Three of these are fixes for ast2600 models that I found when testing
master. The forth is a usability improvement that is helpful when
diagnosing why a watchdog is biting.

v3 adds some comments and fixes whitespace, and r-b from Alex. Thanks
for the review Alex.

v2 fixes some review comments from Cédric and adds his r-b.

Joel Stanley (4):
  aspeed/sdmc: Make ast2600 default 1G
  aspeed/scu: Fix W1C behavior
  watchdog/aspeed: Improve watchdog timeout message
  watchdog/aspeed: Fix AST2600 frequency behaviour

 hw/misc/aspeed_scu.c             | 15 +++++++++++----
 hw/misc/aspeed_sdmc.c            |  6 +++---
 hw/watchdog/wdt_aspeed.c         | 24 +++++++++++++++++++-----
 include/hw/watchdog/wdt_aspeed.h |  1 +
 4 files changed, 34 insertions(+), 12 deletions(-)

-- 
2.24.0

On Wed, 13 Nov 2019 at 22:13, Joel Stanley <joel@jms.id.au> wrote:
>
> Three of these are fixes for ast2600 models that I found when testing
> master. The forth is a usability improvement that is helpful when
> diagnosing why a watchdog is biting.
>
> v3 adds some comments and fixes whitespace, and r-b from Alex. Thanks
> for the review Alex.
>
> v2 fixes some review comments from Cédric and adds his r-b.
>
> Joel Stanley (4):
>   aspeed/sdmc: Make ast2600 default 1G
>   aspeed/scu: Fix W1C behavior
>   watchdog/aspeed: Improve watchdog timeout message
>   watchdog/aspeed: Fix AST2600 frequency behaviour

Were you wanting any of these (or any of the other aspeed
patches currently on the list) to go into 4.2? None of them
are specifically marked as for-4.2 bugfixes, and I don't
know enough about aspeed to be able to judge their importance.

thanks
-- PMM

On 14/11/2019 15:04, Peter Maydell wrote:
> On Wed, 13 Nov 2019 at 22:13, Joel Stanley <joel@jms.id.au> wrote:
>>
>> Three of these are fixes for ast2600 models that I found when testing
>> master. The forth is a usability improvement that is helpful when
>> diagnosing why a watchdog is biting.
>>
>> v3 adds some comments and fixes whitespace, and r-b from Alex. Thanks
>> for the review Alex.
>>
>> v2 fixes some review comments from Cédric and adds his r-b.
>>
>> Joel Stanley (4):
>>   aspeed/sdmc: Make ast2600 default 1G
>>   aspeed/scu: Fix W1C behavior
>>   watchdog/aspeed: Improve watchdog timeout message
>>   watchdog/aspeed: Fix AST2600 frequency behaviour
> 
> Were you wanting any of these (or any of the other aspeed
> patches currently on the list) to go into 4.2? None of them
> are specifically marked as for-4.2 bugfixes, and I don't
> know enough about aspeed to be able to judge their importance.

I think we are done for 4.2, although these are relatively 
simple fixes that could go in 4.2. Nothing critical. As you
wish.

We are preparing, and reviewing, the initial 5.0 patchset 
which will contain the I2C pool buffer model plus the fixes 
Joel and I sent.

Thanks,
C.

On Thu, 14 Nov 2019 at 14:17, Cédric Le Goater <clg@kaod.org> wrote:
> We are preparing, and reviewing, the initial 5.0 patchset
> which will contain the I2C pool buffer model plus the fixes
> Joel and I sent.

OK. I currently have in my 'maybe I should do something with this'
list these patchsets:
aspeed/i2c: Add support for pool and DMA transfer modes
arm/aspeed: Watchdog and SDRAM fixes
aspeed: AST2600 SMC fixes and tacoma-bmc machine
aspeed: rework inter model link properties

(I had a quick scan through and there's nothing in any of
them that looked odd to me, though I haven't done a detailed
review). Should I just assume that you'll send out a rollup
patchset with all of those once 5.0 is out ? Alternatively
I could apply them to the target-arm.for-5.0 branch I'm
currently keeping, which will become the first target-arm
pullreq once 5.0 reopens, if you prefer (and if you've
got on-list reviewed-by tags for them all).

thanks
-- PMM

On 19/11/2019 14:44, Peter Maydell wrote:
> On Thu, 14 Nov 2019 at 14:17, Cédric Le Goater <clg@kaod.org> wrote:
>> We are preparing, and reviewing, the initial 5.0 patchset
>> which will contain the I2C pool buffer model plus the fixes
>> Joel and I sent.
> 
> OK. I currently have in my 'maybe I should do something with this'
> list these patchsets:
> aspeed/i2c: Add support for pool and DMA transfer modes
> arm/aspeed: Watchdog and SDRAM fixes
> aspeed: AST2600 SMC fixes and tacoma-bmc machine
> aspeed: rework inter model link properties
> 
> (I had a quick scan through and there's nothing in any of
> them that looked odd to me, though I haven't done a detailed
> review). Should I just assume that you'll send out a rollup
> patchset with all of those once 5.0 is out ? Alternatively
> I could apply them to the target-arm.for-5.0 branch I'm
> currently keeping, which will become the first target-arm
> pullreq once 5.0 reopens, if you prefer (and if you've
> got on-list reviewed-by tags for them all).

I have a larger list of patches with reviewed-by : 

Cédric Le Goater - aspeed: Change the "nic" property definition
Cédric Le Goater - aspeed: Change the "scu" property definition
PanNengyuan - gpio: fix memory leak in aspeed_gpio_init()
Cédric Le Goater - aspeed: Add support for the tacoma-bmc board
Cédric Le Goater - aspeed: Remove AspeedBoardConfig array and use AspeedMachineClass
Cédric Le Goater - aspeed/smc: Add AST2600 timings registers
Cédric Le Goater - aspeed/smc: Do not map disabled segment on the AST2600
Cédric Le Goater - aspeed/smc: Restore default AHB window mapping at reset
Joel Stanley - watchdog/aspeed: Fix AST2600 frequency behaviour
Joel Stanley - watchdog/aspeed: Improve watchdog timeout message
Joel Stanley - aspeed/scu: Fix W1C behavior
Joel Stanley - aspeed/sdmc: Make ast2600 default 1G
Cédric Le Goater - aspeed/i2c: Add trace events
Cédric Le Goater - aspeed/i2c: Add support for DMA transfers
Cédric Le Goater - aspeed: Add a DRAM memory region at the SoC level
Cédric Le Goater - aspeed/i2c: Check SRAM enablement on AST2500
Cédric Le Goater - aspeed/i2c: Add support for pool buffer transfers


Tests run fine on travis.

I will send them this afternoon.

Thanks,
C.