On 05/02/2015 15:17, Michael Kerrisk (man-pages) wrote:
> On 20 January 2014 at 18:28, Eliezer Tamir
> wrote:
>> On 20/01/2014 18:28, Michael Kerrisk (man-pages) wrote:
>>> On 07/10/2013 04:18 PM, Eliezer Tamir wrote:
>>>> Add description for the SO_BUSY_P
On 03/09/2014 10:51, Michael S. Tsirkin wrote:
> On Wed, Sep 03, 2014 at 09:49:10AM +0300, Eliezer Tamir wrote:
>> On 02/09/2014 11:31, Michael S. Tsirkin wrote:
>>> On Tue, Sep 02, 2014 at 09:15:18AM +0300, Eliezer Tamir wrote:
>> Busy polling is not a general purpose fea
On 02/09/2014 11:31, Michael S. Tsirkin wrote:
> On Tue, Sep 02, 2014 at 09:15:18AM +0300, Eliezer Tamir wrote:
>> On 02/09/2014 06:29, Jason Wang wrote:
>>> On 09/01/2014 02:39 PM, Eliezer Tamir wrote:
>>>> On 29/08/2014 06:08, Jason Wang wrote:
>>>>
On 02/09/2014 09:31, Jason Wang wrote:
> On 09/02/2014 02:03 PM, Eliezer Tamir wrote:
>> On 02/09/2014 06:35, Jason Wang wrote:
>>> Not sure whether this method will scale considering thousands of sockets
>>> and processes.
>> There may be millions of sockets, bu
On 02/09/2014 06:29, Jason Wang wrote:
> On 09/01/2014 02:39 PM, Eliezer Tamir wrote:
>> On 29/08/2014 06:08, Jason Wang wrote:
>>>> Yes, but rx busy polling only works in process context and does not
>>>> disable bh, so it may be not an issue.
>> sk_busy_loo
On 02/09/2014 06:35, Jason Wang wrote:
> On 09/01/2014 02:55 PM, Eliezer Tamir wrote:
>> On 26/08/2014 10:16, Jason Wang wrote:
>>> On 08/25/2014 09:16 PM, Eliezer Tamir wrote:
>> Think about the case where two processes are busy polling on the
>> same CPU and the
On 26/08/2014 10:16, Jason Wang wrote:
> On 08/25/2014 09:16 PM, Eliezer Tamir wrote:
>> Here are my 2 cents:
>> I think Ingo's suggestion of only yielding to tasks with same or higher
>> priority makes sense.
>
> I'm not sure I get your meaning. Do you
On 29/08/2014 06:08, Jason Wang wrote:
> Yes, but rx busy polling only works in process context and does not
> disable bh, so it may be not an issue.
sk_busy_loop() uses rcu_read_lock_bh(), so it does run with bh disabled.
-Eliezer
--
To unsubscribe from this list: send the line "unsubscribe lin
;>
>> And add warns to make sure it can only be called in process context.
>
>
> 1) Any reasons Eliezer Tamir is not included in the CC list ?
Thanks for remembering me, Eric ;)
Here are my 2 cents:
I think Ingo's suggestion of only yielding to tasks with same or higher
On 20/08/2014 14:43, Jiri Slaby wrote:
> From: Jacob Keller
>
> 3.12-stable review patch. If anyone has any objections, please let me know.
No objection, I am however unable to test this myself at the moment.
I would suggest that someone takes a look at other drivers that support
busypoll, to
On 20/01/2014 18:28, Michael Kerrisk (man-pages) wrote:
> On 07/10/2013 04:18 PM, Eliezer Tamir wrote:
>> Add description for the SO_BUSY_POLL socket option to the socket(7) manpage.
>
> Long after the fact, I've applied this. Thanks, Eliezer.
>
> Would you be willing
On 17/12/2013 17:13, Peter Zijlstra wrote:
> On Tue, Dec 17, 2013 at 04:03:58PM +0200, Eliezer Tamir wrote:
>>
>> I'm not sure that subtracting the spin time is the optimal thing to do.
>>
>> The busy poll time is supposed to be limited to something less than 1ms.
n 17/12/2013 15:32, Peter Zijlstra wrote:
>
> Awesomeness.. you'll work on subtracting the spin time from the sleep
> time?
Me or someone on our team will work on it.
I'm not sure that subtracting the spin time is the optimal thing to do.
The busy poll time is supposed to be limited to somethin
On 13/12/2013 15:56, Peter Zijlstra wrote:
> On Fri, Dec 13, 2013 at 12:49:01PM +0200, Eliezer Tamir wrote:
>> On 12/12/2013 16:08, Peter Zijlstra wrote:
>>> This series contains the preempt_enable_no_resched() cleanups that include
>>> spin_lock_bh() optimizations and
On 12/12/2013 16:08, Peter Zijlstra wrote:
> This series contains the preempt_enable_no_resched() cleanups that include
> spin_lock_bh() optimizations and local_clock() optimizations.
I'm trying to test this on tip/master.
Patch 3 fails to apply on kenrel/softirq.c
Thanks,
Eliezer
--
To unsubscri
On 28/11/2013 19:40, Peter Zijlstra wrote:
> On Thu, Nov 28, 2013 at 06:49:00PM +0200, Eliezer Tamir wrote:
>> I have tested this patch and I see a performance regression of about
>> 1.5%.
>
> Cute, can you qualify your metric? Since this is a poll loop the only
> metric t
On 26/11/2013 17:57, Peter Zijlstra wrote:
>
> Replace sched_clock() usage with local_clock() which has a bounded
> drift between CPUs (<2 jiffies).
>
Peter,
I have tested this patch and I see a performance regression of about
1.5%.
Maybe it would be better, rather then testing in the fast pat
On 22/11/2013 13:30, Peter Zijlstra wrote:
> On Fri, Nov 22, 2013 at 08:56:00AM +0200, Eliezer Tamir wrote:
>> On 21/11/2013 15:39, Peter Zijlstra wrote:
>>> On Thu, Nov 21, 2013 at 03:26:17PM +0200, Eliezer Tamir wrote:
>
> Please use local_clock(), yes its slightly mor
On 21/11/2013 15:39, Peter Zijlstra wrote:
> On Thu, Nov 21, 2013 at 03:26:17PM +0200, Eliezer Tamir wrote:
>> We don't override any limit the user has put on the system call.
>
> You are in fact, note how the normal select @endtime argument is only
> set up _after_ you
On 20/11/2013 18:04, Peter Zijlstra wrote:
> The only valid use of preempt_enable_no_resched() is if the very next
> line is schedule() or if we know preemption cannot actually be enabled
> by that statement due to known more preempt_count 'refs'.
The reason I used the no resched version is that b
On 26/08/2013 00:30, Amir Vadai wrote:
> I'm on vacation, will test and have some inputs later this week when I be
> back.
>
Hello Amir,
Ping me when you get back and I will send you my latest so you can
play with it.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
t
Add a cpu_relaxt to sk_busy_loop.
Julie Cummings reported performance issues when hyperthreading is on.
Arjan van de Ven observed that we should have a cpu_relax() in the
busy poll loop.
Reported-by: Julie Cummings
Signed-off-by: Eliezer Tamir
---
include/net/busy_poll.h |1 +
1 files
On 22/08/2013 23:11, Eric Wong wrote:
> Eliezer Tamir wrote:
>> Performance:
>> using sockperf, Intel X520 NICs,
>> Supermicro 6026TT-BTF systems with E5-2690 Xeon CPUs
>> 100 UDP sockets avg. latency 5.756 (std-dev 0.510)
>> 1k UDP sockets avg. latency 5.780 (
On 22/08/2013 23:14, David Miller wrote:
> From: Eliezer Tamir
> Date: Wed, 21 Aug 2013 13:39:54 +0300
>
>> +SK_LL_STATE_MISS_2, /* data came through napi poll twice in a row */
>
> Please rename this to SK_LL_STATE_MISS_MULTI or something like that.
>
> Th
On 21/08/2013 16:53, Eric Dumazet wrote:
> On Wed, 2013-08-21 at 13:39 +0300, Eliezer Tamir wrote:
>
>>
>> Instead of remembering the napi_id for all the sockets in an epoll,
>> we only track the first socket we see with any unique napi_id.
>> The rational for this i
e event handling algorithm.
Willem de Brujin: advice on style, where to put things, RCU.
Special thanks for finding bugs in earlier versions: Julie Cummings.
Signed-off-by: Eliezer Tamir
---
fs/eventpoll.c | 131 +
include/linux/poll.h|2
include
Rename mib counter from "low latency" to "busy poll"
v1 also moved the counter to the ip MIB (suggested by Shawn Bohrer)
Eric Dumazet suggested that the current location is better.
So v2 just renames the counter to fit the new naming convention.
Signed-off-by: Eliezer Tami
On 06/08/2013 13:14, Eric Dumazet wrote:
> On Tue, 2013-08-06 at 12:52 +0300, Eliezer Tamir wrote:
>> Move the low latency mib counter to the ip section.
>> Rename it from low latency to busy poll.
>>
>> Reported-by: Shawn Bohrer
>> Signed-off-by: Eliezer Tamir
&
Move the low latency mib counter to the ip section.
Rename it from low latency to busy poll.
Reported-by: Shawn Bohrer
Signed-off-by: Eliezer Tamir
---
include/net/busy_poll.h |4 ++--
include/uapi/linux/snmp.h |2 +-
net/ipv4/proc.c |2 +-
3 files changed, 4
placeholder.
Signed-off-by: Eliezer Tamir
---
include/net/busy_poll.h |5 -
1 files changed, 0 insertions(+), 5 deletions(-)
diff --git a/include/net/busy_poll.h b/include/net/busy_poll.h
index f18b919..8e2dfc1 100644
--- a/include/net/busy_poll.h
+++ b/include/net/busy_poll.h
@@ -162,11 +162,6
On 10/07/2013 21:03, Michael Kerrisk wrote:
> On 07/10/13 16:18, Eliezer Tamir wrote:
>> Add description for the SO_BUSY_POLL socket option to the socket(7) manpage.
>
> What's the status of this feature. A quick grep seems to show
> that the feature is not yet in mainline.
Rename LL_SO to BUSY_POLL_SO
Rename sysctl_net_ll_{read,poll} to sysctl_busy_{read,poll}
Fix up users of these variables.
Fix documentation for sysctl.
a patch for the socket.7 man page will follow separately,
because of limitations of my mail setup.
Signed-off-by: Eliezer Tamir
Add description for the SO_BUSY_POLL socket option to the socket(7) manpage.
v2
fixed typos reported by Rasmus Villemoes
Signed-off-by: Eliezer Tamir
---
man7/socket.7 | 25 +
1 files changed, 25 insertions(+), 0 deletions(-)
diff --git a/man7/socket.7 b/man7/socket
Rename the file and correct all the places where it is included.
Signed-off-by: Eliezer Tamir
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c |2
drivers/net/ethernet/intel/ixgbe/ixgbe.h|2
drivers/net/ethernet/mellanox/mlx4/en_netdev.c |2
drivers/net/ethernet
Rename ndo_ll_poll to ndo_busy_poll.
Rename sk_mark_ll to sk_mark_napi_id.
Rename skb_mark_ll to skb_mark_napi_id.
Correct all useres of these functions.
Update comments and defines in include/net/busy_poll.h
Signed-off-by: Eliezer Tamir
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
David,
Here are three patches that complete the rename of lls to busy-poll
1. rename include/net/ll_poll.h to include/net/busy_poll.h
2. Rename ndo_ll_poll to ndo_busy_poll.
Rename sk_mark_ll to sk_mark_napi_id.
Rename skb_mark_ll to skb_mark_napi_id.
Correct all useres of these function
On 10/07/2013 16:17, Arend van Spriel wrote:
> On 07/10/2013 02:13 PM, Eliezer Tamir wrote:
>> Rename the file and correct all the places where it is included.
>> Signed-off-by: Eliezer Tamir
>> ---
>> include/net/busy_poll.h
Add description for the SO_BUSY_POLL socket option to the socket(7) manpage.
Signed-off-by: Eliezer Tamir
---
man7/socket.7 | 25 +
1 files changed, 25 insertions(+), 0 deletions(-)
diff --git a/man7/socket.7 b/man7/socket.7
index f2213eb..27603c9 100644
--- a/man7
Rename the file and correct all the places where it is included.
Signed-off-by: Eliezer Tamir
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c |2
drivers/net/ethernet/intel/ixgbe/ixgbe.h|2
drivers/net/ethernet/mellanox/mlx4/en_netdev.c |2
drivers/net/ethernet
: Eliezer Tamir
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c |2 +-
drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c |2 +-
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c|4 ++--
drivers/net/ethernet/mellanox/mlx4/en_netdev.c |2 +-
drivers/net/ethernet/mellanox/mlx4
Rename LL_SO to BUSY_POLL_SO
Rename sysctl_net_ll_{read,poll} to sysctl_busy_{read,poll}
Fix up users of these variables.
Fix documentation for sysctl.
a patch for the socket.7 man page will follow separately,
because of limitations of my mail setup.
Signed-off-by: Eliezer Tamir
Rename ndo_ll_poll to ndo_busy_poll.
Rename sk_mark_ll to sk_mark_napi_id.
Rename skb_mark_ll to skb_mark_napi_id.
Correct all useres of these functions.
Update comments and defines in include/net/busy_poll.h
Signed-off-by: Eliezer Tamir
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
Rename the file and correct all the places where it is included.
Signed-off-by: Eliezer Tamir
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c |2
drivers/net/ethernet/intel/ixgbe/ixgbe.h|2
drivers/net/ethernet/mellanox/mlx4/en_netdev.c |2
drivers/net/ethernet
Rename LL_SO to BUSY_POLL_SO
Rename sysctl_net_ll_{read,poll} to sysctl_busy_{read,poll}
Fix up users of these variables.
Fix documentation for sysctl.
a patch for the socket.7 man page will follow separately,
because of limitations of my mail setup.
Signed-off-by: Eliezer Tamir
On 10/07/2013 07:41, David Miller wrote:
> From: Eliezer Tamir
> Date: Wed, 10 Jul 2013 06:29:16 +0300
>
>> If the following names changes are acceptable I will try to send out
>> a patch today.
>> 2. ndo_ll_poll -> ndo_busy_poll
>>
>> - not technicall
On 10/07/2013 02:06, David Miller wrote:
> From: Jonathan Corbet
> Date: Tue, 9 Jul 2013 16:25:14 -0600
>
>> On Mon, 08 Jul 2013 16:20:34 +0300
>> Eliezer Tamir wrote:
>>
>>> Rename POLL_LL to POLL_BUSY_LOOP.
>>
>> So pardon me if I speak out of
.
Reorder poll/select busy loop conditions:
- Clear busy_flag after one time we can't busy-poll.
- Only init busy_end if we actually are going to busy-poll.
Added one more missing need_resched() test.
Signed-off-by: Eliezer Tamir
---
fs/select.c |
ually are going to busy-poll.
Added one more missing need_resched() test.
Signed-off-by: Eliezer Tamir
---
fs/select.c | 31 ++--
include/net/ll_poll.h | 55 +++--
2 files changed, 37 insertions(+), 49 dele
On 09/07/2013 05:48, Andrew Morton wrote:
> The feature should also be documented in the Linux manpages, please. A
> way to do that is to Cc linux-...@vger.kernel.org and Michael Kerrisk
> , and to work with Michael on developing that
> documentation.
OK
--
To unsubscribe from this list: send th
On 08/07/2013 22:37, Linus Torvalds wrote:
> On Mon, Jul 8, 2013 at 10:14 AM, Eliezer Tamir
> wrote:
>>
>> I think there is no way for the compiler to know the value of
>> can_busy_loop at compile time. It depends on the replies we get
>> from polling the sockets.
On 08/07/2013 19:37, Linus Torvalds wrote:
> On Mon, Jul 8, 2013 at 6:20 AM, Eliezer Tamir
> wrote:
>>
>> - /* only if on, have sockets with POLL_LL and not out of time
>> */
>> - if (ll_flag && can_ll && can_poll_ll(ll_
the existence of supported
sockets with valid queue information.
Signed-off-by: Eliezer Tamir
---
Documentation/sysctl/net.txt| 12 +---
fs/select.c | 60 ---
include/net/ll_poll.h | 46
On 08/07/2013 06:14, David Miller wrote:
From: Linus Torvalds
Date: Sun, 7 Jul 2013 15:33:31 -0700
End result: I think the code is salvageable and people who want this
kind of busy-looping can have it. But I really don't want to merge it
as-is. I think it was badly done, I think it was badly d
On 03/07/2013 10:48, Stephen Rothwell wrote:
Hi all,
After merging the final tree, today's linux-next build (powerpc
allnoconfig) failed like this:
fs/select.c: In function 'do_select':
fs/select.c:408:2: error: implicit declaration of function 'll_start_time'
[-Werror=implicit-function-decla
On 02/07/2013 23:42, Ben Hutchings wrote:
On Tue, 2013-07-02 at 23:28 +0300, Eliezer Tamir wrote:
On 02/07/2013 23:10, Ben Hutchings wrote:
On Tue, 2013-07-02 at 12:49 +0300, Eliezer Tamir wrote:
Time in range will fail safely if we move to a different cpu with an
extremely large clock skew
correct placeholder declarations to prevent build breakage when
!CONFIG_NET_LL_RX_POLL
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/include/net/ll_poll.h b/include/net/ll_poll.h
index b76f004..0d620ba
On 02/07/2013 23:10, Ben Hutchings wrote:
On Tue, 2013-07-02 at 12:49 +0300, Eliezer Tamir wrote:
Time in range will fail safely if we move to a different cpu with an
extremely large clock skew.
Add time_in_range64() and convert lls to use it.
Signed-off-by: Eliezer Tamir
---
v1->v2
fi
Time in range will fail safely if we move to a different cpu with an
extremely large clock skew.
Add time_in_range64() and convert lls to use it.
changelog:
v2
- fixed double call to sched_clock in can_poll_ll
- fixed checkpatchisms
Signed-off-by: Eliezer Tamir
---
fs/select.c
Time in range will fail safely if we move to a different cpu with an
extremely large clock skew.
Add time_in_range64() and convert lls to use it.
Signed-off-by: Eliezer Tamir
---
v1->v2
fixed double call to sched_clock() in can_poll_ll(), checkpatchisms
fs/select.c |
On 02/07/2013 11:38, Eliezer Tamir wrote:
On 02/07/2013 00:08, David Miller wrote:
From: Eliezer Tamir
Date: Fri, 28 Jun 2013 15:59:18 +0300
Here are two cleanup patches.
1. fix warning from debug_smp_processor_id().
- reported by Cody P Schafer.
Applied, but like Ben said perhaps you
On 02/07/2013 00:08, David Miller wrote:
From: Eliezer Tamir
Date: Fri, 28 Jun 2013 15:59:18 +0300
Here are two cleanup patches.
1. fix warning from debug_smp_processor_id().
- reported by Cody P Schafer.
Applied, but like Ben said perhaps you want to remember the last cpu you
got the
On 28/06/2013 19:51, Ben Hutchings wrote:
On Fri, 2013-06-28 at 15:59 +0300, Eliezer Tamir wrote:
Our use of sched_clock is OK because we don't mind the side effects
of calling it and occasionally waking up on a different CPU.
Sure about that? Jitter matters too.
Pretty sure, this
On 28/06/2013 17:38, Andi Kleen wrote:
diff --git a/fs/select.c b/fs/select.c
index 79b876e..3654075 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -402,7 +402,7 @@ int do_select(int n, fd_set_bits *fds, struct timespec
*end_time)
poll_table *wait;
int retval, i, timed_out = 0;
Change Low Latency Sockets code for select and poll so that
when LLS is disabled sched_clock() is never called.
Also, avoid sending POLL_LL to sockets if disabled.
Reported-by: Andi Kleen
Signed-off-by: Eliezer Tamir
---
fs/select.c | 11 +++
include/net/ll_poll.h | 17
David,
Here are two cleanup patches.
1. fix warning from debug_smp_processor_id().
- reported by Cody P Schafer.
2. avoid calling sched_clock() in select and poll when lls is disabled.
- reported by Andi Kleen.
Thanks to everyone for their comments.
-Eliezer
--
To unsubscribe from this list: s
P Schafer
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h | 30 +-
1 files changed, 25 insertions(+), 5 deletions(-)
diff --git a/include/net/ll_poll.h b/include/net/ll_poll.h
index 5bf2b3a..6d45e6f 100644
--- a/include/net/ll_poll.h
+++ b/include/net/ll_poll.h
On 28/06/2013 03:29, Cody P Schafer wrote:
On 06/27/2013 05:25 PM, Cody P Schafer wrote:
On 06/24/2013 12:28 AM, Eliezer Tamir wrote:
select/poll busy-poll support.
...
I'm seeing warnings about using smp_processor_id() while preemptable
(log included below) due to this patch. I expec
On 28/06/2013 07:43, Andi Kleen wrote:
@@ -400,6 +402,8 @@ int do_select(int n, fd_set_bits *fds, struct timespec
*end_time)
poll_table *wait;
int retval, i, timed_out = 0;
unsigned long slack = 0;
+ unsigned int ll_flag = POLL_LL;
+ u64 ll_time = ll_end_time(
On 25/06/2013 17:26, yaniv saar wrote:
On Wed, Jun 19, 2013 at 2:42 PM, Eliezer Tamir
wrote:
[this patch needs the poll patch to be applied first]
with sockperf doing epoll on 1000 sockets I see an avg latency of 6us
hi eliezer,
please consider the following solution for epoll that is
David,
Here is a rework of the select/poll patch.
When select or poll are used on a lot of sockets the sysctl value
needs to be set higher than 50. For 300 sockets a setting of 100 works
for me. For 1000 sockets a setting of 200 works well but the gain is
very small, probably not worth it.
I sho
-off-by: Eliezer Tamir
---
Documentation/sysctl/net.txt| 18 --
fs/select.c | 34 +-
include/net/ll_poll.h | 35 ++-
include/uapi/asm-generic/poll.h |2 ++
net/core/sock.c
On 24/06/2013 04:44, David Miller wrote:
From: Eliezer Tamir
Date: Wed, 19 Jun 2013 13:04:21 +0300
One question: do we need in sock_poll() to test that sock->sk is not null?
(Thanks to Willem de Bruijn for pointing this out.)
We should not have to.
Please clean up various things in t
micro-benchmark.
[this patch needs the poll patch to be applied first]
with sockperf doing epoll on 1000 sockets I see an avg latency of 6us
Signed-off-by: Eliezer Tamir
---
fs/eventpoll.c | 39 +--
1 files changed, 33 insertions(+), 6 deletions(-)
diff --
.
Signed-off-by: Eliezer Tamir
---
Documentation/sysctl/net.txt| 18 --
fs/select.c | 34 +-
include/net/ll_poll.h | 34 ++
include/uapi/asm-generic/poll.h |2 ++
net/core/sock.c
David,
Here is a rework of the select/poll patch.
One question: do we need in sock_poll() to test that sock->sk is not null?
(Thanks to Willem de Bruijn for pointing this out.)
When select or poll are used on a lot of sockets the sysctl value
needs to be set higher than 50. For 300 sockets a se
On 18/06/2013 17:45, Eliezer Tamir wrote:
On 18/06/2013 17:35, Eric Dumazet wrote:
On Tue, 2013-06-18 at 16:25 +0300, Eliezer Tamir wrote:
One other thing,
sock_poll() will only ll_poll if the flag was set _and_ the socket has a
none-zero value in sk->sk_ll_usec so you still only poll
On 18/06/2013 17:35, Eric Dumazet wrote:
On Tue, 2013-06-18 at 16:25 +0300, Eliezer Tamir wrote:
One other thing,
sock_poll() will only ll_poll if the flag was set _and_ the socket has a
none-zero value in sk->sk_ll_usec so you still only poll on sockets
that were enabled for LLS, not on ev
On 18/06/2013 13:25, Eric Dumazet wrote:
On Tue, 2013-06-18 at 11:58 +0300, Eliezer Tamir wrote:
@@ -731,7 +755,11 @@ static inline unsigned int do_pollfd(struct pollfd
*pollfd, poll_table *pwait)
mask = DEFAULT_POLLMASK;
if (f.file->f
On 18/06/2013 13:25, Eric Dumazet wrote:
On Tue, 2013-06-18 at 11:58 +0300, Eliezer Tamir wrote:
@@ -393,6 +394,15 @@ static inline void wait_key_set(poll_table *wait, unsigned
long in,
wait->_key |= POLLOUT_SET;
}
+static inline void wait_key_set_lls(poll_table *w
On 18/06/2013 12:08, Eric Dumazet wrote:
On Tue, 2013-06-18 at 11:58 +0300, Eliezer Tamir wrote:
select/poll busy-poll support.
*/
-static inline u64 ll_end_time(struct sock *sk)
+static inline u64 ll_sk_end_time(struct sock *sk)
{
- u64 end_time = ACCESS_ONCE(sk->sk_ll_u
low-level
sock_poll() again until we are out of time or we find something.
Once the system call finds something, it stops setting POLL_LL, so it can
return the result to the user ASAP.
Signed-off-by: Alexander Duyck
Signed-off-by: Jesse Brandeburg
Signed-off-by: Eliezer Tamir
---
fs/select.c
David,
Here is a rework of the select/poll patch.
(I called this a v2 but we are continuing where we left off in v9
of the original series.)
One question: do we need in sock_poll() to test that sock->sk is not null?
(Thanks to Willem de Bruijn for pointing this out.)
When select or poll are use
Add lls socket option support.
use --lls (value in usecs) to override global setting.
Right now we always get and print the value of the option before we set,
to show that the option works properly.
This should be removed in an official release.
---
src/Defs.h |3 +++
src/SockPerf.cpp
Remove NET_LL_RX_POLL from the config menu.
Change default to y.
Busy polling still needs to be enabled at run time.
Signed-off-by: Eliezer Tamir
---
net/Kconfig | 11 ++-
1 files changed, 2 insertions(+), 9 deletions(-)
diff --git a/net/Kconfig b/net/Kconfig
index e591668..51da839
adds a socket option for low latency polling.
This allows overriding the global sysctl value with a per-socket one.
Unexport sysctl_net_ll_poll since for now it's not needed in modules.
Signed-off-by: Eliezer Tamir
---
arch/alpha/include/uapi/asm/socket.h |2 ++
arch/avr32/include
David,
Here are three clean up patches, plus one that adds a socket option for LLS.
Patch 1 converts sysctl_net_ll_poll into an unsigned int
Patch 2 converts low latency sockets to sched_clock.
Patch 3 removes the config menu for NET_LL_RX_POLL and defaults to y.
Patch 4 adds an SO_LL socket opti
Use sched_clock() instead of get_cycles().
We can use sched_clock() because we don't care much about accuracy.
Remove the dependency on X86_TSC
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h | 33 +
net/Kconfig |1 -
2 files change
There is no reason for sysctl_net_ll_poll to be an unsigned long.
Change it into an unsigned int.
Fix the proc handler.
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h |5 +++--
net/core/sysctl_net_core.c |4 ++--
net/socket.c |2 +-
3 files changed, 6
On 14/06/2013 06:12, Eric Dumazet wrote:
On Fri, 2013-06-14 at 04:57 +0300, Eliezer Tamir wrote:
Use sched_clock() instead of get_cycles().
We can use sched_clock() because we don't care much about accuracy.
Remove the dependency on X86_TSC
Signed-off-by: Eliezer Tamir
---
-static i
Add lls socket option support.
use --lls (value in usecs) to override global setting.
Right now we always get and print the value of the option before we set,
to show that the option works properly.
This should be removed in an official release.
---
src/Defs.h |3 +++
src/SockPerf.cpp
Use sched_clock() instead of get_cycles().
We can use sched_clock() because we don't care much about accuracy.
Remove the dependency on X86_TSC
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h | 32 +---
net/Kconfig |1 -
2 files change
adds a socket option for low latency polling.
This allows overriding the global sysctl value with a per-socket one.
Unexport sysctl_net_ll_poll since for now it's not needed in modules.
Signed-off-by: Eliezer Tamir
---
arch/alpha/include/uapi/asm/socket.h |2 ++
arch/avr32/include
Remove NET_LL_RX_POLL from the config menu.
Change default to y.
Busy polling still needs to be enabled at run time.
Signed-off-by: Eliezer Tamir
---
net/Kconfig | 11 ++-
1 files changed, 2 insertions(+), 9 deletions(-)
diff --git a/net/Kconfig b/net/Kconfig
index e591668..51da839
There is no reason for sysctl_net_ll_poll to be an unsigned long.
Change it into an unsigned int.
Fix the proc handler.
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h |5 +++--
net/core/sysctl_net_core.c |4 ++--
net/socket.c |2 +-
3 files changed, 6
David,
Here are three clean up patches, plus one that adds a socket option for LLS.
Patch 1 converts sysctl_net_ll_poll into an unsigned int
Patch 2 converts low latency sockets to sched_clock.
Patch 3 removes the config menu for NET_LL_RX_POLL and defaults to y.
Patch 4 adds an SO_LL socket opti
On 13/06/2013 18:31, Eric Dumazet wrote:
On Thu, 2013-06-13 at 17:46 +0300, Eliezer Tamir wrote:
+extern unsigned int sysctl_net_ll_poll __read_mostly;
Oh well, you re-introduce the overflow
hint : you defined
#define USECS_TO_NS(us) (us << 10)
ouch!
(thanks)
--
To unsubscribe fro
Add lls socket option support.
use --lls (value in usecs) to override global setting.
Right now we always get and print the value of the option before we set,
to show that the option works properly.
This should be removed in an official release.
---
src/Defs.h |3 +++
src/SockPerf.cpp
On 13/06/2013 17:49, Eliezer Tamir wrote:
Add lls socket option support to sockperf.
Right now we always get the option before set to show the option is
working properly. We should probably remove that in an official release.
use --lls (value in usecs) to override global setting.
Sorry, this
Add lls socket option support to sockperf.
Right now we always get the option before set to show the option is
working properly. We should probably remove that in an official release.
use --lls (value in usecs) to override global setting.
---
src/Defs.h |3 +++
src/SockPerf.cpp | 5
There is no reason for sysctl_net_ll_poll to be an unsigned long.
Change it into an unsigned int.
Fix the proc handler.
Signed-off-by: Eliezer Tamir
---
include/net/ll_poll.h |2 +-
net/core/sysctl_net_core.c |4 ++--
net/socket.c |2 +-
3 files changed, 4
1 - 100 of 241 matches
Mail list logo
