aboutsummaryrefslogtreecommitdiffstats
path: root/fs/pipe.c
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2012-04-29 16:12:42 -0400
committerLinus Torvalds <torvalds@linux-foundation.org>2012-04-29 16:12:42 -0400
commit9883035ae7edef3ec62ad215611cb8e17d6a1a5d (patch)
treeab4afff1603e0f1e85e349b8a1fdb8415cc457cf /fs/pipe.c
parentde9e24eda331bbefb9195a4d646c786bdcbba7d4 (diff)
pipes: add a "packetized pipe" mode for writing
The actual internal pipe implementation is already really about individual packets (called "pipe buffers"), and this simply exposes that as a special packetized mode. When we are in the packetized mode (marked by O_DIRECT as suggested by Alan Cox), a write() on a pipe will not merge the new data with previous writes, so each write will get a pipe buffer of its own. The pipe buffer is then marked with the PIPE_BUF_FLAG_PACKET flag, which in turn will tell the reader side to break the read at that boundary (and throw away any partial packet contents that do not fit in the read buffer). End result: as long as you do writes less than PIPE_BUF in size (so that the pipe doesn't have to split them up), you can now treat the pipe as a packet interface, where each read() system call will read one packet at a time. You can just use a sufficiently big read buffer (PIPE_BUF is sufficient, since bigger than that doesn't guarantee atomicity anyway), and the return value of the read() will naturally give you the size of the packet. NOTE! We do not support zero-sized packets, and zero-sized reads and writes to a pipe continue to be no-ops. Also note that big packets will currently be split at write time, but that the size at which that happens is not really specified (except that it's bigger than PIPE_BUF). Currently that limit is the system page size, but we might want to explicitly support bigger packets some day. The main user for this is going to be the autofs packet interface, allowing us to stop having to care so deeply about exact packet sizes (which have had bugs with 32/64-bit compatibility modes). But user space can create packetized pipes with "pipe2(fd, O_DIRECT)", which will fail with an EINVAL on kernels that do not support this interface. Tested-by: Michael Tokarev <mjt@tls.msk.ru> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: David Miller <davem@davemloft.net> Cc: Ian Kent <raven@themaw.net> Cc: Thomas Meyer <thomas@m3y3r.de> Cc: stable@kernel.org # needed for systemd/autofs interaction fix Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'fs/pipe.c')
-rw-r--r--fs/pipe.c31
1 files changed, 29 insertions, 2 deletions
diff --git a/fs/pipe.c b/fs/pipe.c
index 25feaa3faac..fec5e4ad071 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -346,6 +346,16 @@ static const struct pipe_buf_operations anon_pipe_buf_ops = {
346 .get = generic_pipe_buf_get, 346 .get = generic_pipe_buf_get,
347}; 347};
348 348
349static const struct pipe_buf_operations packet_pipe_buf_ops = {
350 .can_merge = 0,
351 .map = generic_pipe_buf_map,
352 .unmap = generic_pipe_buf_unmap,
353 .confirm = generic_pipe_buf_confirm,
354 .release = anon_pipe_buf_release,
355 .steal = generic_pipe_buf_steal,
356 .get = generic_pipe_buf_get,
357};
358
349static ssize_t 359static ssize_t
350pipe_read(struct kiocb *iocb, const struct iovec *_iov, 360pipe_read(struct kiocb *iocb, const struct iovec *_iov,
351 unsigned long nr_segs, loff_t pos) 361 unsigned long nr_segs, loff_t pos)
@@ -407,6 +417,13 @@ redo:
407 ret += chars; 417 ret += chars;
408 buf->offset += chars; 418 buf->offset += chars;
409 buf->len -= chars; 419 buf->len -= chars;
420
421 /* Was it a packet buffer? Clean up and exit */
422 if (buf->flags & PIPE_BUF_FLAG_PACKET) {
423 total_len = chars;
424 buf->len = 0;
425 }
426
410 if (!buf->len) { 427 if (!buf->len) {
411 buf->ops = NULL; 428 buf->ops = NULL;
412 ops->release(pipe, buf); 429 ops->release(pipe, buf);
@@ -459,6 +476,11 @@ redo:
459 return ret; 476 return ret;
460} 477}
461 478
479static inline int is_packetized(struct file *file)
480{
481 return (file->f_flags & O_DIRECT) != 0;
482}
483
462static ssize_t 484static ssize_t
463pipe_write(struct kiocb *iocb, const struct iovec *_iov, 485pipe_write(struct kiocb *iocb, const struct iovec *_iov,
464 unsigned long nr_segs, loff_t ppos) 486 unsigned long nr_segs, loff_t ppos)
@@ -593,6 +615,11 @@ redo2:
593 buf->ops = &anon_pipe_buf_ops; 615 buf->ops = &anon_pipe_buf_ops;
594 buf->offset = 0; 616 buf->offset = 0;
595 buf->len = chars; 617 buf->len = chars;
618 buf->flags = 0;
619 if (is_packetized(filp)) {
620 buf->ops = &packet_pipe_buf_ops;
621 buf->flags = PIPE_BUF_FLAG_PACKET;
622 }
596 pipe->nrbufs = ++bufs; 623 pipe->nrbufs = ++bufs;
597 pipe->tmp_page = NULL; 624 pipe->tmp_page = NULL;
598 625
@@ -1013,7 +1040,7 @@ struct file *create_write_pipe(int flags)
1013 goto err_dentry; 1040 goto err_dentry;
1014 f->f_mapping = inode->i_mapping; 1041 f->f_mapping = inode->i_mapping;
1015 1042
1016 f->f_flags = O_WRONLY | (flags & O_NONBLOCK); 1043 f->f_flags = O_WRONLY | (flags & (O_NONBLOCK | O_DIRECT));
1017 f->f_version = 0; 1044 f->f_version = 0;
1018 1045
1019 return f; 1046 return f;
@@ -1057,7 +1084,7 @@ int do_pipe_flags(int *fd, int flags)
1057 int error; 1084 int error;
1058 int fdw, fdr; 1085 int fdw, fdr;
1059 1086
1060 if (flags & ~(O_CLOEXEC | O_NONBLOCK)) 1087 if (flags & ~(O_CLOEXEC | O_NONBLOCK | O_DIRECT))
1061 return -EINVAL; 1088 return -EINVAL;
1062 1089
1063 fw = create_write_pipe(flags); 1090 fw = create_write_pipe(flags);