Evolution-X-Devices/kernel_google_wahoo
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

binder: always allocate/map first BINDER_MIN_ALLOC pages

Browse Source 
Certain usecases like camera are constantly allocating and freeing
binder buffers beyond the first 4k resulting in mmap_sem contention.

If we expand the allocated range from 4k to something higher, we can
reduce the contention. Tests show that 6 pages is enough to cause very
little update_page_range operations and reduces contention.

Bug: 36727951

Change-Id: I28bc3fb9b33c764c257e28487712fce2a3c1078b
Reported-by: Tim Murray <timmurray@google.com>
Signed-off-by: Joel Fernandes <joelaf@google.com>

Pre-allocate 1 instead of 6 pages as in the original patch,
as we use this pre-allocated page to prevent the first page
from getting unpinned after removing the buffer headers,
rather than pinning pages to speedup larger transactions.

Change-Id: Id027adcfd61b2d6b37f69a3f6009a068e90e84f0
Signed-off-by: Sherry Yang <sherryy@android.com>
This commit is contained in:
Joel Fernandes
2017-06-05 18:05:54 -07:00
committed by John Dias
parent b487b37b84
commit 2410d72197
1 changed files with 21 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
drivers/android/binder_alloc.c
View File

@@ -29,6 +29,8 @@
#include "binder_alloc.h"
#include "binder_trace.h"
#define BINDER_MIN_ALLOC (1 * PAGE_SIZE)
static DEFINE_MUTEX(binder_alloc_mmap_lock);
enum {
@@ -155,9 +157,9 @@ struct binder_buffer *binder_alloc_prepare_to_free(struct binder_alloc *alloc,
return NULL;
}
static int binder_update_page_range(struct binder_alloc *alloc, int allocate,
void *start, void *end,
struct vm_area_struct *vma)
static int __binder_update_page_range(struct binder_alloc *alloc, int allocate,
void *start, void *end,
struct vm_area_struct *vma)
{
void *page_addr;
unsigned long user_page_addr;
@@ -257,6 +259,20 @@ err_no_vma:
return vma ? -ENOMEM : -ESRCH;
}
static int binder_update_page_range(struct binder_alloc *alloc, int allocate,
void *start, void *end,
struct vm_area_struct *vma)
{
/*
* For regular updates, move up start if needed since MIN_ALLOC pages
* are always mapped
*/
if (start - alloc->buffer < BINDER_MIN_ALLOC)
start = alloc->buffer + BINDER_MIN_ALLOC;
return __binder_update_page_range(alloc, allocate, start, end, vma);
}
struct binder_buffer *binder_alloc_new_buf(struct binder_alloc *alloc,
size_t data_size,
size_t offsets_size,
@@ -587,8 +603,8 @@ int binder_alloc_mmap_handler(struct binder_alloc *alloc,
}
alloc->buffer_size = vma->vm_end - vma->vm_start;
if (binder_update_page_range(alloc, 1, alloc->buffer,
alloc->buffer + PAGE_SIZE, vma)) {
if (__binder_update_page_range(alloc, 1, alloc->buffer,
alloc->buffer + BINDER_MIN_ALLOC, vma)) {
ret = -ENOMEM;
failure_string = "alloc small buf";
goto err_alloc_small_buf_failed;