summaryrefslogtreecommitdiff
path: root/fs
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2015-11-01 16:43:24 -0800
committerLinus Torvalds <torvalds@linux-foundation.org>2015-11-01 16:43:24 -0800
commit2e002662973fd8d67d5a760776a5d3ea3d3399a9 (patch)
tree4c6feb1e7a10d602ef7c3f9e017cad48e52cc8a1 /fs
parent6a13feb9c82803e2b815eca72fa7a9f5561d7861 (diff)
parentfc90888d07b8e17eec49c04bdb26344fdea96c3b (diff)
Merge branch 'fs-file-descriptor-optimization'
Merge file descriptor allocation speedup. Eric Dumazet has a test-case for a fairly common network deamon load pattern: openign and closing a lot of sockets that each have very little work done on them. It turns out that in that case, the cost of just finding the correct file descriptor number can be a dominating factor. We've long had a trivial optimization for allocating file descriptors sequentially, but that optimization ends up being not very effective when other file descriptors are being closed concurrently, and the fd patterns are not some simple FIFO pattern. In such cases we ended up spending a lot of time just scanning the bitmap of open file descriptors in order to find the next file descriptor number to open. This trivial patch-series mitigates that by simply introducing a second-level bitmap of which words in the first bitmap are already fully allocated. That cuts down the cost of scanning by an order of magnitude in some pathological (but realistic) cases. The second patch is an even more trivial patch to avoid unnecessarily dirtying the cacheline for the close-on-exec bit array that normally ends up being all empty. * fs-file-descriptor-optimization: vfs: conditionally clear close-on-exec flag vfs: Fix pathological performance case for __alloc_fd()
Diffstat (limited to 'fs')
-rw-r--r--fs/file.c42
1 files changed, 37 insertions, 5 deletions
diff --git a/fs/file.c b/fs/file.c
index 6c672ad329e9..c6986dce0334 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -56,6 +56,9 @@ static void free_fdtable_rcu(struct rcu_head *rcu)
__free_fdtable(container_of(rcu, struct fdtable, rcu));
}
+#define BITBIT_NR(nr) BITS_TO_LONGS(BITS_TO_LONGS(nr))
+#define BITBIT_SIZE(nr) (BITBIT_NR(nr) * sizeof(long))
+
/*
* Expand the fdset in the files_struct. Called with the files spinlock
* held for write.
@@ -77,6 +80,11 @@ static void copy_fdtable(struct fdtable *nfdt, struct fdtable *ofdt)
memset((char *)(nfdt->open_fds) + cpy, 0, set);
memcpy(nfdt->close_on_exec, ofdt->close_on_exec, cpy);
memset((char *)(nfdt->close_on_exec) + cpy, 0, set);
+
+ cpy = BITBIT_SIZE(ofdt->max_fds);
+ set = BITBIT_SIZE(nfdt->max_fds) - cpy;
+ memcpy(nfdt->full_fds_bits, ofdt->full_fds_bits, cpy);
+ memset(cpy+(char *)nfdt->full_fds_bits, 0, set);
}
static struct fdtable * alloc_fdtable(unsigned int nr)
@@ -115,12 +123,14 @@ static struct fdtable * alloc_fdtable(unsigned int nr)
fdt->fd = data;
data = alloc_fdmem(max_t(size_t,
- 2 * nr / BITS_PER_BYTE, L1_CACHE_BYTES));
+ 2 * nr / BITS_PER_BYTE + BITBIT_SIZE(nr), L1_CACHE_BYTES));
if (!data)
goto out_arr;
fdt->open_fds = data;
data += nr / BITS_PER_BYTE;
fdt->close_on_exec = data;
+ data += nr / BITS_PER_BYTE;
+ fdt->full_fds_bits = data;
return fdt;
@@ -226,17 +236,22 @@ static inline void __set_close_on_exec(int fd, struct fdtable *fdt)
static inline void __clear_close_on_exec(int fd, struct fdtable *fdt)
{
- __clear_bit(fd, fdt->close_on_exec);
+ if (test_bit(fd, fdt->close_on_exec))
+ __clear_bit(fd, fdt->close_on_exec);
}
-static inline void __set_open_fd(int fd, struct fdtable *fdt)
+static inline void __set_open_fd(unsigned int fd, struct fdtable *fdt)
{
__set_bit(fd, fdt->open_fds);
+ fd /= BITS_PER_LONG;
+ if (!~fdt->open_fds[fd])
+ __set_bit(fd, fdt->full_fds_bits);
}
-static inline void __clear_open_fd(int fd, struct fdtable *fdt)
+static inline void __clear_open_fd(unsigned int fd, struct fdtable *fdt)
{
__clear_bit(fd, fdt->open_fds);
+ __clear_bit(fd / BITS_PER_LONG, fdt->full_fds_bits);
}
static int count_open_files(struct fdtable *fdt)
@@ -280,6 +295,7 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
new_fdt->max_fds = NR_OPEN_DEFAULT;
new_fdt->close_on_exec = newf->close_on_exec_init;
new_fdt->open_fds = newf->open_fds_init;
+ new_fdt->full_fds_bits = newf->full_fds_bits_init;
new_fdt->fd = &newf->fd_array[0];
spin_lock(&oldf->file_lock);
@@ -323,6 +339,7 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
memcpy(new_fdt->open_fds, old_fdt->open_fds, open_files / 8);
memcpy(new_fdt->close_on_exec, old_fdt->close_on_exec, open_files / 8);
+ memcpy(new_fdt->full_fds_bits, old_fdt->full_fds_bits, BITBIT_SIZE(open_files));
for (i = open_files; i != 0; i--) {
struct file *f = *old_fds++;
@@ -454,10 +471,25 @@ struct files_struct init_files = {
.fd = &init_files.fd_array[0],
.close_on_exec = init_files.close_on_exec_init,
.open_fds = init_files.open_fds_init,
+ .full_fds_bits = init_files.full_fds_bits_init,
},
.file_lock = __SPIN_LOCK_UNLOCKED(init_files.file_lock),
};
+static unsigned long find_next_fd(struct fdtable *fdt, unsigned long start)
+{
+ unsigned long maxfd = fdt->max_fds;
+ unsigned long maxbit = maxfd / BITS_PER_LONG;
+ unsigned long bitbit = start / BITS_PER_LONG;
+
+ bitbit = find_next_zero_bit(fdt->full_fds_bits, maxbit, bitbit) * BITS_PER_LONG;
+ if (bitbit > maxfd)
+ return maxfd;
+ if (bitbit > start)
+ start = bitbit;
+ return find_next_zero_bit(fdt->open_fds, maxfd, start);
+}
+
/*
* allocate a file descriptor, mark it busy.
*/
@@ -476,7 +508,7 @@ repeat:
fd = files->next_fd;
if (fd < fdt->max_fds)
- fd = find_next_zero_bit(fdt->open_fds, fdt->max_fds, fd);
+ fd = find_next_fd(fdt, fd);
/*
* N.B. For clone tasks sharing a files structure, this test