rockowitz

The large sleep and the syslog message reflect the same problem. If the index into the steps array is beyond the end, the sleep multiplier used is whatever happens to...

By any chance are you using an explicit ***--sleep-multiplier*** value greater than 2.0 for this monitor?

@digitaltrails Please ignore the last comment I just deleted. I'm having a hard time coming with a reason why the monitor keeps returning the DDC Null Message when all that...

Your conjecture that the Nvidia card/Benq monitor combination is being driven too hard makes sense. The question is which is the culprit - that would take sniffing the line. The...

@digitaltrails Based on your tests, I have modified **multi_part_read_with_retry()** to sleep for an optional amount of time after **try_multi_part_read()** indicates that all its calls to **ddc_write_read_with_retry()** received a NULL message...

@digitaltrails I have installed a Nvidia 660ti using driver 535.54 on a test system, and put my most ornery monitor, a Dell P2411h, on it. Lots of DDCRC_DDC_DATA errors, but...

Re the first log, tryctr is 1 because on the first iteration of the loop the call to try_multi_part_read() returned NULL, i.e. it succeeded. ***--i2*** only applies when multi_part_read() returns...

The latest set of changes implement an earlier mechanism for recovering from Null Message errors in ddc_write_read_with_retry(), which is (ultimately) called in multi_part_read_with_retry()'s loop. Utility option ***--f7*** enables it. If...

Passing --f7 on the command line works for me: ~~~ rock@banner:/shared/playproj/i2c$ ./test_dsa_cap --f7 rm: cannot remove '/home/rock/.cache/ddcutil/dsa': No such file or directory =====> Starting loop 1 cat: /home/rock/.cache/ddcutil/dsa: No such...

Try adding ***--trcfunc ddc_initial_checks_by_dh***. DREF_DDC_USES_DDC_FLAG_FOR_UNSUPPORTED should be set, not some other DREF_DDC_...FOR_UNSUPPORTED flag.