Age | Commit message (Collapse) | Author |
|
Introduce the smp_load_acquire() and smp_store_release() macros. Fix
synchronization in io_uring_cq_advance() and __io_uring_get_cqe().
Remove a superfluous local variable, if-test and write barrier from
__io_uring_submit(). Remove a superfluous barrier from
test/io_uring_enter.c.
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Use io_uring_for_each_cqe() inside __io_uring_get_cqe() such that it
becomes possible to test the io_uring_for_each_cqe() implementation
from inside the liburing project.
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Don't check for IORING_SQ_NEED_WAKEUP twice, put it inside
sq_ring_needs_enter().
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We don't modify them, make that clear.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This example misses an increment though the iovecs array. This causes
the same buffer to be filled from the block device every time. It would
be good to fix this since it is one of the first examples a new-comer to
io_uring is exposed too.
Signed-off-by: Stephen Bates <sbates@raithlin.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Works just like io_uring_submit(), but also allows retrieving events
(or waiting/polling for them) in the same call.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Don't leak the memory. We never need to reference it from the
test app once the io_uring_register() call is done.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Ring memory is released by umapping {cq,sq}->khead, which assumes that
p->{cq,sq}_off.head will always be 0.
Add another field to for the mmaped region and use that instead of
->khead when umapping.
Signed-off-by: Kornilios Kourtis <kkourt@kkourt.io>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_uring_queue_mmap wasn't exported from the ld map file, meaning the
linker gave it private, static linkage in the resulting shared object,
as opposed to global linkage. This results in the symbol seemingly
missing at link-time, despite the existence of the symbol in the library
itself.
Signed-off-by: Austin Seipp <aseipp@pobox.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_uring_cqe_seen() contains write memory barrier (at least for !x86),
so calling it can be very expensive for a big number of cqes.
Would be better to have a loop over all available cqes and only then
advance the cq ring once, e.g.:
struct io_uring_cqe *cqe;
unsigned nr, head;
nr = 0;
io_uring_for_each_cqe(&ring, head, cqe) {
/* handle cqe */
if (++nr == max)
break;
}
io_uring_cq_advance(&ring, nr);
Signed-off-by: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_uring_submit() is a last commit point when a caller has to be sure
that all prepared sqes eventually lands to the kernel.
If SQ thread is running it is quite possible that khead != ktail, thus
charging of sq->array is skipped, which leads to IO hang.
Nasty hang is well reproduced when submitter and completion harverster
are different threads.
Signed-off-by: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Enter kernel only if SQ thread is off or wakeup is needed.
Signed-off-by: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
pkgconfig file is installed but not referenced in the
rpm's spec file so rpmbuild fails.
Signed-off-by: Hrvoje Zeba <zeba.hrvoje@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
libdir is hardcoded to ${prefix}/lib in Makefile. Fedora x86_64 uses
/usr/lib64 and this means libaries will be installed in the wrong place.
This patch moves prefix, includedir, libdir, and mandir into ./configure
for easier customization. To build and install on Fedora x86_64:
# ./configure --libdir=/usr/lib64
# make && make install
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
pkg-config (https://pkgconfig.freedesktop.org/) makes it easier to build
applications that have library dependencies. Libraries ship .pc files
containing the compiler and linker flags needed to build successfully.
This saves applications from hardcoding these details into their build
scripts, especially when these details can change between operating
systems or distributions.
To build a liburing application:
gcc $(pkg-config --cflags --libs liburing) -o myapp myapp.c
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This syncs with io_uring-next, which has support for doing recvmsg
and sendmsg through io_uring.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
In case malloc fails, the fix returns NULL to avoid NULL pointer
dereference.
Signed-off-by: zhangliguang <zhangliguang@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Use proper unsigned math to figure out how many entries we have. If
we have head and tail on either side of UINT_MAX, then we currently
don't submit anything as:
while (head < tail) {
is never true.
Reported-by: Weiping Zhang <zhangweiping@didiglobal.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
If the ring setup fails, close the fd before returning failure.
Reported-by: Kornilios Kourtis <kkourt@kkourt.io>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We really should use a ';' to terminate the line. This is just a
cosmetic issue, no functional changes.
Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
Modified wording
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Some of the symbols were missing from linker version
script so they weren't exported in the dynamic library.
Signed-off-by: Hrvoje Zeba <zeba.hrvoje@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Fortunately, all architectures except alpha have common numbers
for new system calls nowadays, so support of non-alpha architectures
is as simple as support of x86.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Rename __NR_sys_io_uring* to __NR_io_uring* to match kernel headers.
The prefix traditionally used for macros describing syscall numbers
is "__NR_", not "__NR_sys_".
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
* 'const-matching' of https://github.com/flamingcow66/liburing:
Match const signatures in io_uring_prep_* to native functions
|
|
io_uring_prep_{readv,writev,read_fixed,write_fixed}() take
non-const arguments that are const in the native functions
that they mimic. Make those arguments const.
Signed-off-by: Ian Gulliver <git@flamingcow.io>
|
|
* 'missing-header' of https://github.com/flamingcow66/liburing:
Fix liburing.h memset() missing declaration
|
|
liburing.h uses memset() but doesn't include string.h.
If nothing else includes it, this causes compile errors.
Add the include.
Signed-off-by: Ian Gulliver <git@flamingcow.io>
|
|
Add sync_range_file opcode and related fields, and the
io_uring_register(2) commands for eventfd registration.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The examples in the io_uring_setup.2 man patch erroneously ORs
the protection bits and mmap flags, fix it up.
Signed-off-by: Ian Gulliver <git@flamingcow.io>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Also kill reference to long dead TODO file.
Patch from Hrvoje Zeba <zeba.hrvoje@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Put the data at the end so we don't have to offset the actual buffer.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Add short read requeue and abort on error.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Still need to handle broken chains, must resubmit those.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We need a lot of RAM for this case, my vm always fails it. Don't
hard fail due to ENOMEM, just skip the test.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Just a basic test case that does various forms of linked nops, and
a sample bare bones copy program using linked reads and writes.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Make header includable from C++ projects.
Signed-off-by: Kevin Vigor <kvigor@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Static libraries are generally built without -fPIC for slight
performance gain, make that the case for liburing.
Signed-off-by: Kevin Vigor <kvigor@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|