Re: [PATCH V1] i2c: tegra: increase transfer timeout

[Thierry Reding <thierry.reding@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2678 bytes --]

On Fri, Jan 18, 2019 at 05:21:28PM +0000, Sowjanya Komatineni wrote:
> 
> 
> >> increase transfer timeout to 10s to allow enough time during max 
> >> transfer size.
> >> 
> >> Signed-off-by: Sowjanya Komatineni <skomatineni@nvidia.com>
> >> ---
> >>  drivers/i2c/busses/i2c-tegra.c | 2 +-
> >>  1 file changed, 1 insertion(+), 1 deletion(-)
> >> 
> >> diff --git a/drivers/i2c/busses/i2c-tegra.c 
> >> b/drivers/i2c/busses/i2c-tegra.c index e417ebf7628c..ca7c581fb4c0 
> >> 100644
> >> --- a/drivers/i2c/busses/i2c-tegra.c
> >> +++ b/drivers/i2c/busses/i2c-tegra.c
> >> @@ -25,7 +25,7 @@
> >>  
> >>  #include <asm/unaligned.h>
> >>  
> >> -#define TEGRA_I2C_TIMEOUT (msecs_to_jiffies(1000))
> >> +#define TEGRA_I2C_TIMEOUT (msecs_to_jiffies(10000))
> >>  #define BYTES_PER_FIFO_WORD 4
> >>  
> >>  #define I2C_CNFG				0x000
> >
> >Should the timeout be set depending on the max transfer size? 10s seems an age if the max transfer size is 4KB. In other words, we should this only be applied for >T194+?
> >
> >Furthermore, in tegra_i2c_xfer_msg() we know the len of the message and so maybe it would be better to dynamically set the timeout depending on length?
> >
> >Cheers
> >Jon
> 
> Yes, that’s the ideal way to compute timeout based on msg len and bus rate. 
> To do this I had to update TEGRA_I2C_TIMEOUT macro to take arg and
> there are 3 different patches for tegra i2c under review and all of
> those will effect as the patch changes use TEGRA_I2C_TIMEOUT. 
> 
> So, Should I hold on to this change for now till those patches are merged?

If you have a number of patches with interdependencies, it's best to
send them out as a whole series. So you'd typically apply them in order
to a single branch, then use:

	$ git format-patch first^..last

where first is the SHA1 of the first commit you want to send, and last
is the the SHA1 of the last patch. The carret (^) means the parent
commit of the specified one and is needed because git format-patch
doesn't include the start of the sequence.

If the commits are at the top of your branch you can use something like
this:

	$ git format-patch -3

which will generate a series for the last three patches in the branch
(more specifically starting from HEAD).

If you send them as a series, it's immediately obvious in what order
they should be applied and generally makes it easier for people to
review and test.

I think in this case you can probably just have the other two patches
first in the series, then apply the timeout patch on top. That way you
can resolve the conflicts between patches 1 and 2, and patch 3 before
sending out.

Thierry

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]