linux/drivers/gpu/drm/drm_mm.c
Vlastimil Babka 2dba5eb1c7 lib/stackdepot: allow optional init and stack_table allocation by kvmalloc()
Currently, enabling CONFIG_STACKDEPOT means its stack_table will be
allocated from memblock, even if stack depot ends up not actually used.
The default size of stack_table is 4MB on 32-bit, 8MB on 64-bit.

This is fine for use-cases such as KASAN which is also a config option
and has overhead on its own.  But it's an issue for functionality that
has to be actually enabled on boot (page_owner) or depends on hardware
(GPU drivers) and thus the memory might be wasted.  This was raised as
an issue [1] when attempting to add stackdepot support for SLUB's debug
object tracking functionality.  It's common to build kernels with
CONFIG_SLUB_DEBUG and enable slub_debug on boot only when needed, or
create only specific kmem caches with debugging for testing purposes.

It would thus be more efficient if stackdepot's table was allocated only
when actually going to be used.  This patch thus makes the allocation
(and whole stack_depot_init() call) optional:

 - Add a CONFIG_STACKDEPOT_ALWAYS_INIT flag to keep using the current
   well-defined point of allocation as part of mem_init(). Make
   CONFIG_KASAN select this flag.

 - Other users have to call stack_depot_init() as part of their own init
   when it's determined that stack depot will actually be used. This may
   depend on both config and runtime conditions. Convert current users
   which are page_owner and several in the DRM subsystem. Same will be
   done for SLUB later.

 - Because the init might now be called after the boot-time memblock
   allocation has given all memory to the buddy allocator, change
   stack_depot_init() to allocate stack_table with kvmalloc() when
   memblock is no longer available. Also handle allocation failure by
   disabling stackdepot (could have theoretically happened even with
   memblock allocation previously), and don't unnecessarily align the
   memblock allocation to its own size anymore.

[1] https://lore.kernel.org/all/CAMuHMdW=eoVzM1Re5FVoEN87nKfiLmM2+Ah7eNu2KXEhCvbZyA@mail.gmail.com/

Link: https://lkml.kernel.org/r/20211013073005.11351-1-vbabka@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Marco Elver <elver@google.com> # stackdepot
Cc: Marco Elver <elver@google.com>
Cc: Vijayanand Jitta <vjitta@codeaurora.org>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: David Airlie <airlied@linux.ie>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Oliver Glitta <glittao@gmail.com>
Cc: Imran Khan <imran.f.khan@oracle.com>
From: Colin Ian King <colin.king@canonical.com>
Subject: lib/stackdepot: fix spelling mistake and grammar in pr_err message

There is a spelling mistake of the work allocation so fix this and
re-phrase the message to make it easier to read.

Link: https://lkml.kernel.org/r/20211015104159.11282-1-colin.king@canonical.com
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
From: Vlastimil Babka <vbabka@suse.cz>
Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() - fixup

On FLATMEM, we call page_ext_init_flatmem_late() just before
kmem_cache_init() which means stack_depot_init() (called by page owner
init) will not recognize properly it should use kvmalloc() and not
memblock_alloc().  memblock_alloc() will also not issue a warning and
return a block memory that can be invalid and cause kernel page fault when
saving stacks, as reported by the kernel test robot [1].

Fix this by moving page_ext_init_flatmem_late() below kmem_cache_init() so
that slab_is_available() is true during stack_depot_init().  SPARSEMEM
doesn't have this issue, as it doesn't do page_ext_init_flatmem_late(),
but a different page_ext_init() even later in the boot process.

Thanks to Mike Rapoport for pointing out the FLATMEM init ordering issue.

While at it, also actually resolve a checkpatch warning in stack_depot_init()
from DRM CI, which was supposed to be in the original patch already.

[1] https://lore.kernel.org/all/20211014085450.GC18719@xsang-OptiPlex-9020/

Link: https://lkml.kernel.org/r/6abd9213-19a9-6d58-cedc-2414386d2d81@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reported-by: kernel test robot <oliver.sang@intel.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
From: Vlastimil Babka <vbabka@suse.cz>
Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() - fixup3

Due to cd06ab2fd4 ("drm/locking: add backtrace for locking contended
locks without backoff") landing recently to -next adding a new stack depot
user in drivers/gpu/drm/drm_modeset_lock.c we need to add an appropriate
call to stack_depot_init() there as well.

Link: https://lkml.kernel.org/r/2a692365-cfa1-64f2-34e0-8aa5674dce5e@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Naresh Kamboju <naresh.kamboju@linaro.org>
Cc: Marco Elver <elver@google.com>
Cc: Vijayanand Jitta <vjitta@codeaurora.org>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: David Airlie <airlied@linux.ie>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Oliver Glitta <glittao@gmail.com>
Cc: Imran Khan <imran.f.khan@oracle.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
From: Vlastimil Babka <vbabka@suse.cz>
Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() - fixup4

Due to 4e66934eaa ("lib: add reference counting tracking
infrastructure") landing recently to net-next adding a new stack depot
user in lib/ref_tracker.c we need to add an appropriate call to
stack_depot_init() there as well.

Link: https://lkml.kernel.org/r/45c1b738-1a2f-5b5f-2f6d-86fab206d01c@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Cc: Jiri Slab <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2022-01-22 08:33:37 +02:00

1042 lines
30 KiB
C

/**************************************************************************
*
* Copyright 2006 Tungsten Graphics, Inc., Bismarck, ND., USA.
* Copyright 2016 Intel Corporation
* 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.
*
*
**************************************************************************/
/*
* Generic simple memory manager implementation. Intended to be used as a base
* class implementation for more advanced memory managers.
*
* Note that the algorithm used is quite simple and there might be substantial
* performance gains if a smarter free list is implemented. Currently it is
* just an unordered stack of free regions. This could easily be improved if
* an RB-tree is used instead. At least if we expect heavy fragmentation.
*
* Aligned allocations can also see improvement.
*
* Authors:
* Thomas Hellström <thomas-at-tungstengraphics-dot-com>
*/
#include <linux/export.h>
#include <linux/interval_tree_generic.h>
#include <linux/seq_file.h>
#include <linux/slab.h>
#include <linux/stacktrace.h>
#include <drm/drm_mm.h>
/**
* DOC: Overview
*
* drm_mm provides a simple range allocator. The drivers are free to use the
* resource allocator from the linux core if it suits them, the upside of drm_mm
* is that it's in the DRM core. Which means that it's easier to extend for
* some of the crazier special purpose needs of gpus.
*
* The main data struct is &drm_mm, allocations are tracked in &drm_mm_node.
* Drivers are free to embed either of them into their own suitable
* datastructures. drm_mm itself will not do any memory allocations of its own,
* so if drivers choose not to embed nodes they need to still allocate them
* themselves.
*
* The range allocator also supports reservation of preallocated blocks. This is
* useful for taking over initial mode setting configurations from the firmware,
* where an object needs to be created which exactly matches the firmware's
* scanout target. As long as the range is still free it can be inserted anytime
* after the allocator is initialized, which helps with avoiding looped
* dependencies in the driver load sequence.
*
* drm_mm maintains a stack of most recently freed holes, which of all
* simplistic datastructures seems to be a fairly decent approach to clustering
* allocations and avoiding too much fragmentation. This means free space
* searches are O(num_holes). Given that all the fancy features drm_mm supports
* something better would be fairly complex and since gfx thrashing is a fairly
* steep cliff not a real concern. Removing a node again is O(1).
*
* drm_mm supports a few features: Alignment and range restrictions can be
* supplied. Furthermore every &drm_mm_node has a color value (which is just an
* opaque unsigned long) which in conjunction with a driver callback can be used
* to implement sophisticated placement restrictions. The i915 DRM driver uses
* this to implement guard pages between incompatible caching domains in the
* graphics TT.
*
* Two behaviors are supported for searching and allocating: bottom-up and
* top-down. The default is bottom-up. Top-down allocation can be used if the
* memory area has different restrictions, or just to reduce fragmentation.
*
* Finally iteration helpers to walk all nodes and all holes are provided as are
* some basic allocator dumpers for debugging.
*
* Note that this range allocator is not thread-safe, drivers need to protect
* modifications with their own locking. The idea behind this is that for a full
* memory manager additional data needs to be protected anyway, hence internal
* locking would be fully redundant.
*/
#ifdef CONFIG_DRM_DEBUG_MM
#include <linux/stackdepot.h>
#define STACKDEPTH 32
#define BUFSZ 4096
static noinline void save_stack(struct drm_mm_node *node)
{
unsigned long entries[STACKDEPTH];
unsigned int n;
n = stack_trace_save(entries, ARRAY_SIZE(entries), 1);
/* May be called under spinlock, so avoid sleeping */
node->stack = stack_depot_save(entries, n, GFP_NOWAIT);
}
static void show_leaks(struct drm_mm *mm)
{
struct drm_mm_node *node;
char *buf;
buf = kmalloc(BUFSZ, GFP_KERNEL);
if (!buf)
return;
list_for_each_entry(node, drm_mm_nodes(mm), node_list) {
if (!node->stack) {
DRM_ERROR("node [%08llx + %08llx]: unknown owner\n",
node->start, node->size);
continue;
}
stack_depot_snprint(node->stack, buf, BUFSZ, 0);
DRM_ERROR("node [%08llx + %08llx]: inserted at\n%s",
node->start, node->size, buf);
}
kfree(buf);
}
#undef STACKDEPTH
#undef BUFSZ
#else
static void save_stack(struct drm_mm_node *node) { }
static void show_leaks(struct drm_mm *mm) { }
#endif
#define START(node) ((node)->start)
#define LAST(node) ((node)->start + (node)->size - 1)
INTERVAL_TREE_DEFINE(struct drm_mm_node, rb,
u64, __subtree_last,
START, LAST, static inline, drm_mm_interval_tree)
struct drm_mm_node *
__drm_mm_interval_first(const struct drm_mm *mm, u64 start, u64 last)
{
return drm_mm_interval_tree_iter_first((struct rb_root_cached *)&mm->interval_tree,
start, last) ?: (struct drm_mm_node *)&mm->head_node;
}
EXPORT_SYMBOL(__drm_mm_interval_first);
static void drm_mm_interval_tree_add_node(struct drm_mm_node *hole_node,
struct drm_mm_node *node)
{
struct drm_mm *mm = hole_node->mm;
struct rb_node **link, *rb;
struct drm_mm_node *parent;
bool leftmost;
node->__subtree_last = LAST(node);
if (drm_mm_node_allocated(hole_node)) {
rb = &hole_node->rb;
while (rb) {
parent = rb_entry(rb, struct drm_mm_node, rb);
if (parent->__subtree_last >= node->__subtree_last)
break;
parent->__subtree_last = node->__subtree_last;
rb = rb_parent(rb);
}
rb = &hole_node->rb;
link = &hole_node->rb.rb_right;
leftmost = false;
} else {
rb = NULL;
link = &mm->interval_tree.rb_root.rb_node;
leftmost = true;
}
while (*link) {
rb = *link;
parent = rb_entry(rb, struct drm_mm_node, rb);
if (parent->__subtree_last < node->__subtree_last)
parent->__subtree_last = node->__subtree_last;
if (node->start < parent->start) {
link = &parent->rb.rb_left;
} else {
link = &parent->rb.rb_right;
leftmost = false;
}
}
rb_link_node(&node->rb, rb, link);
rb_insert_augmented_cached(&node->rb, &mm->interval_tree, leftmost,
&drm_mm_interval_tree_augment);
}
#define HOLE_SIZE(NODE) ((NODE)->hole_size)
#define HOLE_ADDR(NODE) (__drm_mm_hole_node_start(NODE))
static u64 rb_to_hole_size(struct rb_node *rb)
{
return rb_entry(rb, struct drm_mm_node, rb_hole_size)->hole_size;
}
static void insert_hole_size(struct rb_root_cached *root,
struct drm_mm_node *node)
{
struct rb_node **link = &root->rb_root.rb_node, *rb = NULL;
u64 x = node->hole_size;
bool first = true;
while (*link) {
rb = *link;
if (x > rb_to_hole_size(rb)) {
link = &rb->rb_left;
} else {
link = &rb->rb_right;
first = false;
}
}
rb_link_node(&node->rb_hole_size, rb, link);
rb_insert_color_cached(&node->rb_hole_size, root, first);
}
RB_DECLARE_CALLBACKS_MAX(static, augment_callbacks,
struct drm_mm_node, rb_hole_addr,
u64, subtree_max_hole, HOLE_SIZE)
static void insert_hole_addr(struct rb_root *root, struct drm_mm_node *node)
{
struct rb_node **link = &root->rb_node, *rb_parent = NULL;
u64 start = HOLE_ADDR(node), subtree_max_hole = node->subtree_max_hole;
struct drm_mm_node *parent;
while (*link) {
rb_parent = *link;
parent = rb_entry(rb_parent, struct drm_mm_node, rb_hole_addr);
if (parent->subtree_max_hole < subtree_max_hole)
parent->subtree_max_hole = subtree_max_hole;
if (start < HOLE_ADDR(parent))
link = &parent->rb_hole_addr.rb_left;
else
link = &parent->rb_hole_addr.rb_right;
}
rb_link_node(&node->rb_hole_addr, rb_parent, link);
rb_insert_augmented(&node->rb_hole_addr, root, &augment_callbacks);
}
static void add_hole(struct drm_mm_node *node)
{
struct drm_mm *mm = node->mm;
node->hole_size =
__drm_mm_hole_node_end(node) - __drm_mm_hole_node_start(node);
node->subtree_max_hole = node->hole_size;
DRM_MM_BUG_ON(!drm_mm_hole_follows(node));
insert_hole_size(&mm->holes_size, node);
insert_hole_addr(&mm->holes_addr, node);
list_add(&node->hole_stack, &mm->hole_stack);
}
static void rm_hole(struct drm_mm_node *node)
{
DRM_MM_BUG_ON(!drm_mm_hole_follows(node));
list_del(&node->hole_stack);
rb_erase_cached(&node->rb_hole_size, &node->mm->holes_size);
rb_erase_augmented(&node->rb_hole_addr, &node->mm->holes_addr,
&augment_callbacks);
node->hole_size = 0;
node->subtree_max_hole = 0;
DRM_MM_BUG_ON(drm_mm_hole_follows(node));
}
static inline struct drm_mm_node *rb_hole_size_to_node(struct rb_node *rb)
{
return rb_entry_safe(rb, struct drm_mm_node, rb_hole_size);
}
static inline struct drm_mm_node *rb_hole_addr_to_node(struct rb_node *rb)
{
return rb_entry_safe(rb, struct drm_mm_node, rb_hole_addr);
}
static struct drm_mm_node *best_hole(struct drm_mm *mm, u64 size)
{
struct rb_node *rb = mm->holes_size.rb_root.rb_node;
struct drm_mm_node *best = NULL;
do {
struct drm_mm_node *node =
rb_entry(rb, struct drm_mm_node, rb_hole_size);
if (size <= node->hole_size) {
best = node;
rb = rb->rb_right;
} else {
rb = rb->rb_left;
}
} while (rb);
return best;
}
static bool usable_hole_addr(struct rb_node *rb, u64 size)
{
return rb && rb_hole_addr_to_node(rb)->subtree_max_hole >= size;
}
static struct drm_mm_node *find_hole_addr(struct drm_mm *mm, u64 addr, u64 size)
{
struct rb_node *rb = mm->holes_addr.rb_node;
struct drm_mm_node *node = NULL;
while (rb) {
u64 hole_start;
if (!usable_hole_addr(rb, size))
break;
node = rb_hole_addr_to_node(rb);
hole_start = __drm_mm_hole_node_start(node);
if (addr < hole_start)
rb = node->rb_hole_addr.rb_left;
else if (addr > hole_start + node->hole_size)
rb = node->rb_hole_addr.rb_right;
else
break;
}
return node;
}
static struct drm_mm_node *
first_hole(struct drm_mm *mm,
u64 start, u64 end, u64 size,
enum drm_mm_insert_mode mode)
{
switch (mode) {
default:
case DRM_MM_INSERT_BEST:
return best_hole(mm, size);
case DRM_MM_INSERT_LOW:
return find_hole_addr(mm, start, size);
case DRM_MM_INSERT_HIGH:
return find_hole_addr(mm, end, size);
case DRM_MM_INSERT_EVICT:
return list_first_entry_or_null(&mm->hole_stack,
struct drm_mm_node,
hole_stack);
}
}
/**
* DECLARE_NEXT_HOLE_ADDR - macro to declare next hole functions
* @name: name of function to declare
* @first: first rb member to traverse (either rb_left or rb_right).
* @last: last rb member to traverse (either rb_right or rb_left).
*
* This macro declares a function to return the next hole of the addr rb tree.
* While traversing the tree we take the searched size into account and only
* visit branches with potential big enough holes.
*/
#define DECLARE_NEXT_HOLE_ADDR(name, first, last) \
static struct drm_mm_node *name(struct drm_mm_node *entry, u64 size) \
{ \
struct rb_node *parent, *node = &entry->rb_hole_addr; \
\
if (!entry || RB_EMPTY_NODE(node)) \
return NULL; \
\
if (usable_hole_addr(node->first, size)) { \
node = node->first; \
while (usable_hole_addr(node->last, size)) \
node = node->last; \
return rb_hole_addr_to_node(node); \
} \
\
while ((parent = rb_parent(node)) && node == parent->first) \
node = parent; \
\
return rb_hole_addr_to_node(parent); \
}
DECLARE_NEXT_HOLE_ADDR(next_hole_high_addr, rb_left, rb_right)
DECLARE_NEXT_HOLE_ADDR(next_hole_low_addr, rb_right, rb_left)
static struct drm_mm_node *
next_hole(struct drm_mm *mm,
struct drm_mm_node *node,
u64 size,
enum drm_mm_insert_mode mode)
{
switch (mode) {
default:
case DRM_MM_INSERT_BEST:
return rb_hole_size_to_node(rb_prev(&node->rb_hole_size));
case DRM_MM_INSERT_LOW:
return next_hole_low_addr(node, size);
case DRM_MM_INSERT_HIGH:
return next_hole_high_addr(node, size);
case DRM_MM_INSERT_EVICT:
node = list_next_entry(node, hole_stack);
return &node->hole_stack == &mm->hole_stack ? NULL : node;
}
}
/**
* drm_mm_reserve_node - insert an pre-initialized node
* @mm: drm_mm allocator to insert @node into
* @node: drm_mm_node to insert
*
* This functions inserts an already set-up &drm_mm_node into the allocator,
* meaning that start, size and color must be set by the caller. All other
* fields must be cleared to 0. This is useful to initialize the allocator with
* preallocated objects which must be set-up before the range allocator can be
* set-up, e.g. when taking over a firmware framebuffer.
*
* Returns:
* 0 on success, -ENOSPC if there's no hole where @node is.
*/
int drm_mm_reserve_node(struct drm_mm *mm, struct drm_mm_node *node)
{
struct drm_mm_node *hole;
u64 hole_start, hole_end;
u64 adj_start, adj_end;
u64 end;
end = node->start + node->size;
if (unlikely(end <= node->start))
return -ENOSPC;
/* Find the relevant hole to add our node to */
hole = find_hole_addr(mm, node->start, 0);
if (!hole)
return -ENOSPC;
adj_start = hole_start = __drm_mm_hole_node_start(hole);
adj_end = hole_end = hole_start + hole->hole_size;
if (mm->color_adjust)
mm->color_adjust(hole, node->color, &adj_start, &adj_end);
if (adj_start > node->start || adj_end < end)
return -ENOSPC;
node->mm = mm;
__set_bit(DRM_MM_NODE_ALLOCATED_BIT, &node->flags);
list_add(&node->node_list, &hole->node_list);
drm_mm_interval_tree_add_node(hole, node);
node->hole_size = 0;
rm_hole(hole);
if (node->start > hole_start)
add_hole(hole);
if (end < hole_end)
add_hole(node);
save_stack(node);
return 0;
}
EXPORT_SYMBOL(drm_mm_reserve_node);
static u64 rb_to_hole_size_or_zero(struct rb_node *rb)
{
return rb ? rb_to_hole_size(rb) : 0;
}
/**
* drm_mm_insert_node_in_range - ranged search for space and insert @node
* @mm: drm_mm to allocate from
* @node: preallocate node to insert
* @size: size of the allocation
* @alignment: alignment of the allocation
* @color: opaque tag value to use for this node
* @range_start: start of the allowed range for this node
* @range_end: end of the allowed range for this node
* @mode: fine-tune the allocation search and placement
*
* The preallocated @node must be cleared to 0.
*
* Returns:
* 0 on success, -ENOSPC if there's no suitable hole.
*/
int drm_mm_insert_node_in_range(struct drm_mm * const mm,
struct drm_mm_node * const node,
u64 size, u64 alignment,
unsigned long color,
u64 range_start, u64 range_end,
enum drm_mm_insert_mode mode)
{
struct drm_mm_node *hole;
u64 remainder_mask;
bool once;
DRM_MM_BUG_ON(range_start > range_end);
if (unlikely(size == 0 || range_end - range_start < size))
return -ENOSPC;
if (rb_to_hole_size_or_zero(rb_first_cached(&mm->holes_size)) < size)
return -ENOSPC;
if (alignment <= 1)
alignment = 0;
once = mode & DRM_MM_INSERT_ONCE;
mode &= ~DRM_MM_INSERT_ONCE;
remainder_mask = is_power_of_2(alignment) ? alignment - 1 : 0;
for (hole = first_hole(mm, range_start, range_end, size, mode);
hole;
hole = once ? NULL : next_hole(mm, hole, size, mode)) {
u64 hole_start = __drm_mm_hole_node_start(hole);
u64 hole_end = hole_start + hole->hole_size;
u64 adj_start, adj_end;
u64 col_start, col_end;
if (mode == DRM_MM_INSERT_LOW && hole_start >= range_end)
break;
if (mode == DRM_MM_INSERT_HIGH && hole_end <= range_start)
break;
col_start = hole_start;
col_end = hole_end;
if (mm->color_adjust)
mm->color_adjust(hole, color, &col_start, &col_end);
adj_start = max(col_start, range_start);
adj_end = min(col_end, range_end);
if (adj_end <= adj_start || adj_end - adj_start < size)
continue;
if (mode == DRM_MM_INSERT_HIGH)
adj_start = adj_end - size;
if (alignment) {
u64 rem;
if (likely(remainder_mask))
rem = adj_start & remainder_mask;
else
div64_u64_rem(adj_start, alignment, &rem);
if (rem) {
adj_start -= rem;
if (mode != DRM_MM_INSERT_HIGH)
adj_start += alignment;
if (adj_start < max(col_start, range_start) ||
min(col_end, range_end) - adj_start < size)
continue;
if (adj_end <= adj_start ||
adj_end - adj_start < size)
continue;
}
}
node->mm = mm;
node->size = size;
node->start = adj_start;
node->color = color;
node->hole_size = 0;
__set_bit(DRM_MM_NODE_ALLOCATED_BIT, &node->flags);
list_add(&node->node_list, &hole->node_list);
drm_mm_interval_tree_add_node(hole, node);
rm_hole(hole);
if (adj_start > hole_start)
add_hole(hole);
if (adj_start + size < hole_end)
add_hole(node);
save_stack(node);
return 0;
}
return -ENOSPC;
}
EXPORT_SYMBOL(drm_mm_insert_node_in_range);
static inline bool drm_mm_node_scanned_block(const struct drm_mm_node *node)
{
return test_bit(DRM_MM_NODE_SCANNED_BIT, &node->flags);
}
/**
* drm_mm_remove_node - Remove a memory node from the allocator.
* @node: drm_mm_node to remove
*
* This just removes a node from its drm_mm allocator. The node does not need to
* be cleared again before it can be re-inserted into this or any other drm_mm
* allocator. It is a bug to call this function on a unallocated node.
*/
void drm_mm_remove_node(struct drm_mm_node *node)
{
struct drm_mm *mm = node->mm;
struct drm_mm_node *prev_node;
DRM_MM_BUG_ON(!drm_mm_node_allocated(node));
DRM_MM_BUG_ON(drm_mm_node_scanned_block(node));
prev_node = list_prev_entry(node, node_list);
if (drm_mm_hole_follows(node))
rm_hole(node);
drm_mm_interval_tree_remove(node, &mm->interval_tree);
list_del(&node->node_list);
if (drm_mm_hole_follows(prev_node))
rm_hole(prev_node);
add_hole(prev_node);
clear_bit_unlock(DRM_MM_NODE_ALLOCATED_BIT, &node->flags);
}
EXPORT_SYMBOL(drm_mm_remove_node);
/**
* drm_mm_replace_node - move an allocation from @old to @new
* @old: drm_mm_node to remove from the allocator
* @new: drm_mm_node which should inherit @old's allocation
*
* This is useful for when drivers embed the drm_mm_node structure and hence
* can't move allocations by reassigning pointers. It's a combination of remove
* and insert with the guarantee that the allocation start will match.
*/
void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new)
{
struct drm_mm *mm = old->mm;
DRM_MM_BUG_ON(!drm_mm_node_allocated(old));
*new = *old;
__set_bit(DRM_MM_NODE_ALLOCATED_BIT, &new->flags);
list_replace(&old->node_list, &new->node_list);
rb_replace_node_cached(&old->rb, &new->rb, &mm->interval_tree);
if (drm_mm_hole_follows(old)) {
list_replace(&old->hole_stack, &new->hole_stack);
rb_replace_node_cached(&old->rb_hole_size,
&new->rb_hole_size,
&mm->holes_size);
rb_replace_node(&old->rb_hole_addr,
&new->rb_hole_addr,
&mm->holes_addr);
}
clear_bit_unlock(DRM_MM_NODE_ALLOCATED_BIT, &old->flags);
}
EXPORT_SYMBOL(drm_mm_replace_node);
/**
* DOC: lru scan roster
*
* Very often GPUs need to have continuous allocations for a given object. When
* evicting objects to make space for a new one it is therefore not most
* efficient when we simply start to select all objects from the tail of an LRU
* until there's a suitable hole: Especially for big objects or nodes that
* otherwise have special allocation constraints there's a good chance we evict
* lots of (smaller) objects unnecessarily.
*
* The DRM range allocator supports this use-case through the scanning
* interfaces. First a scan operation needs to be initialized with
* drm_mm_scan_init() or drm_mm_scan_init_with_range(). The driver adds
* objects to the roster, probably by walking an LRU list, but this can be
* freely implemented. Eviction candidates are added using
* drm_mm_scan_add_block() until a suitable hole is found or there are no
* further evictable objects. Eviction roster metadata is tracked in &struct
* drm_mm_scan.
*
* The driver must walk through all objects again in exactly the reverse
* order to restore the allocator state. Note that while the allocator is used
* in the scan mode no other operation is allowed.
*
* Finally the driver evicts all objects selected (drm_mm_scan_remove_block()
* reported true) in the scan, and any overlapping nodes after color adjustment
* (drm_mm_scan_color_evict()). Adding and removing an object is O(1), and
* since freeing a node is also O(1) the overall complexity is
* O(scanned_objects). So like the free stack which needs to be walked before a
* scan operation even begins this is linear in the number of objects. It
* doesn't seem to hurt too badly.
*/
/**
* drm_mm_scan_init_with_range - initialize range-restricted lru scanning
* @scan: scan state
* @mm: drm_mm to scan
* @size: size of the allocation
* @alignment: alignment of the allocation
* @color: opaque tag value to use for the allocation
* @start: start of the allowed range for the allocation
* @end: end of the allowed range for the allocation
* @mode: fine-tune the allocation search and placement
*
* This simply sets up the scanning routines with the parameters for the desired
* hole.
*
* Warning:
* As long as the scan list is non-empty, no other operations than
* adding/removing nodes to/from the scan list are allowed.
*/
void drm_mm_scan_init_with_range(struct drm_mm_scan *scan,
struct drm_mm *mm,
u64 size,
u64 alignment,
unsigned long color,
u64 start,
u64 end,
enum drm_mm_insert_mode mode)
{
DRM_MM_BUG_ON(start >= end);
DRM_MM_BUG_ON(!size || size > end - start);
DRM_MM_BUG_ON(mm->scan_active);
scan->mm = mm;
if (alignment <= 1)
alignment = 0;
scan->color = color;
scan->alignment = alignment;
scan->remainder_mask = is_power_of_2(alignment) ? alignment - 1 : 0;
scan->size = size;
scan->mode = mode;
DRM_MM_BUG_ON(end <= start);
scan->range_start = start;
scan->range_end = end;
scan->hit_start = U64_MAX;
scan->hit_end = 0;
}
EXPORT_SYMBOL(drm_mm_scan_init_with_range);
/**
* drm_mm_scan_add_block - add a node to the scan list
* @scan: the active drm_mm scanner
* @node: drm_mm_node to add
*
* Add a node to the scan list that might be freed to make space for the desired
* hole.
*
* Returns:
* True if a hole has been found, false otherwise.
*/
bool drm_mm_scan_add_block(struct drm_mm_scan *scan,
struct drm_mm_node *node)
{
struct drm_mm *mm = scan->mm;
struct drm_mm_node *hole;
u64 hole_start, hole_end;
u64 col_start, col_end;
u64 adj_start, adj_end;
DRM_MM_BUG_ON(node->mm != mm);
DRM_MM_BUG_ON(!drm_mm_node_allocated(node));
DRM_MM_BUG_ON(drm_mm_node_scanned_block(node));
__set_bit(DRM_MM_NODE_SCANNED_BIT, &node->flags);
mm->scan_active++;
/* Remove this block from the node_list so that we enlarge the hole
* (distance between the end of our previous node and the start of
* or next), without poisoning the link so that we can restore it
* later in drm_mm_scan_remove_block().
*/
hole = list_prev_entry(node, node_list);
DRM_MM_BUG_ON(list_next_entry(hole, node_list) != node);
__list_del_entry(&node->node_list);
hole_start = __drm_mm_hole_node_start(hole);
hole_end = __drm_mm_hole_node_end(hole);
col_start = hole_start;
col_end = hole_end;
if (mm->color_adjust)
mm->color_adjust(hole, scan->color, &col_start, &col_end);
adj_start = max(col_start, scan->range_start);
adj_end = min(col_end, scan->range_end);
if (adj_end <= adj_start || adj_end - adj_start < scan->size)
return false;
if (scan->mode == DRM_MM_INSERT_HIGH)
adj_start = adj_end - scan->size;
if (scan->alignment) {
u64 rem;
if (likely(scan->remainder_mask))
rem = adj_start & scan->remainder_mask;
else
div64_u64_rem(adj_start, scan->alignment, &rem);
if (rem) {
adj_start -= rem;
if (scan->mode != DRM_MM_INSERT_HIGH)
adj_start += scan->alignment;
if (adj_start < max(col_start, scan->range_start) ||
min(col_end, scan->range_end) - adj_start < scan->size)
return false;
if (adj_end <= adj_start ||
adj_end - adj_start < scan->size)
return false;
}
}
scan->hit_start = adj_start;
scan->hit_end = adj_start + scan->size;
DRM_MM_BUG_ON(scan->hit_start >= scan->hit_end);
DRM_MM_BUG_ON(scan->hit_start < hole_start);
DRM_MM_BUG_ON(scan->hit_end > hole_end);
return true;
}
EXPORT_SYMBOL(drm_mm_scan_add_block);
/**
* drm_mm_scan_remove_block - remove a node from the scan list
* @scan: the active drm_mm scanner
* @node: drm_mm_node to remove
*
* Nodes **must** be removed in exactly the reverse order from the scan list as
* they have been added (e.g. using list_add() as they are added and then
* list_for_each() over that eviction list to remove), otherwise the internal
* state of the memory manager will be corrupted.
*
* When the scan list is empty, the selected memory nodes can be freed. An
* immediately following drm_mm_insert_node_in_range_generic() or one of the
* simpler versions of that function with !DRM_MM_SEARCH_BEST will then return
* the just freed block (because it's at the top of the free_stack list).
*
* Returns:
* True if this block should be evicted, false otherwise. Will always
* return false when no hole has been found.
*/
bool drm_mm_scan_remove_block(struct drm_mm_scan *scan,
struct drm_mm_node *node)
{
struct drm_mm_node *prev_node;
DRM_MM_BUG_ON(node->mm != scan->mm);
DRM_MM_BUG_ON(!drm_mm_node_scanned_block(node));
__clear_bit(DRM_MM_NODE_SCANNED_BIT, &node->flags);
DRM_MM_BUG_ON(!node->mm->scan_active);
node->mm->scan_active--;
/* During drm_mm_scan_add_block() we decoupled this node leaving
* its pointers intact. Now that the caller is walking back along
* the eviction list we can restore this block into its rightful
* place on the full node_list. To confirm that the caller is walking
* backwards correctly we check that prev_node->next == node->next,
* i.e. both believe the same node should be on the other side of the
* hole.
*/
prev_node = list_prev_entry(node, node_list);
DRM_MM_BUG_ON(list_next_entry(prev_node, node_list) !=
list_next_entry(node, node_list));
list_add(&node->node_list, &prev_node->node_list);
return (node->start + node->size > scan->hit_start &&
node->start < scan->hit_end);
}
EXPORT_SYMBOL(drm_mm_scan_remove_block);
/**
* drm_mm_scan_color_evict - evict overlapping nodes on either side of hole
* @scan: drm_mm scan with target hole
*
* After completing an eviction scan and removing the selected nodes, we may
* need to remove a few more nodes from either side of the target hole if
* mm.color_adjust is being used.
*
* Returns:
* A node to evict, or NULL if there are no overlapping nodes.
*/
struct drm_mm_node *drm_mm_scan_color_evict(struct drm_mm_scan *scan)
{
struct drm_mm *mm = scan->mm;
struct drm_mm_node *hole;
u64 hole_start, hole_end;
DRM_MM_BUG_ON(list_empty(&mm->hole_stack));
if (!mm->color_adjust)
return NULL;
/*
* The hole found during scanning should ideally be the first element
* in the hole_stack list, but due to side-effects in the driver it
* may not be.
*/
list_for_each_entry(hole, &mm->hole_stack, hole_stack) {
hole_start = __drm_mm_hole_node_start(hole);
hole_end = hole_start + hole->hole_size;
if (hole_start <= scan->hit_start &&
hole_end >= scan->hit_end)
break;
}
/* We should only be called after we found the hole previously */
DRM_MM_BUG_ON(&hole->hole_stack == &mm->hole_stack);
if (unlikely(&hole->hole_stack == &mm->hole_stack))
return NULL;
DRM_MM_BUG_ON(hole_start > scan->hit_start);
DRM_MM_BUG_ON(hole_end < scan->hit_end);
mm->color_adjust(hole, scan->color, &hole_start, &hole_end);
if (hole_start > scan->hit_start)
return hole;
if (hole_end < scan->hit_end)
return list_next_entry(hole, node_list);
return NULL;
}
EXPORT_SYMBOL(drm_mm_scan_color_evict);
/**
* drm_mm_init - initialize a drm-mm allocator
* @mm: the drm_mm structure to initialize
* @start: start of the range managed by @mm
* @size: end of the range managed by @mm
*
* Note that @mm must be cleared to 0 before calling this function.
*/
void drm_mm_init(struct drm_mm *mm, u64 start, u64 size)
{
DRM_MM_BUG_ON(start + size <= start);
mm->color_adjust = NULL;
INIT_LIST_HEAD(&mm->hole_stack);
mm->interval_tree = RB_ROOT_CACHED;
mm->holes_size = RB_ROOT_CACHED;
mm->holes_addr = RB_ROOT;
/* Clever trick to avoid a special case in the free hole tracking. */
INIT_LIST_HEAD(&mm->head_node.node_list);
mm->head_node.flags = 0;
mm->head_node.mm = mm;
mm->head_node.start = start + size;
mm->head_node.size = -size;
add_hole(&mm->head_node);
mm->scan_active = 0;
#ifdef CONFIG_DRM_DEBUG_MM
stack_depot_init();
#endif
}
EXPORT_SYMBOL(drm_mm_init);
/**
* drm_mm_takedown - clean up a drm_mm allocator
* @mm: drm_mm allocator to clean up
*
* Note that it is a bug to call this function on an allocator which is not
* clean.
*/
void drm_mm_takedown(struct drm_mm *mm)
{
if (WARN(!drm_mm_clean(mm),
"Memory manager not clean during takedown.\n"))
show_leaks(mm);
}
EXPORT_SYMBOL(drm_mm_takedown);
static u64 drm_mm_dump_hole(struct drm_printer *p, const struct drm_mm_node *entry)
{
u64 start, size;
size = entry->hole_size;
if (size) {
start = drm_mm_hole_node_start(entry);
drm_printf(p, "%#018llx-%#018llx: %llu: free\n",
start, start + size, size);
}
return size;
}
/**
* drm_mm_print - print allocator state
* @mm: drm_mm allocator to print
* @p: DRM printer to use
*/
void drm_mm_print(const struct drm_mm *mm, struct drm_printer *p)
{
const struct drm_mm_node *entry;
u64 total_used = 0, total_free = 0, total = 0;
total_free += drm_mm_dump_hole(p, &mm->head_node);
drm_mm_for_each_node(entry, mm) {
drm_printf(p, "%#018llx-%#018llx: %llu: used\n", entry->start,
entry->start + entry->size, entry->size);
total_used += entry->size;
total_free += drm_mm_dump_hole(p, entry);
}
total = total_free + total_used;
drm_printf(p, "total: %llu, used %llu free %llu\n", total,
total_used, total_free);
}
EXPORT_SYMBOL(drm_mm_print);