All,
As of the end of the day yesterday, Dmitry was still working on closed-loop
handling for GFX CPR. He has spent just over a full day on that task since
the last update.
At this point, we have exhausted the first tranche of 80 engineering hours
and are 2-3 hours into the second. Project management time is at 7.5 hours
of the 8 in the first tranche.
Updates will continue in my absence next week, most likely directly from
Dmitry. I will be checking emails from time to time, so if there are any
questions or concerns, please let me know.
Thanks.
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
Since the meeting on Monday, Dmitry continues to work on the implementation
of closed-loop for GFX CPR. He spent .6 engineering days yesterday.
Just a note: I will be away on vacation from Friday of this week and for
all of next week. We will continue giving updates during that time, but
they may be a bit more sporadic.
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
Apologies for the lack of updates this week. We had a slow start to the
week but are back in full activity now.
Dmitry has spent 1 additional engineer-day working on the GFX CPR,
programming voltages, and coding the boundaries. This work is ongoing.
We would request a quick engineering sync meeting early next week if
that works for you. We would suggest Monday early in your day. By that time
the work on the GPX CPR should be at a point where we can have more
detailed discussion about where we are, how Sightline might take a new
image for testing, and what remains. I don't know who other than Kim might
want to attend, but I think at least she should be part of that
conversation, as will Nico and Dmitry from the Linaro side. Kim, can you
suggest a date and time that work well for you?
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
As Dmitry is returning from vacation next week, it seemed a good time to
give everyone a quick status update to remind ourselves of where we are.
The PO for the additional tranche of hours has been received and an invoice
set to Sightline. Work on the project will continue as we await payment.
Before he left on vacation, Dmitry had the GFX CPR work mostly completed.
The voltage boundaries are all set correctly, and he only needs to finish
programming the target quotients. He will start on that when he returns.
This should finish off the GPU power regulator work, at which point he will
return to completing the CPU power regulator tasks.
For project spend, we have the following (all figures compared against the
total budget of the two PO's and are current through 15 August):
Engineering Time: 71.65 of 160 hours used (8.95 of 20
engineer days). This is 44.9% of the revised budget.
Project Management Time: 6 of 16 hours used (0.75 of 2 PM days). This is
37.5% of the revised budget.
As Dmitry gets working on the project again, daily status updates will
resume.
As always, if you have any questions or concerns, please let me know.
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
Over the course of Wednesday and Thursday, Dmitry spent 8 hours working on
the GFX CPR implementation. He estimates the work to now be approximately
50% complete on that item. He is continuing work on that today.
A full project burn report will come on Monday.
As always, if you have any questions, please let me know.
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
Dmitry has spent time working on the GFX power regulator which he hopes to
finish by the end of the week. He spent approximately 6.5 hours on the work
yesterday after having spent no appreciable time on it on Monday.
Dmitry is scheduled for vacation for the next two weeks, so progress will
be suspended during that time. He will return to work on Monday, 22 August.
Linaro is in receipt of the signed SoW for the project extension, but we do
not yet have the Purchase Order.
More updates during the week.
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
As was noted in the last update, Dmitry worked Friday and over the weekend
to catch up with work not completed last week.
Over Friday and the weekend, he spent 10.5 hours and accomplished:
- Complete a fix for the OPP tables for CPU frequencies
- Worked on implementing the CPUs LDO and APM voltage regulators.
Budget burn:
- As of the start of the day today, we have consumed 54.75 of the 80
engineering hours available, leaving 25.25 hours remaining. I have consumed
a total of 4.50 of the 8 project management hours.
At this point, I would recommend we have a conversation about prioritising
any remaining work, including support for Sightline's efforts to build and
test the kernels Linaro have provided. Along with that, I would suggest a
quick review of what (if anything) would be left that Sightline would want
Linaro to accomplish and whether or not a contract extension is warranted
to accomplish that work. Would either Tuesday or Wednesday work for you? I
would expect that conversation to take no longer than 30 minutes, and I
would be happy to make myself available at any time you wish.
Thank you.
Kind regards,
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
As we discussed in yesterday’s status meeting, There have been some
unexpected complexities in delivering all of the components of the 5.15
kernel as planned. Let me give a quick status update on where we are and
what we recommend continuing with.
Still in progress and yet to be delivered pending testing: cpufreq and
cpuidle changes. With the fixes he has, Dmitry is able to boot and run the
board with 4 cores and without the RCU messages previously seen.
With the remaining hours left in the original contract, Dmitry can work on
the CPR and voltage regulator support issues which include additional GPU
support, but those fixes will take more effort than currently remains in
the budget. Dmitry currently estimates that full regulator support will
require 5-10 more working days of effort. At present, there are roughly 3
days of effort left in the original budget, and at least one of those days
will likely be consumed with cpufreq testing, branch creation, and support
for SightLine testing.
Dmitry also believes that the fixes to CPR and the regulators will correct
the spurious rebooting issue SightLine is seeing in one of the test images
provided to date.
Linaro recommends that the agreement be extended out for an additional 10
days of budget in order to complete the CPR and regulator work. This should
allow us to deliver the remaining work to provide SightLine with a 5.15
kernel fit for purpose on your production platforms.
If you agree with this approach, and I believe you have agreed in principle
pending details, then I will forward an extension of the first contract for
your review and signature.
As always, please do not hesitate to contact me with any questions or
concerns.
Thank you.
Kind regards,
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
Progress since Tuesday:
- Kim and Dmitry continue to work through the build and image
difficulties they have been having. The latest effort is to have Sightline
disable cpufreq for early testing and until Dmitry can finish the rest of
the work.
- In the meantime, Dimitry has done some testing of the images for
undervolting and has been working on understanding and modifying the CPU
core regulator code.
Work in Progress:
- In addition to the CPU core regulator work and the cpufreq issues,
Dmitry continues to focus on gfx freq, CPR, and CBF work.
Budget burn:
- Across Wednesday and Thursday, Dmitry was only able to spend about
3.25 hours on the project due to personal time off work. This leaves 35.75
hours of engineering time in the budget (44.7%). He will be working over
this weekend to catch back up with pending work. To date, I have now
consumed 4 hours (50%) of the project's PM budget.
As always, if you have questions or concerns, please reach out to me.
Kind regards,
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
All,
Here is where we stand on the project at the end of the day on Thursday:
Progress since yesterday:
- As promised, Dmitry turned his attention to preparing a preliminary
5.15 kernel with his clock driver patches for Sightline testing. That
kernel has been published on the following public branch and is ready for
Sightline's initial testing:
https://git.linaro.org/landing-teams/working/qualcomm/kernel.git/log/?h=rel…
Dmitry notes that this kernel should be passed the kernel parameter
'maxcpus=2'.
Work in Progress:
- With the initial kernel published, work will return to the CBF/CPR
patches, which when complete will be published on a separate branch.
Planned Work:
- Finish the CBF special handling in the CPR driver
- Utilize CBF for CPUfreq scaling
- Consider utilizing CPR driver for the GPU regulator powerup
- Finalize extra corner cases in the CPR driver
- Support for Sightline as required.
Budget burn: Yesterday, Dmitry spent 1.5 hours on the project, equating to
0.2 person-days. This leaves 5.6 person-days (45.0 hours) remaining in the
engineering budget. I spent .25 hours (for a running total of 3 hours
including the kickoff meeting and project set-up), meaning 37% of the PM
time has been consumed.
We feel we are still on-track to to deliver the expected outcomes within
current budget. We are especially pleased that we have been able to deliver
an initial, testable 5.15 kernel with over half of the engineering budget
remaining, and we hope Sightline will be pleased with the results.
As always, if you have any questions or concerns please reach out at any
time.
Kind regards,
Paul
--
[image: Linaro] <http://www.linaro.org/>
Paul Neuhardt | *Sr. Program Manager*
T: +44 0771 377 8664
paul.neuhardt(a)linaro.org <glen.valante(a)linaro.org>
IRC: pneuhardt
