On Tue, Sep 08, 2020 at 10:23:33PM -0500, Jassi Brar wrote:
> On Tue, Sep 8, 2020 at 4:15 AM Arnd Bergmann wrote:
> >
> > Picking up the old thread again after and getting pinged by multiple
> > colleagues about it (thanks!) reading through the history.
> >
> > On Fri, Jun 12, 2020 at 7:29 AM
On 08-09-20, 22:23, Jassi Brar wrote:
> From the test case Sudeep last shared, the scmi usage on mhu doesn't
> not even hit any bottleneck ... the test "failed" because of the too
> small hardcoded timeout value. Otherwise the current code actually
> shows better numbers.
Its not important on why
On Tue, Sep 8, 2020 at 4:15 AM Arnd Bergmann wrote:
>
> Picking up the old thread again after and getting pinged by multiple
> colleagues about it (thanks!) reading through the history.
>
> On Fri, Jun 12, 2020 at 7:29 AM Viresh Kumar wrote:
> >
> > On 11-06-20, 19:34, Jassi Brar wrote:
> > > In
On Tue, Sep 08, 2020 at 11:14:50AM +0200, Arnd Bergmann wrote:
> Picking up the old thread again after and getting pinged by multiple
> colleagues about it (thanks!) reading through the history.
>
> On Fri, Jun 12, 2020 at 7:29 AM Viresh Kumar wrote:
> >
> > On 11-06-20, 19:34, Jassi Brar wrote:
On 08-09-20, 11:14, Arnd Bergmann wrote:
> Picking up the old thread again after and getting pinged by multiple
> colleagues about it (thanks!) reading through the history.
Thanks for your inputs Arnd.
> Earlier, Jassi also commented "Linux does not provide real-time
> guarantees", which to me
Picking up the old thread again after and getting pinged by multiple
colleagues about it (thanks!) reading through the history.
On Fri, Jun 12, 2020 at 7:29 AM Viresh Kumar wrote:
>
> On 11-06-20, 19:34, Jassi Brar wrote:
> > In the first post in this thread, Viresh lamented that mailbox
> >
On 11-06-20, 19:34, Jassi Brar wrote:
> In the first post in this thread, Viresh lamented that mailbox
> introduces "a few ms" delay in the scheduler path.
> Your own tests show that is certainly not the case -- average is the
> same as proposed virtual channels 50-100us, the best case is 3us vs
>
On Thu, Jun 11, 2020 at 5:00 AM Sudeep Holla wrote:
> >
> > > Interesting logs ! The time taken to complete _successful_ requests
> > > are arguably better in bad_trace ... there are many <10usec responses
> > > in bad_trace, while the fastest response in good_trace is 53usec.
> >
> > Indeed
Hi Viresh,
Thanks for summarising the thoughts quite nicely.
On Wed, Jun 10, 2020 at 03:03:34PM +0530, Viresh Kumar wrote:
> On 05-06-20, 10:42, Jassi Brar wrote:
> > Since origin upto scmi_xfer, there can be many forms of sleep like
> > schedule/mutexlock etc think of some userspace
On 05-06-20, 10:42, Jassi Brar wrote:
> Since origin upto scmi_xfer, there can be many forms of sleep like
> schedule/mutexlock etc think of some userspace triggering sensor
> or dvfs operation. Linux does not provide real-time guarantees. Even
> if remote (scmi) firmware guarantee RT
On Fri, Jun 5, 2020 at 3:58 AM Sudeep Holla wrote:
>
> > > > >> bash-1526 [000] 1149.472553: scmi_xfer_begin:
> > > > >> transfer_id=1538 msg_id=6 protocol_id=21 seq=0 poll=0
> > > > >> -0 [001] 1149.472733: scmi_xfer_begin:
> > > > >> transfer_id=1539 msg_id=7
On Fri, Jun 05, 2020 at 01:30:54AM -0500, Jassi Brar wrote:
> On Thu, Jun 4, 2020 at 11:56 PM Sudeep Holla wrote:
>
> >
> > > >> bash-1526 [000] 1149.472553: scmi_xfer_begin:
> > > >> transfer_id=1538 msg_id=6 protocol_id=21 seq=0 poll=0
> > > >> -0 [001] 1149.472733:
On Thu, Jun 4, 2020 at 11:56 PM Sudeep Holla wrote:
>
> > >> bash-1526 [000] 1149.472553: scmi_xfer_begin:
> > >> transfer_id=1538 msg_id=6 protocol_id=21 seq=0 poll=0
> > >> -0 [001] 1149.472733: scmi_xfer_begin:
> > >> transfer_id=1539 msg_id=7 protocol_id=19 seq=1
On Thu, Jun 04, 2020 at 10:15:55AM -0500, Jassi Brar wrote:
> On Thu, Jun 4, 2020 at 4:20 AM Sudeep Holla wrote:
> >
> > On Wed, Jun 03, 2020 at 01:32:42PM -0500, Jassi Brar wrote:
> > > On Wed, Jun 3, 2020 at 1:04 PM Sudeep Holla wrote:
> > > >
> > > > On Fri, May 29, 2020 at 09:37:58AM +0530,
On Thu, Jun 4, 2020 at 4:20 AM Sudeep Holla wrote:
>
> On Wed, Jun 03, 2020 at 01:32:42PM -0500, Jassi Brar wrote:
> > On Wed, Jun 3, 2020 at 1:04 PM Sudeep Holla wrote:
> > >
> > > On Fri, May 29, 2020 at 09:37:58AM +0530, Viresh Kumar wrote:
> > > > On 28-05-20, 13:20, Rob Herring wrote:
> > >
On Wed, Jun 03, 2020 at 01:32:42PM -0500, Jassi Brar wrote:
> On Wed, Jun 3, 2020 at 1:04 PM Sudeep Holla wrote:
> >
> > On Fri, May 29, 2020 at 09:37:58AM +0530, Viresh Kumar wrote:
> > > On 28-05-20, 13:20, Rob Herring wrote:
> > > > Whether Linux
> > > > requires serializing mailbox accesses
On Thu, Jun 04, 2020 at 11:29:03AM +0530, Viresh Kumar wrote:
> On 03-06-20, 19:17, Sudeep Holla wrote:
> > I just realised that we have the timing info in the traces and you will
> > observe the sensor readings take something in order of 100us to 500-600us
> > or even more based on which sensor
On 03-06-20, 19:17, Sudeep Holla wrote:
> I just realised that we have the timing info in the traces and you will
> observe the sensor readings take something in order of 100us to 500-600us
> or even more based on which sensor is being read. While we have 100us
> timeout for cpufreq opp set.
On Wed, Jun 3, 2020 at 1:31 PM Sudeep Holla wrote:
>
> > >
> > H/W is actually fine :) Its just that the driver is written to
> > _also_ support a platform (my original) that doesn't have shmem and
> > need to pass data via 32bit registers.
> > Frankly, I am not against the doorbell mode, I am
On Wed, Jun 3, 2020 at 1:04 PM Sudeep Holla wrote:
>
> On Fri, May 29, 2020 at 09:37:58AM +0530, Viresh Kumar wrote:
> > On 28-05-20, 13:20, Rob Herring wrote:
> > > Whether Linux
> > > requires serializing mailbox accesses is a separate issue. On that side,
> > > it seems silly to not allow
On Mon, May 18, 2020 at 11:05:03PM -0500, Jassi Brar wrote:
> On Mon, May 18, 2020 at 10:40 PM Viresh Kumar wrote:
> >
> > On 18-05-20, 18:29, Bjorn Andersson wrote:
> > > On Thu 14 May 22:17 PDT 2020, Viresh Kumar wrote:
> > > > This stuff has been doing rounds on the mailing list since several
Hi Bjorn,
Thanks for the details response.
On Mon, May 18, 2020 at 06:29:27PM -0700, Bjorn Andersson wrote:
> On Thu 14 May 22:17 PDT 2020, Viresh Kumar wrote:
>
[...]
I find this part nicely summarise your response.
> > - With serialization, if we use only one channel as today at every
> >
On Wed, Jun 03, 2020 at 07:04:35PM +0100, Sudeep Holla wrote:
> On Fri, May 29, 2020 at 09:37:58AM +0530, Viresh Kumar wrote:
> > On 28-05-20, 13:20, Rob Herring wrote:
> > > Whether Linux
> > > requires serializing mailbox accesses is a separate issue. On that side,
> > > it seems silly to not
On Fri, May 29, 2020 at 09:37:58AM +0530, Viresh Kumar wrote:
> On 28-05-20, 13:20, Rob Herring wrote:
> > Whether Linux
> > requires serializing mailbox accesses is a separate issue. On that side,
> > it seems silly to not allow driving the h/w in the most efficient way
> > possible.
>
> That's
On 29-05-20, 00:20, Jassi Brar wrote:
> The fact that all these bits are backed by one physical signal makes
I believe that within the IP itself, there must be 32 signals, just
that only one comes out of it all OR'd together. Maybe that can be
verified by writing 0x01 to it multiple times and it
On Thu, May 28, 2020 at 2:20 PM Rob Herring wrote:
>
> On Fri, May 15, 2020 at 10:47:38AM +0530, Viresh Kumar wrote:
> > From: Sudeep Holla
> >
> > Hi Rob, Arnd and Jassi,
> >
> > This stuff has been doing rounds on the mailing list since several years
> > now with no agreed conclusion by all
On 28-05-20, 13:20, Rob Herring wrote:
> Whether Linux
> requires serializing mailbox accesses is a separate issue. On that side,
> it seems silly to not allow driving the h/w in the most efficient way
> possible.
That's exactly what we are trying to say. The hardware allows us to
write all 32
On Fri, May 15, 2020 at 10:47:38AM +0530, Viresh Kumar wrote:
> From: Sudeep Holla
>
> Hi Rob, Arnd and Jassi,
>
> This stuff has been doing rounds on the mailing list since several years
> now with no agreed conclusion by all the parties. And here is another
> attempt to get some feedback from
On 18-05-20, 19:53, Jassi Brar wrote:
> That is a client/protocol property and has nothing to do with the
> controller dt node.
That's what I am concerned about, i.e. different ways of passing the
doorbell number via DT.
--
viresh
On Mon, May 18, 2020 at 10:40 PM Viresh Kumar wrote:
>
> On 18-05-20, 18:29, Bjorn Andersson wrote:
> > On Thu 14 May 22:17 PDT 2020, Viresh Kumar wrote:
> > > This stuff has been doing rounds on the mailing list since several years
> > > now with no agreed conclusion by all the parties. And here
On 18-05-20, 18:29, Bjorn Andersson wrote:
> On Thu 14 May 22:17 PDT 2020, Viresh Kumar wrote:
> > This stuff has been doing rounds on the mailing list since several years
> > now with no agreed conclusion by all the parties. And here is another
> > attempt to get some feedback from everyone
On Thu 14 May 22:17 PDT 2020, Viresh Kumar wrote:
> From: Sudeep Holla
>
> Hi Rob, Arnd and Jassi,
>
> This stuff has been doing rounds on the mailing list since several years
> now with no agreed conclusion by all the parties. And here is another
> attempt to get some feedback from everyone
On Mon, May 18, 2020 at 2:42 AM Viresh Kumar wrote:
>
> On 15-05-20, 11:46, Jassi Brar wrote:
> > As I asked you yesterday over the call, it may help if you could share
> > some numbers to back up the doomsday scenario.
>
> Yes, I have already asked Sudeep to get some numbers for this. He will
>
On 15-05-20, 11:46, Jassi Brar wrote:
> As I asked you yesterday over the call, it may help if you could share
> some numbers to back up the doomsday scenario.
Yes, I have already asked Sudeep to get some numbers for this. He will
get back to us.
> > - With the current approach it isn't possible
On Fri, May 15, 2020 at 12:17 AM Viresh Kumar wrote:
>
> - The hardware gives us the capability to write the register in
> parallel, i.e. we can write 0x800 and 0x400 together without any
> software locks, and so these 32 bits should be considered as separate
> channel even if only one
From: Sudeep Holla
Hi Rob, Arnd and Jassi,
This stuff has been doing rounds on the mailing list since several years
now with no agreed conclusion by all the parties. And here is another
attempt to get some feedback from everyone involved to close this once
and for ever. Your comments will very
36 matches
Mail list logo