Hi Darrick,
Am So., 2. Dez. 2018 um 19:13 Uhr schrieb Darrick J. Wong
<[email protected]>:
> From: Darrick J. Wong <[email protected]>
>
> In commit 4721a601099, we tried to fix a problem wherein directio reads
> into a splice pipe will bounce EFAULT/EAGAIN all the way out to
> userspace by simulating a zero-byte short read. This happens because
> some directio read implementations (xfs) will call
> bio_iov_iter_get_pages to grab pipe buffer pages and issue asynchronous
> reads, but as soon as we run out of pipe buffers that _get_pages call
> returns EFAULT, which the splice code translates to EAGAIN and bounces
> out to userspace.
>
> In that commit, the iomap code catches the EFAULT and simulates a
> zero-byte read, but that causes assertion errors on regular splice reads
> because xfs doesn't allow short directio reads. This causes infinite
> splice() loops and assertion failures on generic/095 on overlayfs
> because xfs only permit total success or total failure of a directio
> operation. The underlying issue in the pipe splice code has now been
> fixed by changing the pipe splice loop to avoid avoid reading more data
> than there is space in the pipe.
>
> Therefore, it's no longer necessary to simulate the short directio, so
> remove the hack from iomap.
>
> Fixes: 4721a601099 ("iomap: dio data corruption and spurious errors when
> pipes fill")
> Reported-by: Amir Goldstein <[email protected]>
> Reviewed-by: Christoph Hellwig <[email protected]>
> Signed-off-by: Darrick J. Wong <[email protected]>
> ---
> v2: split into two patches per hch request
> ---
> fs/iomap.c | 9 ---------
> 1 file changed, 9 deletions(-)
>
> diff --git a/fs/iomap.c b/fs/iomap.c
> index 3ffb776fbebe..d6bc98ae8d35 100644
> --- a/fs/iomap.c
> +++ b/fs/iomap.c
> @@ -1877,15 +1877,6 @@ iomap_dio_rw(struct kiocb *iocb, struct iov_iter *iter,
> dio->wait_for_completion = true;
> ret = 0;
> }
> -
> - /*
> - * Splicing to pipes can fail on a full pipe. We have
> to
> - * swallow this to make it look like a short IO
> - * otherwise the higher splice layers will completely
> - * mishandle the error and stop moving data.
> - */
> - if (ret == -EFAULT)
> - ret = 0;
> break;
> }
> pos += ret;
I'm afraid this breaks the following test case on xfs and gfs2, the
two current users of iomap_dio_rw.
Here, the splice system call fails with errno = EAGAIN when trying to
"move data" from a file opened with O_DIRECT into a pipe.
The test case can be run with option -d to not use O_DIRECT, which
makes the test succeed.
The -r option switches from reading from the pipe sequentially to
reading concurrently with the splice, which doesn't change the
behavior.
Any thoughts?
Thanks,
Andreas
=================================== 8< ===================================
#define _GNU_SOURCE
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/wait.h>
#include <unistd.h>
#include <fcntl.h>
#include <err.h>
#include <stdlib.h>
#include <stdio.h>
#include <stdbool.h>
#include <string.h>
#include <errno.h>
#define SECTOR_SIZE 512
#define BUFFER_SIZE (150 * SECTOR_SIZE)
void read_from_pipe(int fd, const char *filename, size_t size)
{
char buffer[SECTOR_SIZE];
size_t sz;
ssize_t ret;
while (size) {
sz = size;
if (sz > sizeof buffer)
sz = sizeof buffer;
ret = read(fd, buffer, sz);
if (ret < 0)
err(1, "read: %s", filename);
if (ret == 0) {
fprintf(stderr, "read: %s: unexpected EOF\n", filename);
exit(1);
}
size -= sz;
}
}
void do_splice1(int fd, const char *filename, size_t size)
{
bool retried = false;
int pipefd[2];
if (pipe(pipefd) == -1)
err(1, "pipe");
while (size) {
ssize_t spliced;
spliced = splice(fd, NULL, pipefd[1], NULL, size, SPLICE_F_MOVE);
if (spliced == -1) {
if (errno == EAGAIN && !retried) {
retried = true;
fprintf(stderr, "retrying splice\n");
sleep(1);
continue;
}
err(1, "splice");
}
read_from_pipe(pipefd[0], filename, spliced);
size -= spliced;
}
close(pipefd[0]);
close(pipefd[1]);
}
void do_splice2(int fd, const char *filename, size_t size)
{
bool retried = false;
int pipefd[2];
int pid;
if (pipe(pipefd) == -1)
err(1, "pipe");
pid = fork();
if (pid == 0) {
close(pipefd[1]);
read_from_pipe(pipefd[0], filename, size);
exit(0);
} else {
close(pipefd[0]);
while (size) {
ssize_t spliced;
spliced = splice(fd, NULL, pipefd[1], NULL, size, SPLICE_F_MOVE);
if (spliced == -1) {
if (errno == EAGAIN && !retried) {
retried = true;
fprintf(stderr, "retrying splice\n");
sleep(1);
continue;
}
err(1, "splice");
}
size -= spliced;
}
close(pipefd[1]);
waitpid(pid, NULL, 0);
}
}
void usage(const char *argv0)
{
fprintf(stderr, "USAGE: %s [-rd] {filename}\n", basename(argv0));
exit(2);
}
int main(int argc, char *argv[])
{
void (*do_splice)(int fd, const char *filename, size_t size);
const char *filename;
char *buffer;
int opt, open_flags, fd;
ssize_t ret;
do_splice = do_splice1;
open_flags = O_CREAT | O_TRUNC | O_RDWR | O_DIRECT;
while ((opt = getopt(argc, argv, "rd")) != -1) {
switch(opt) {
case 'r':
do_splice = do_splice2;
break;
case 'd':
open_flags &= ~O_DIRECT;
break;
default: /* '?' */
usage(argv[0]);
}
}
if (optind >= argc)
usage(argv[0]);
filename = argv[optind];
printf("%s reader %s O_DIRECT\n",
do_splice == do_splice1 ? "sequential" : "concurrent",
(open_flags & O_DIRECT) ? "with" : "without");
buffer = aligned_alloc(SECTOR_SIZE, BUFFER_SIZE);
if (buffer == NULL)
err(1, "aligned_alloc");
fd = open(filename, open_flags, 0666);
if (fd == -1)
err(1, "open: %s", filename);
memset(buffer, 'x', BUFFER_SIZE);
ret = write(fd, buffer, BUFFER_SIZE);
if (ret < 0)
err(1, "write: %s", filename);
if (ret != BUFFER_SIZE) {
fprintf(stderr, "%s: short write\n", filename);
exit(1);
}
ret = lseek(fd, 0, SEEK_SET);
if (ret != 0)
err(1, "lseek: %s", filename);
do_splice(fd, filename, BUFFER_SIZE);
if (unlink(filename) == -1)
err(1, "unlink: %s", filename);
return 0;
}
=================================== 8< ===================================
