This decreases the number of syscalls per read/write by half.
Signed-off-by: Anton Ivanov
---
arch/um/drivers/ubd_kern.c | 27 +--
arch/um/include/shared/os.h | 2 ++
arch/um/os-Linux/file.c | 19 +++
3 files changed, 26
Modifying RLIMIT requires including sys/time.h and sys/resource.h
Signed-off-by: Anton Ivanov
---
arch/um/os-Linux/start_up.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/arch/um/os-Linux/start_up.c b/arch/um/os-Linux/start_up.c
index 425162e..c1811df 100644
---
Hi Richard,
This is vs your tree, 4.x master branch in linus repo actually has
sys/resource.h
Pls ignore the patch.
A.
On 08/11/15 14:58, Anton Ivanov wrote:
> Modifying RLIMIT requires including sys/time.h and sys/resource.h
>
> Signed-off-by: Anton Ivanov
> ---
>
I am going to send in a couple of more which bulk-up the transactions
(this forms most of the improvement) later this week.
This one is obvious - qemu has been using it for ages.
A.
On 08/11/15 15:20, Anton Ivanov wrote:
> This decreases the number of syscalls per read/write by half.
>
>
No worries.
A completely rewritten IRQ controller patch is coming in ~ 15 minutes
or so. 20% gain across the board for a basic config, up to 100s% gain if
you register large numbers of devices (in the tens).
Just finishing to run it through its paces :)
It is also a pre-requisite to do
On Sun, Nov 8, 2015 at 4:18 PM, Anton Ivanov wrote:
> Hi Richard,
>
> This is vs your tree, 4.x master branch in linus repo actually has
> sys/resource.h
>
> Pls ignore the patch.
Sorry for the trouble!
I forgot to mention that linux-next is the only valid branch in my tree.
Epoll based interrupt controller.
IMPROVES: IO loop performance - no per fd lookups, allowing for
15% IO speedup in minimal config going to 100s of % with many
devices - a N^N lookup is now replaced by a log(N)
ADDS: True Write IRQ functionality
OBSOLETES: The need to call reactivate_fd() in
This works cleanly and is understandable (something I would not say
about the original version I wrote a couple of years back).
It emits some minor nags on shutdown related to cleaning up the term
descriptors, but as far as I can see they are mostly harmless. I suspect
that once I do an
I just noted one minor issue with it (which existed in the earlier
version as well) - it leaks one FD per reboot.
I will fix it later on during the week in a revised version.
A.
On 08/11/15 22:50, Anton Ivanov wrote:
> Epoll based interrupt controller.
>
> IMPROVES: IO loop performance - no