On Saturday 12 May 2012 06:37 PM, Wolfram Sang wrote:
* PGP Signed by an unknown key
On Mon, May 07, 2012 at 12:16:19PM +0530, Laxman Dewangan wrote:
The notification of the transfer complete by calling complete()
should be done after clearing all interrupt status.
This may avoid the race condition of misconfigure the i2c controller
in multi-core environment.
"may avoid" or "does avoid"? I.e. does it fix any bug that really
occured? Then I'd add a stable tag.
So far, applied to next.
"does avoid".
We had some of stability issue in long hour stress test with i2c and
this is one the fix towards the stability issue. Most of fixes are
already there in the mainline.
It is fine to stable tag on this change.
if (status& I2C_INT_PACKET_XFER_COMPLETE) {
BUG_ON(i2c_dev->msg_buf_remaining);
complete(&i2c_dev->msg_complete);
}
Side note: If you see a way to handle the BUG invocations more
gracefully, that would be appreciated.
Ok, I will work towards that.
Thanks,
Laxman
--
To unsubscribe from this list: send the line "unsubscribe linux-i2c" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
