- 23 Aug, 2018 1 commit
-
-
Federico Vaga authored
The bug is that on a second TDC instance, the driver will try to register again the trigger type. The trigger type needs to be registered only once at module_init Signed-off-by:Federico Vaga <federico.vaga@cern.ch>
-
- 20 Aug, 2018 10 commits
-
-
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
The driver exports a sysfs attrubte to change the coalescing timeout. A new library funciton can be used to set/get this value; and the test program allows the configuration before an acquisition. Keep in mind that, for FIFOs the configuration is global (users are warned on dmesg) while for DMA it is per-channel. I though it was better to have a common interface for both. Users are warnd in the documentation as well. In the long run, perhaps also the FIFO can have its own register. Signed-off-by: -
Federico Vaga authored
It is not an IRQ coalescing size, but the size that triggers the dump from the FIFO to the DDR. so, it is internal stuff that does not need to get exposed to the user Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
This way the name is properly terminated. Perhaps there is a bug in ZIO that expect a \0 and it does not check any upper limit. Signed-off-by:
-
- 16 Aug, 2018 4 commits
-
-
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by:
-
- 15 Aug, 2018 25 commits
-
-
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
They should be implemented in ZIO Signed-off-by: -
Federico Vaga authored
the HW_BUSY flag must be cleared before doing data_done in order to hold only during DMA. Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
it happened that ZIO was taking some extra characters after the name. to play safe I put a terminator there Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
This also fix a bug. The bug was that I was coping data at the wrong offset withing the ZIO block. It starts always at offset 0, but I was shifing by "start" bytes Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
For the time being is used only to change the BURST size using POST-SAMPLEs This solves also a bug with the ZIO trigger "user" which does arm for us, but this is not exactly what we want because we need to arm only when we know the number of samples to acquire Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Signed-off-by: -
Federico Vaga authored
Not reallyused at this point Signed-off-by: -
Federico Vaga authored
Signed-off-by:
-
