = -ENOMEM;
@@ -2077,7 +2077,7 @@ get_fence_array(struct drm_i915_gem_exec
return ERR_PTR(-EFAULT);
fences = kvmalloc_array(args->num_cliprects, sizeof(*fences),
- __GFP_NOWARN | GFP_TEMPORARY);
+ __GFP_NOWARN | GFP_KERNEL);
if (!fences)
return ERR_PTR(-ENOMEM);
@@ -2463,9 +2463,9 @@ i915_gem_execbuffer(struct drm_device *d
/* Copy in the exec list from userland */
exec_list = kvmalloc_array(args->buffer_count, sizeof(*exec_list),
- __GFP_NOWARN | GFP_TEMPORARY);
+ __GFP_NOWARN | GFP_KERNEL);
exec2_list = kvmalloc_array(args->buffer_count + 1, sz,
- __GFP_NOWARN | GFP_TEMPORARY);
+ __GFP_NOWARN | GFP_KERNEL);
if (exec_list == NULL || exec2_list == NULL) {
DRM_DEBUG("Failed to allocate exec list for %d buffers\n",
args->buffer_count);
@@ -2543,7 +2543,7 @@ i915_gem_execbuffer2(struct drm_device *
/* Allocate an extra slot for use by the command parser */
exec2_list = kvmalloc_array(args->buffer_count + 1, sz,
- __GFP_NOWARN | GFP_TEMPORARY);
+ __GFP_NOWARN | GFP_KERNEL);
if (exec2_list == NULL) {
DRM_DEBUG("Failed to allocate exec list for %d buffers\n",
args->buffer_count);
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -3231,7 +3231,7 @@ intel_rotate_pages(struct intel_rotation
/* Allocate a temporary list of source pages for random access. */
page_addr_list = kvmalloc_array(n_pages,
sizeof(dma_addr_t),
- GFP_TEMPORARY);
+ GFP_KERNEL);
if (!page_addr_list)
return ERR_PTR(ret);
--- a/drivers/gpu/drm/i915/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -507,7 +507,7 @@ __i915_gem_userptr_get_pages_worker(stru
ret = -ENOMEM;
pinned = 0;
- pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_TEMPORARY);
+ pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
if (pvec != NULL) {
struct mm_struct *mm = obj->userptr.mm->mm;
unsigned int flags = 0;
@@ -643,7 +643,7 @@ i915_gem_userptr_get_pages(struct drm_i9
if (mm == current->mm) {
pvec = kvmalloc_array(num_pages, sizeof(struct page *),
- GFP_TEMPORARY |
+ GFP_KERNEL |
__GFP_NORETRY |
__GFP_NOWARN);
if (pvec) /* defer to worker if malloc fails */
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -787,16 +787,16 @@ int i915_error_state_buf_init(struct drm
*/
ebuf->size = count + 1 > PAGE_SIZE ? count + 1 : PAGE_SIZE;
ebuf->buf = kmalloc(ebuf->size,
- GFP_TEMPORARY | __GFP_NORETRY | __GFP_NOWARN);
+ GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN);
if (ebuf->buf == NULL) {
ebuf->size = PAGE_SIZE;
- ebuf->buf = kmalloc(ebuf->size, GFP_TEMPORARY);
+ ebuf->buf = kmalloc(ebuf->size, GFP_KERNEL);
}
if (ebuf->buf == NULL) {
ebuf->size = 128;
- ebuf->buf = kmalloc(ebuf->size, GFP_TEMPORARY);
+ ebuf->buf = kmalloc(ebuf->size, GFP_KERNEL);
}
if (ebuf->buf == NULL)
--- a/drivers/gpu/drm/i915/selftests/i915_random.c
+++ b/drivers/gpu/drm/i915/selftests/i915_random.c
@@ -62,7 +62,7 @@ unsigned int *i915_random_order(unsigned
{
unsigned int *order, i;
- order = kmalloc_array(count, sizeof(*order), GFP_TEMPORARY);
+ order = kmalloc_array(count, sizeof(*order), GFP_KERNEL);
if (!order)
return order;
--- a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
@@ -117,12 +117,12 @@ static int igt_random_insert_remove(void
mock_engine_reset(engine);
- waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY);
+ waiters = kvmalloc_array(count, sizeof(*waiters), GFP_KERNEL);
if (!waiters)
goto out_engines;
bitmap = kcalloc(DIV_ROUND_UP(count, BITS_PER_LONG), sizeof(*bitmap),
- GFP_TEMPORARY);
+ GFP_KERNEL);
if (!bitmap)
goto out_waiters;
@@ -187,12 +187,12 @@ static int igt_insert_complete(void *arg
mock_engine_reset(engine);
- waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY);
+ waiters = kvmalloc_array(count, sizeof(*waiters), GFP_KERNEL);
if (!waiters)
goto out_engines;
bitmap = kcalloc(DIV_ROUND_UP(count, BITS_PER_LONG), sizeof(*bitmap),
- GFP_TEMPORARY);
+ GFP_KERNEL);
if (!bitmap)
goto out_waiters;
@@ -368,7 +368,7 @@ static int igt_wakeup(void *arg)
mock_engine_reset(engine);
- waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY);
+ waiters = kvmalloc_array(count, sizeof(*waiters), GFP_KERNEL);
if (!waiters)
goto out_engines;
--- a/drivers/gpu/drm/i915/selftests/intel_uncore.c
+++ b/drivers/gpu/drm/i915/selftests/intel_uncore.c
@@ -127,7 +127,7 @@ static int intel_uncore_check_forcewake_
return 0;
valid = kzalloc(BITS_TO_LONGS(FW_RANGE) * sizeof(*valid),
- GFP_TEMPORARY);
+ GFP_KERNEL);
if (!valid)
return -ENOMEM;
--- a/drivers/gpu/drm/lib/drm_random.c
+++ b/drivers/gpu/drm/lib/drm_random.c
@@ -28,7 +28,7 @@ unsigned int *drm_random_order(unsigned
{
unsigned int *order, i;
- order = kmalloc_array(count, sizeof(*order), GFP_TEMPORARY);
+ order = kmalloc_array(count, sizeof(*order), GFP_KERNEL);
if (!order)
return order;
--- a/drivers/gpu/drm/msm/msm_gem_submit.c
+++ b/drivers/gpu/drm/msm/msm_gem_submit.c
@@ -40,7 +40,7 @@ static struct msm_gem_submit *submit_cre
if (sz > SIZE_MAX)
return NULL;
- submit = kmalloc(sz, GFP_TEMPORARY | __GFP_NOWARN | __GFP_NORETRY);
+ submit = kmalloc(sz, GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY);
if (!submit)
return NULL;
--- a/drivers/gpu/drm/selftests/test-drm_mm.c
+++ b/drivers/gpu/drm/selftests/test-drm_mm.c
@@ -1627,7 +1627,7 @@ static int igt_topdown(void *ignored)
goto err;
bitmap = kzalloc(count / BITS_PER_LONG * sizeof(unsigned long),
- GFP_TEMPORARY);
+ GFP_KERNEL);
if (!bitmap)
goto err_nodes;
@@ -1741,7 +1741,7 @@ static int igt_bottomup(void *ignored)
goto err;
bitmap = kzalloc(count / BITS_PER_LONG * sizeof(unsigned long),
- GFP_TEMPORARY);
+ GFP_KERNEL);
if (!bitmap)
goto err_nodes;