Dear Jon,
In message 9e4733910903221959n452a067kd46b313b9b453ec3@mail.gmail.com you wrote:
You did not answer my question in which way this actually makes I2C faster? Where do we save time, and how much?
I increased the retry loop to 10,000. This definitely makes my system faster. On my bus the actual i2c delay is 40-55us. The original code always delayed 1,000us so for me it a gain of 940us on each i2c operation. This is visible during things like probe.
Hm... you don't really convince me. Being "visible" is nice, but is it a reason to change the code? I mean, if you said something like: "without that patch operation FOO takes 20 seconds, and with the patch it takes less than 3 seconds", then everybody can understand why this is a good thing to do.
And: the I2C probe operation itself is just a debug thing, nothing you do in normal operation, so it doesn't look that time critical to me.
I agree that the change itself looks harmless, and I'm willing to accept it, but at least make sure not to reduce the total timeout values - that would require re-testing on some boards.
Best regards,
Wolfgang Denk