Thread (14 messages) 14 messages, 3 authors, 2014-09-12

Re: [PATCH v2 0/2] Add DRA7xx CPSW Ethernet support in Device Tree

From: Mugunthan V N <hidden>
Date: 2014-09-11 06:32:06
Also in: linux-omap, lkml

On Wednesday 10 September 2014 07:31 PM, Nishanth Menon wrote:
On 09/10/2014 08:37 AM, Mugunthan V N wrote:
quoted
Adding device tree entry for CPSW to make it work in Dual EMAC mode.
These patches were tested with DRA7 hwmod patches on top of linux-next.
Patches are tested on top of Nishanth's PM tree for v3.17 [1] and pushed
my tree to [2].

Did a boot test with CPSW and ping test with suspend/resume, the boot logs
on DRA7xx EVM are posted at [3]

[1] git://github.com/nmenon/linux-2.6-playground.git testing/v3.17/cpu-idle-suspend-dra7-omap5-framework
[2] git://git.ti.com/~mugunthanvnm/ti-linux-kernel/linux.git v3.17/dra7-evm-cpsw
[3] http://pastebin.ubuntu.com/8309289/

Changes from initial version:
* Dropped patch for pinoff states
* Changed pinoff state to mode15

Mugunthan V N (2):
  ARM: dts: dra7: Add CPSW and MDIO module nodes for dra7
  ARM: dts: dra7-evm: Enable CPSW and MDIO for dra7xx EVM

 arch/arm/boot/dts/dra7-evm.dts | 107 +++++++++++++++++++++++++++++++++++++++++
 arch/arm/boot/dts/dra7.dtsi    |  59 +++++++++++++++++++++++
 2 files changed, 166 insertions(+)
I did apply these on Tony's
"omap-for-v3.18/dt 6e55426 Merge branch .."
they do apply.

However: why are we not sending dra72-evm.dts changes in the same
series since we claim it to be dra7xx support?
Will add dra72x support and submit the next version.
Also, in your tests (http://pastebin.ubuntu.com/8309289/)
echo -n "20">/sys/kernel/debug/pm_debug/wakeup_timer_seconds
ping 172.24.188.1  & (or even better do a transfer of large file in
background)
echo mem > /sys/power/state

transfer/ping should resume seamlessly. (only thing you should be
careful is that the server should not timeout for 20 second inactivity)..
ping will resume seamlessly will try tftp and update on my next version
submission. Will add ping test across suspend resume with both ports
while posting my next version

Regards
Mugunthan V N
It is a valid test to verify peripheral activity after resume, BUT,
equally important is to verify suspend-resume in the middle of activity.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help