Hi Stephen,
On 14.09.2012 05:17, Stephen Rothwell wrote:
Hi Tejun,
After merging the cgroup tree, today's linux-next build (powerpc
ppc64_defconfig) failed like this:
drivers/net/tun.c: In function 'tun_alloc_skb':
drivers/net/tun.c:589:2: error: implicit declaration of function
'sock_update_
From: Daniel Wagner
Hi,
Sorry for breaking Tejun's cgroup for-3.7 branch. Here are two patches
which fix this problem. Either of them will do.
cheers,
daniel
Cc: "David S. Miller"
Cc: "Michael S. Tsirkin"
Cc: Gao feng
Cc: Jamal Hadi Salim
Cc: Joe Perches
C
From: Daniel Wagner
The two users of sock_update_classid() are net/socket.c and
drivers/net/tun.c. socket.c includes cls_cgroup.h already. Update
tun.c to inlcude the header.
Signed-off-by: Daniel Wagner
Acked-by: Tejun Heo
Acked-by: Li Zefan
Cc: Gao feng
Cc: Jamal Hadi Salim
Cc: John
From: Daniel Wagner
commit 1f66c0a8833c3974ab6b35edcf4f8585b2f94592
Author: Daniel Wagner
Date: Wed Sep 12 16:12:01 2012 +0200
cgroup: net_cls: Move sock_update_classid() declaration to cls_cgroup.h
Claimed that there was only net/socket.c depending on
sock_update_class(). That is not
Hi Sedat,
On 14.09.2012 16:43, Sedat Dilek wrote:
On Fri, Sep 14, 2012 at 4:33 PM, Daniel Wagner wrote:
From: Daniel Wagner
commit 1f66c0a8833c3974ab6b35edcf4f8585b2f94592
Author: Daniel Wagner
Date: Wed Sep 12 16:12:01 2012 +0200
cgroup: net_cls: Move sock_update_classid
Hi Tejun,
On 09.11.2012 18:22, Tejun Heo wrote:
Hey, Daniel.
On Fri, Nov 09, 2012 at 12:09:38PM +0100, Daniel Wagner wrote:
On 08.11.2012 20:07, Tejun Heo wrote:> Subject: cgroup: add
cgroup_subsys->post_create()
Currently, there's no way for a controller to find out whether a
}
part before create() or add some protection between create() and
post_create() callback in net_prio. I have a patch but I see
I could drop it completely if post_create() is there.
cheers,
daniel
From 84fbbdf0dc5d3460389e39a00a3ee553ee55b563 Mon Sep 17 00:00:00 2001
From: Daniel Wagner
Hi Tejun,
On 17.11.2012 04:31, Tejun Heo wrote:
-static int write_classid(struct cgroup *cgrp, struct cftype *cft, u64 value)
+static int write_classid(struct cgroup *cgrp, struct cftype *cft,
+const char *buf)
{
+ struct cgroup_cls_state *cs = cgrp_cls_state(cgrp
Hi Tejun,
On 17.11.2012 04:31, Tejun Heo wrote:
@@ -112,15 +161,6 @@ struct cgroup_subsys net_cls_subsys = {
.subsys_id = net_cls_subsys_id,
.base_cftypes = ss_files,
.module = THIS_MODULE,
-
- /*
-* While net_cls cgroup has the rudimentary hi
ds, this
currently doesn't make any userland-visible difference, but will help
implementing proper hierarchy support.
Signed-off-by: Tejun Heo
Acked-by: Daniel Wagner
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord.
Hi Tejun,
On 17.11.2012 04:30, Tejun Heo wrote:
> Hello, guys.
>
> This patchset implements proper hierarchy support for netcls_cgroup.
> Pretty simliar to the netprio one[3]. Simpler as each cgroup has
> single config value instead of array of them.
>
> This patchset contains the following thr
Hi Tejun,
On 19.11.2012 18:08, Tejun Heo wrote:
> Hello, Daniel.
>
> On Mon, Nov 19, 2012 at 02:47:06PM +0100, Daniel Wagner wrote:
>> On 17.11.2012 04:31, Tejun Heo wrote:
>>> @@ -112,15 +161,6 @@ struct cgroup_subsys net_cls_subsys = {
>>> .s
2: Daniel pointed out that cftype->write_u64() accepts base prefix
(e.g. 0x10 for 16). Updated "%lld" to "%lli".
Signed-off-by: Tejun Heo
Acked-by: David S. Miller
Acked-by: Neil Horman
Cc: Daniel Wagner
Acked-by: Daniel Wagner
thanks,
daniel
--
To unsubscribe fro
Hi Tejun,
I have tested this series and it works nicely. I didn't mentioned yet to
review it completely. I will do it tomorrow.
cheers,
daniel
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at
On 19.11.2012 20:25, Daniel Wagner wrote:
Hi Tejun,
I have tested this series and it works nicely. I didn't mentioned yet to
review it completely. I will do it tomorrow.
s/mentioned/managed/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body
Not introducing the 'is_local' is a good thing.
Tested and Acked-by: Daniel Wagner
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordo
On 20.11.2012 09:30, Tejun Heo wrote:
sscanf() doesn't bite.
Signed-off-by: Tejun Heo
Acked-by: Neil Horman
Tested and Acked-by: Daniel Wagner
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
On 20.11.2012 09:30, Tejun Heo wrote:
The function is about to go through a rewrite. In preparation,
shorten the variable names so that we don't repeat "priomap" so often.
This patch is cosmetic.
Signed-off-by: Tejun Heo
Acked-by: Neil Horman
Tested and Acked-by: Daniel
Hi Tejun,
On 20.11.2012 09:30, Tejun Heo wrote:
> netprio kept track of the highest prioidx allocated and resized
> priomaps accordingly when necessary. This makes it necessary to keep
> track of prioidx allocation and may end up resizing on every new
> prioidx.
>
> Update extend_netdev_table()
ed above @cs allocation
to simplify error path.
* In cgrp_css_free(), @cs assignment is made initialization.
Signed-off-by: Tejun Heo
Acked-by: Neil Horman
Tested and Acked-by: Daniel Wagner
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body
On 20.11.2012 09:30, Tejun Heo wrote:
Introduce two helpers - netprio_prio() and netprio_set_prio() - which
hide the details of priomap access and expansion. This will help
implementing hierarchy support.
Signed-off-by: Tejun Heo
Acked-by: Neil Horman
Tested and Acked-by: Daniel Wagner
s
allowed and the first level cgroups below the root cgroup behave
differently - they inherit priorities from the root cgroup on creation
instead of starting with 0. This is unfortunate but not doing so is
much crazier.
Signed-off-by: Tejun Heo
Tested and Acked-by: Daniel Wagner
---
Document
On 20.11.2012 15:38, Tejun Heo wrote:
Hello, Daniel.
On Tue, Nov 20, 2012 at 09:46:22AM +0100, Daniel Wagner wrote:
struct netprio_map {
struct rcu_head rcu;
struct netprio_aux *aux;/* auxiliary config array */
u32 priomap_len;
u32 priomap[];
};
Is
Hi Tejun,
On 20.11.2012 15:40, Tejun Heo wrote:
Hello, Daniel.
On Tue, Nov 20, 2012 at 09:57:14AM +0100, Daniel Wagner wrote:
-static void cgrp_css_free(struct cgroup *cgrp)
+static int cgrp_css_online(struct cgroup *cgrp)
{
- struct cgroup_netprio_state *cs = cgrp_netprio_state(cgrp
On 20.11.2012 16:13, Tejun Heo wrote:
Hello,
On Tue, Nov 20, 2012 at 04:09:22PM +0100, Daniel Wagner wrote:
Thanks for the explanation. I was pondering if the new size in power
of two could be a bit too excessive and the allocation step could be
linear, e.g. stick at 4096. target_id will
Hi,
I experience a regression on my Macbook Air 2012 model with the 3.9-rc
kernels (v3.9-rc2-284-gaea8b5d) The symptoms are that there is just no
sound from the loudspeakers. The headphone output works fine.
I was able to pin point it down to one commit
1077a024 "ALSA: hda - Use generic pars
On 03/25/2013 07:46 PM, Daniel Wagner wrote:
I experience a regression on my Macbook Air 2012 model with the 3.9-rc
kernels (v3.9-rc2-284-gaea8b5d) The symptoms are that there is just no
sound from the loudspeakers. The headphone output works fine.
Oh, there is already a fix in the mainline
_cls: cgrp_cls_state() doesn't have any user left. Removed.
Signed-off-by: Tejun Heo
Cc: Li Zefan
Cc: Peter Zijlstra
Cc: Ingo Molnar
Cc: Johannes Weiner
Cc: Michal Hocko
Cc: Balbir Singh
Cc: Aristeu Rozanski
Cc: Matt Helsley
Cc: Daniel Wagner
Cc: Vivek Goyal
Cc: Jens Axboe
Cc: Steven R
Zefan
Cc: Ingo Molnar
Cc: Matt Helsley
Cc: Daniel Wagner
Cc: Steven Rostedt
Nice cleanup.
Acked-By: Daniel Wagner
cheers,
daniel
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
Mor
_init() fails") is converted too. Suggested
by Li Zefan.
Signed-off-by: Tejun Heo
Acked-by: Li Zefan
Acked-by: Michal Hocko
Cc: Peter Zijlstra
Cc: Ingo Molnar
Cc: Johannes Weiner
Cc: Balbir Singh
Cc: Aristeu Rozanski
Cc: Matt Helsley
Cc: Daniel Wagner
Cc: Vivek Goyal
Cc: Jens
Hi,
I just updated my laptop (MacBook Air 2012) from 3.11-rc3+ to 3.11-rc4+
and since then brcsmac crashes short after enabling it. Unfortunately, I
did also change some config flags in the area of cpufreq (enabling
p-state driver) so the configuration is not exactly the same. Before I
try to
On 08/13/2013 03:41 PM, Arend van Spriel wrote:
On 08/12/2013 10:11 PM, Daniel Wagner wrote:
On 08/12/2013 12:38 PM, Arend van Spriel wrote:
On 08/11/2013 06:21 PM, Daniel Wagner wrote:
I just updated my laptop (MacBook Air 2012) from 3.11-rc3+ to 3.11-rc4+
and since then brcsmac crashes
On 14.09.2012 10:24, Glauber Costa wrote:
On 09/13/2012 09:48 PM, Tejun Heo wrote:
Hello, Glauber.
On Thu, Sep 13, 2012 at 03:53:56PM +0400, Glauber Costa wrote:
Here is where the Kconfig option comes to play. If we do it in the
kernel, userspace doesn't have to do anything. I spoke with Lenna
Hi Li,
On 21.01.2013 07:08, Li Zefan wrote:
> I'm not a network developer, so correct me if I'm wrong.
>
> Since commit 7955490f732c2b8
> ("net: netprio_cgroup: rework update socket logic"), sock->sk->sk_cgrp_prioidx
> is set when the socket is created, and won't be updated unless the task is
> m
On 21.01.2013 10:01, Li Zefan wrote:
> On 2013/1/21 16:50, Daniel Wagner wrote:
>> Hi Li,
>>
>> On 21.01.2013 07:08, Li Zefan wrote:
>>> I'm not a network developer, so correct me if I'm wrong.
>>>
>>> Since commit 7955490f732c2b8
>>
From: Daniel Wagner
Commit 6a328d8c6f03501657ad580f6f98bf9a42583ff7 changed the update
logic for the socket but it does not update the SCM_RIGHTS update
as well. This patch is based on the net_prio fix commit
48a87cc26c13b68f6cce4e9d769fcb17a6b3e4b8
net: netprio: fd passed in SCM_RIGHTS
On Mon, Jan 21, 2013 at 09:18:38AM -0800, John Fastabend wrote:
> On 01/21/2013 01:57 AM, Li Zefan wrote:
> > On 2013/1/21 17:27, Daniel Wagner wrote:
> >> On 21.01.2013 10:01, Li Zefan wrote:
> >>> On 2013/1/21 16:50, Daniel Wagner wrote:
> >>>> Hi Li
his I have written a test program, which
seems to test the right thing. Please have a look and let me know if
it is correct: http://www.monom.org/misc/scm_rights.c
And here a dirty first version of the patch:
From 49a78d907eaf31c16673025e7e3b4844e419e416 Mon Sep 17 00:00:00 2001
From: Daniel W
Hi,
I updated the firmware of Macbook Air Mid 2012 model to
version 2.5. The changelog from Apple
http://support.apple.com/kb/DL1592
"This update fixes an issue where Turbo Boost does not
activate when using Boot Camp"
indicates that they changed stuff on the ACPI tables.
Now I have noticed
Hi Casey,
On 26.09.2012 18:34, Casey Schaufler wrote:
On 9/24/2012 5:21 AM, Daniel Wagner wrote:
From: Daniel Wagner
The userland git repository has been moved to a new place.
Signed-off-by: Daniel Wagner
Applied to git://git.gitorious.org/smack-next/kernel.git#for-1209
Thanks.
BTW, I
hi,
i tried to test the hp5300 usb module for my scanner but it wasn't buildt.
any hints?
thx,
daniel
ps: experimental drivers, scsi and usb are enabled.
--
@gpg: http://stud3.tuwien.ac.at/~e9925791/daniel_wagner.asc
C63A 06F0 3E2A A039 E830 83A0 C1DA 3479 803F 078F
-
To unsubscribe
Pieter, do applications like yours need the cycle counter only for a few
predetermined packets or for each and every packet?
We need it for every packet for two reasons:
1) it's the only way to determine how many packets were dropped when
packet drops are flagged in the callback
Your applicatio
From: Daniel Wagner
Hi Jeff,
I had a bunch of small patches in my tree, which I forgot to send.
So here they are...
cheers,
daniel
Daniel Wagner (3):
Synchronize all clients on start up
posix03, posix04: Use '-l' instead of '-i' as option argument name
posix03: Do no
From: Daniel Wagner
kill(0, SIGINT) sends to all in the process group the signal
including the parent shell.
Instead remember the PIDs of all children and just send the signal
to these processes.
Reported-by: Dave Chinner
Signed-off-by: Daniel Wagner
---
posix03.c | 17 +++--
1
From: Daniel Wagner
The child process start working as soon as they are forked. Sometimes
that leads to a workload pattern that no contention happens at all even
with a high number of processes. Since the main motivation of this
this is to benchmark the contention overhead let the clients wait
From: Daniel Wagner
All other test use '=l' instead of '-i' to as option name for defining how many
loops should be executed. Let's streamline posix03 and posix04.
Signed-off-by: Daniel Wagner
---
posix03.c | 6 +++---
posix04.c | 6 +++---
2 files changed, 6 i
From: Daniel Wagner
Hi,
Using complete_all() is not wrong per se but it suggest that there
might be more than one waiter. For -rt I am reviewing all
complete_all() users and would like to leave only the real ones in the
tree. The main problem for -rt about complete_all() is that it can be
uses
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The generic caching code from sunrpc is calling revisit() only once.
The usage pattern of the completi
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
nfs_file_direct_write() or nfs_file_direct_read() allocated a request
object via nfs_direct_req_alloc(),
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
While we are at it, we also change the init_completion() to
reinit_completion() which slightly lighter.
Hi Anna,
On 09/23/2016 03:48 PM, Anna Schumaker wrote:
Besides trying to analys all the code paths to the wait_for_completion()
call and convince myself that there is only one waiter, I also run
a few tests:
- some fio benchmarks
- pynfs
-cthon04
Thanks for the patches, and for the extens
Hi Anna,
On 09/26/2016 03:31 PM, Anna Schumaker wrote:
> No, there are a handful that we know don't pass. When I run on NFS v4.1, I
> get:
>
> Failures: generic/035 generic/071 generic/128 generic/294
> generic/306 generic/309 generic/310 generic/313 generic/314
> generic/315 generic/316 generi
On 09/27/2016 09:42 AM, Daniel Wagner wrote:
> Failures: generic/035 generic/087 generic/088 generic/089 generic/126
> generic/184 generic/285 generic/347 generic/350
> Failed 9 of 62 tests
>
> Hmm, not sure if this good or bad. Does it help?
FWIW, running the same tests on non
From: Daniel Wagner
Hi,
While reviewing all the complete_all() users, I realized there is
recouring pattern how the completion API is used to synchronize the
stages of the firmware loading. Since firmware_class.c contains a
fairly complete implemetation for synching the loading, it worthwhile
From: Daniel Wagner
Loading firmware is an operation many drivers implement in various ways
around the completion API. And most of them do it almost in the same
way. Let's reuse the firmware_stat API which is used also by the
firmware_class loader. Apart of streamlining the firmware lo
From: Daniel Wagner
Loading firmware is an operation many drivers implement in various ways
around the completion API. And most of them do it almost in the same
way. Let's reuse the firmware_stat API which is used also by the
firmware_class loader. Apart of streamlining the firmware lo
From: Daniel Wagner
Some of the test are supposed to fail but we get a few ouptput lines:
./fw_filesystem.sh: line 51: printf: write error: Invalid argument
./fw_filesystem.sh: line 56: printf: write error: No such device
./fw_filesystem.sh: line 62: echo: write error: No such file or directory
From: Daniel Wagner
When running the test script you will get:
kselftest/firmware/fw_userhelper.sh: line 69: echo: write error: Resource
temporarily unavailable
and stops right there. Because the script runs with the '-e' option
which will stop the script at any error.
We should
From: Daniel Wagner
Loading firmware is an operation many drivers implement in various ways
around the completion API. And most of them do it almost in the same
way. Let's reuse the firmware_stat API which is used also by the
firmware_class loader. Apart of streamlining the firmware lo
From: Daniel Wagner
Factor out the firmware loading synchronization code in order to allow
drivers to reuse it. This also documents more clearly what is
happening. This is especial useful the drivers which will be converted
afterwards. Not everyone has to come with yet another way to handle it
From: Daniel Wagner
Loading firmware is an operation many drivers implement in various ways
around the completion API. And most of them do it almost in the same
way. Let's reuse the firmware_stat API which is used also by the
firmware_class loader. Apart of streamlining the firmware lo
From: Daniel Wagner
Loading firmware is an operation many drivers implement in various ways
around the completion API. And most of them do it almost in the same
way. Let's reuse the firmware_stat API which is used also by the
firmware_class loader. Apart of streamlining the firmware lo
On 07/28/2016 01:22 PM, Bastien Nocera wrote:
On Thu, 2016-07-28 at 09:55 +0200, Daniel Wagner wrote:
From: Daniel Wagner
Loading firmware is an operation many drivers implement in various
ways
around the completion API. And most of them do it almost in the same
way. Let's reus
Looking at the API, I really don't like the mixing of namespaces.
Either it's fw_ or it's firmware_ but not a mix of both.
I agree, that was not really clever.
struct fw_loading { ... }
__fw_loading_*()
fw_loading_*()
Would that make more sense? firmware_loading_ as pre
On 07/28/2016 03:10 PM, Steven Rostedt wrote:
Note, Daniel did start work on upstreaming this previously:
http://lkml.kernel.org/r/1430388385-29558-1-git-send-email-daniel.wag...@bmw-carit.de
There are a few things that do need to be cleaned up before this can go
upstream. Please coordinate
It's somewhat odd to me that the structure is "firmware_stat" but most
of the functions are "fw_loading_*". That seems inconsistent for a
structure that is (currently) only used by these functions.
I agree, my proposal is odd.
I would personally do either:
a) "struct fw_load_status" and "fw_
On 07/28/2016 07:57 PM, Dmitry Torokhov wrote:
On Thu, Jul 28, 2016 at 09:55:07AM +0200, Daniel Wagner wrote:
+int __firmware_stat_wait(struct firmware_stat *fwst,
+ long timeout)
+{
+ int err;
+ err = swait_event_interruptible_timeout(fwst->
On 07/28/2016 09:01 PM, Bjorn Andersson wrote:
On Thu 28 Jul 11:33 PDT 2016, Dmitry Torokhov wrote:
On Thu, Jul 28, 2016 at 09:55:11AM +0200, Daniel Wagner wrote:
From: Daniel Wagner
[..]
Do not quite like it... I'd rather asynchronous request give out a
firmware status pointer
On 07/31/2016 09:23 AM, Dmitry Torokhov wrote:
On July 30, 2016 9:58:17 AM PDT, "Luis R. Rodriguez" wrote:
On Sat, Jul 30, 2016 at 02:42:41PM +0200, Arend van Spriel wrote:
On 29-07-16 08:13, Daniel Wagner wrote:
On 07/28/2016 09:01 PM, Bjorn Andersson wrote:
On Thu 28 Jul 11:3
From: Daniel Wagner
When we load the firmware directly we don't need to take the umh
lock. So move this part inside fw_load_from_user_helper which is only
available when CONFIG_FW_LOADER_USER_HELPER is set.
This avoids a dependency on firmware_loading_timeout() even
From: Daniel Wagner
We track the state of the loading with bit ops. Since the state machine
has only a couple of states and there are only a few simple state
transition we can model this simplify.
UNKNOWN -> LOADING -> DONE | ABORTED
Signed-off-by: Daniel Wagner
Cc: Ming L
From: Daniel Wagner
The firmware user helper code tracks the current state of the loading
process via unsigned long status and a complection in struct
firmware_buf. We only need this for the usermode helper as such we can
encapsulate all this data into its own data structure.
Signed-off-by
From: Daniel Wagner
Hi,
This is version depends on Luis' '[PATCH v4 0/5] firmware: add SmPL
grammar to avoid issues' series [1]. Because of that I was able to get
rid of loading_timeout and firmware_loading_timeout() in the
!FW_LOADER_USER_HELPER_FALLBACK case.
cheers,
dani
From: Daniel Wagner
fw_lock is to use to protect 'corner cases' inside firmware_class. It
is not exactly clear what those corner cases are nor what it exactly
protects. fw_status can be used without needing the fw_lock to protect
its state transition and wake ups.
fw_status is holds
On 09/07/2016 06:42 PM, Thomas Gleixner wrote:
On Wed, 7 Sep 2016, Binoy Jayan wrote:
This work is based on work by Daniel Wagner. A few tracepoints are added
Interesting. The patch you are referring to was written by Carsten
Emde. Can you please get these attributions right? They want to be
On 09/08/2016 03:39 AM, Luis R. Rodriguez wrote:
On Wed, Sep 07, 2016 at 10:45:06AM +0200, Daniel Wagner wrote:
From: Daniel Wagner
timeout = MAX_JIFFY_OFFSET;
}
- retval = wait_for_completion_interruptible_timeout(&buf->com
On 09/08/2016 10:05 AM, Daniel Wagner wrote:
This while loop was originally a goto loop:
1f2b79599ee8 ("firmware loader: always let firmware_buf own the pages
buffer")
I don't think the code doesn't do what it was indented to do. The reason
is that callin
Hi Ming,
On 09/08/2016 01:26 PM, Ming Lei wrote:
On Wed, Sep 7, 2016 at 4:45 PM, Daniel Wagner wrote:
From: Daniel Wagner
+static int fw_status_wait_timeout(struct fw_status *fw_st, long timeout)
+{
+ int ret;
+
+ ret = wait_for_completion_interruptible_timeout(&fw_st->com
On 09/07/2016 10:45 AM, Daniel Wagner wrote:
@@ -702,7 +760,7 @@ static ssize_t firmware_loading_store(struct device *dev,
dev_err(dev, "%s: unexpected value (%d)\n", __func__, loading);
/* fallthrough */
case -1:
- fw_load_abo
esume.
Okay, so let's finish this round of refactoring first. I prefer going in
smaller steps and see if there are any regressions with those changes.
Signed-off-by: Daniel Wagner
Cc: Ming Lei
Cc: Luis R. Rodriguez
Cc: Greg Kroah-Hartman
---
drivers/base/firmware_class.c | 52 ++
On 09/08/2016 03:45 AM, Luis R. Rodriguez wrote:
On Wed, Sep 07, 2016 at 10:45:07AM +0200, Daniel Wagner wrote:
From: Daniel Wagner
We track the state of the loading with bit ops. Since the state machine
has only a couple of states and there are only a few simple state
transition
And they
From: Daniel Wagner
The (re)initialzing of the completion object should be done before we
trigger the transfer. Doing this after triggering the hardware opens up
a race window. Without the timeout we would problaly even deadlock. Use
also reinit_completion because we initalize the whole data
From: Daniel Wagner
Hi,
Using complete_all() is not wrong per se but it suggest that there
might be more than one waiter. For -rt I am reviewing all
complete_all() users and would like to leave only the real ones in the
tree. The main problem for -rt about complete_all() is that it can be
uses
From: Daniel Wagner
There is only one waiter for the completion, therefore there is no need
to use complete_all(). Let's make that clear by using complete() instead
of complete_all().
The usage pattern of the completion is:
waiter context waker context
virtscs
From: Daniel Wagner
There is only one waiter for the completion, therefore there is no need
to use complete_all(). Let's make that clear by using complete() instead
of complete_all().
The usage pattern of the completion is:
waiter context waker context
sym_eh_ha
Hi Luis,
On 09/09/2016 02:12 PM, Daniel Wagner wrote:
The firmware user helper code tracks the current state of the loading
process via unsigned long status and a complection in struct
firmware_buf. We only need this for the usermode helper as such we can
encapsulate all this data into its own
Hi Luis,
On 08/24/2016 02:45 AM, mcg...@kernel.org wrote:
diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c
index 22d1760a4278..dca4f9cbf4db 100644
--- a/drivers/base/firmware_class.c
+++ b/drivers/base/firmware_class.c
@@ -1412,7 +1412,7 @@ static int cache_firmware(con
Hi Luis,
On 08/30/2016 09:34 PM, Luis R. Rodriguez wrote:
On Mon, Aug 29, 2016 at 04:18:33PM +0200, Daniel Wagner wrote:
On 08/29/2016 11:50 AM, Daniel Wagner wrote:
I forgot to answer your question. So we have the dependency to
loading_timeout/firmware_loading_timeout from the firmware
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
From: Daniel Wagner
Hi,
Using complete_all() is not wrong per se but it suggest that there
might be more than one reader. For -rt I am reviewing all
complete_all() users and would like to leave only the real ones in the
tree. The main problem for -rt about complete_all() is that it can be
uses
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
From: Daniel Wagner
There is only one waiter for the completion, therefore there
is no need to use complete_all(). Let's make that clear by
using complete() instead of complete_all().
The usage pattern of the completion is:
waiter context waker co
Hi Binoy,
On 08/24/2016 01:17 PM, Binoy Jayan wrote:
Histogram output:
cat /sys/kernel/debug/tracing/events/latency/latency_irqs/hist
cat /sys/kernel/debug/tracing/events/latency/latency_preempt/hist
cat /sys/kernel/debug/tracing/events/latency/latency_critical_timings/hist
cat /sys/kernel/debug
From: Daniel Wagner
fw_lock is to use to protect 'corner cases' inside firmware_class. It
is not exactly clear what those corner cases are nor what it exactly
protects. fw_status can be used without needing the fw_lock to protect
its state transition and wake ups.
fw_status is holds
From: Daniel Wagner
This version should address all comments from Luis. In the last patch
the fw_lock dependency is dropped. This only works if we garantee not
to race between the reader and the writer side in combination in going
to sleep and waking up. So here that should now be a good
From: Daniel Wagner
The firmware user helper code tracks the current state of the loading
process via unsigned long status and a complection in struct
firmware_buf. We only need this for the usermode helper as such we can
encapsulate all this data into its own data structure.
Signed-off-by
1 - 100 of 742 matches
Mail list logo
