On Mon, Feb 21, 2011 at 5:06 AM, Paul E. McKenney
<paulmck(a)linux.vnet.ibm.com> wrote:
> Thank you, Dave!
>
> Any thoughts on prioritization? On which tasks others might be able
> to take on?
I could look at
* ftrace
* kprobes
* VFP/NEON coredump support
* Thumb2 kernel benchmarking (not included in the original blueprints)
In the meantime, I'll continue pushing the OMAP Thumb2 kernel stuff,
but the real work is complete.
I've had a brief look at the coredump stuff, but I'm not committed yet.
I'm keen to hear people's views on which task sounds the most useful.
Cheers
---Dave
>
> Thanx, Paul
>
> On Fri, Feb 18, 2011 at 06:04:57PM +0000, Dave Martin wrote:
>> == [ACTIVITY] (Dave Martin) 2011-02-14 - 2011-02-18 ==
>>
>> === Activity ===
>> ==== hardware-n-kernel-standard-architecture ====
>> * Upstream link-time discard versus SMP_ON_UP problem still
>> not fixed. Russell suggested a patch; currently trying to
>> confirm whether it got merged anywhere.
>> * omap3/4 Thumb-2 compatiblity patches re-posted (again). By
>> reverting the low-level PM functionality to ARM for correct
>> interoperation with firmware, positive PM test results have
>> now been seen on omap3.
>> * Spent a bit of time trying to understand coredump generation
>> on ARM.
>>
>> ==== miscellaneous ====
>> * Cambridge validation lab support:
>> * Tried to rescue the new, broken vexpress board. No
>> success yet, but I now have a firmware upgrade which I
>> think should fix the problem.
>>
>> * Tried to set up u8500 boards, but failed due to a lack of
>> serial breakout boards.
>> ==== Plans ====
>> * Discuss prioritisation for outstanding kernel tasks.
>>
>> * Write up the output from the Freescale i.MX BSP review
>> discussion, and post for comment.
>>
>> ==== Absences ====
>> * Possible absense of a couple of days in early March
>> (unconfirmed).
>
Hello!
Just a reminder, we are going to be using IRC rather than phone for
today's linaro-kernel meetings.
Let's use #linaro-kernel on Freenode for today's linaro-kernel meetings.
Thanx, Paul
=== Device Tree ===
* Studying clkdev.c and trying to come up with the approach to
match clock node with struct clk pointer
* Reviewed and tested the basically mx51 dt patch set posted by
Jason
=== Misc ===
* Sent out mxs dma and mmc patch v3, and hopefully this version
could get merged
--
Regards,
Shawn
== U-Boot Cache Maintenance ==
* Working on v2 series - fixing review comments
== U-Boot SPL ==
* Sent an early drop to John Rigby.
* Added DMM initialization
* Added mux support and corrected mux values
* Testing with different versions of OMAP4 board
== Device Tree ==
* Modified s5pv310 device tree support as per comments from Grant for
the v2 patchset.
* Working on device tree based probe for uart driver, console init
from dt took up lot of time.
* Planning to submit v3 s5pv310 device tree patchset this week.
* sdhci driver will be modified next for device tree based probe.
== Storage performance eMMC ==
* Got trim to work on my mmci host driver
* Rebasing async mmc requests on top of 2.6.38-rc4 mmci (testing is not done).
* Started adding tests in mmc_test.c in kernel to investigate trim,
bg and mmc async requests.
* mmc_test results so far: Write after TRIM show 10 % max increase
for linear writes on my particular eMMC.
== Kernel work ==
* Reviewed TI pruss drivers
* Reviewed new unicore patches
== Flash drive performance ==
* Completed lwn.net article
* Many discussions based on this work
== Noteworthy ==
* Article now online: http://lwn.net/Articles/428584/
* Got invited to do the keynote at ELC
== RCU priority boosting ==
* Between fixing some Power bugs and finding an 8-CPU system that avoids others, testing of RCU priority boosting has restarted.
== Processes ==
* Completed a couple of TSC license requests.
* Started one more TSC license request.
== Misc ==
* Continuing chasing additional Power bugs to enable tests on larger systems.
== [ACTIVITY] (Dave Martin) 2011-02-14 - 2011-02-18 ==
=== Activity ===
==== hardware-n-kernel-standard-architecture ====
* Upstream link-time discard versus SMP_ON_UP problem still
not fixed. Russell suggested a patch; currently trying to
confirm whether it got merged anywhere.
* omap3/4 Thumb-2 compatiblity patches re-posted (again). By
reverting the low-level PM functionality to ARM for correct
interoperation with firmware, positive PM test results have
now been seen on omap3.
* Spent a bit of time trying to understand coredump generation
on ARM.
==== miscellaneous ====
* Cambridge validation lab support:
* Tried to rescue the new, broken vexpress board. No
success yet, but I now have a firmware upgrade which I
think should fix the problem.
* Tried to set up u8500 boards, but failed due to a lack of
serial breakout boards.
==== Plans ====
* Discuss prioritisation for outstanding kernel tasks.
* Write up the output from the Freescale i.MX BSP review
discussion, and post for comment.
==== Absences ====
* Possible absense of a couple of days in early March
(unconfirmed).
