Commits · 82f87f3414e4ba8b97c1d2842dccb1cfc95c0309 · Projects / FMC TDC 1ns 5cha

28 Aug, 2018 2 commits

kernel: use correct buffer in test_dma() · 82f87f34
Manohar Vanga authored Aug 28, 2018

82f87f34

kernel: fixed the F***ING DMA issues (I mean they were fascinating). · 4f7cfb30

Manohar Vanga authored Aug 28, 2018

- needs new bitstream (28 August onwards)
- fixed scatter list allocation (needs one more entry for the start item)
- fixed DMA direction attribute (should be ORed)
- fixed scatter list page calculation to point to different pages

4f7cfb30

24 Aug, 2018 2 commits
- drv: bugfix in test_dma_sg · 8451407c
  Federico Vaga authored Aug 24, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
Signed-off-by: Tomasz Wlostowski <tomasz.wlostowski@cern.ch>
```
  8451407c
- drv: don't do anything if test data is not enabled · 0a416f20
  Federico Vaga authored Aug 24, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  0a416f20
23 Aug, 2018 6 commits
- drv: fix DMA wait helper · 8f269c14
  Federico Vaga authored Aug 23, 2018
  
  8f269c14
- drv: gennum status register has debug information about errors · 4fd46dd3
  Federico Vaga authored Aug 23, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  4fd46dd3
- drv: do not compile on warning · dc969cf9
  Federico Vaga authored Aug 23, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  dc969cf9
- drv: export testing DMA function to user-space · b3f907e8
  Federico Vaga authored Aug 23, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  b3f907e8
- Merge branch 'devfede-tst' into proposed_master · 954a5ad6
  Federico Vaga authored Aug 23, 2018
```
Mixed improvements
```
  954a5ad6
- drv: bugfix, register trigger type only once · 2d45c267
  Federico Vaga authored Aug 23, 2018
```
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>
```
  2d45c267
22 Aug, 2018 5 commits
- drv: remove print · c21b972a
  Federico Vaga authored Aug 22, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  c21b972a
- drv: debug print to inspect timestamps · ff36c141
  Federico Vaga authored Aug 22, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  ff36c141
- drv: reduce DMA transfer to maximum PAGE_SIZE · 3edd1f2b
  Federico Vaga authored Aug 22, 2018
```
Apparently the gn4124 does not handle properly bigger pages. I think I have
seen this before in the ADC and indeed thare we do PAGE_SIZE transfers.
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  3edd1f2b
- tools: add option to stop on error · b4eb3036
  Federico Vaga authored Aug 22, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  b4eb3036
- tools: handle seq overflow during test mode 1 · 2fe17105
  Federico Vaga authored Aug 22, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  2fe17105
21 Aug, 2018 7 commits
- drv: set default coalescing time to 10ms · 68d63acc
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  68d63acc
- drv: restore buffer burst size configuration · 398cd3ee
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  398cd3ee
- drv: add test mode on channel 0 · bf09403f
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  bf09403f
- lib: add comment · 007d1f05
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  007d1f05
- tools: set irq coalesing only when user ask for it · 62fe0b3a
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  62fe0b3a
- tools: improve printing code · f55e86f9
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  f55e86f9
- drv: improve debug message · 45003981
  Federico Vaga authored Aug 21, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  45003981
20 Aug, 2018 10 commits

tools:tstamp: improve testing mode output · c3f23abb
Federico Vaga authored Aug 20, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
c3f23abb
lib: make the code simpler to read · 2ef88fe4
Federico Vaga authored Aug 20, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
2ef88fe4
lib: fix to_ns convertion · 716b7a61
Federico Vaga authored Aug 17, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
716b7a61
lib: convert all aquired timestamps in a dedicated func · a869ec6b
Federico Vaga authored Aug 17, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
a869ec6b

*: set timeout coalescing dynamically · 013a37f9

Federico Vaga authored Aug 16, 2018

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 <federico.vaga@cern.ch>

013a37f9

drv: bugfix, do not touch burst size dynamically · 8f34868a

Federico Vaga authored Aug 17, 2018

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 <federico.vaga@cern.ch>

8f34868a

drv: WARN if spurious IRQ · 19d22fd4
Federico Vaga authored Aug 20, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
19d22fd4
drv: continue acquisition on ZIO failure · 4280de0d
Federico Vaga authored Aug 20, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
4280de0d
drv: add bitfield shift value for channels · 1627ac98
Federico Vaga authored Aug 20, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
1627ac98

drv: bugfix ZIO names are 12bytes, make this 11 · cc39d6ac

Federico Vaga authored Aug 17, 2018

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: Federico Vaga <federico.vaga@cern.ch>

cc39d6ac

16 Aug, 2018 4 commits
- tools:tstamp: init pollfd array · a91b9e35
  Federico Vaga authored Aug 16, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  a91b9e35
- tools:tstamp: add testing mode · 5a8e8f55
  Federico Vaga authored Aug 16, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  5a8e8f55
- drv: fix fmc-bus type · 34941a92
  Federico Vaga authored Aug 16, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  34941a92
- drv: print only on DMA error · 3219d260
  Federico Vaga authored Aug 16, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  3219d260
15 Aug, 2018 4 commits
- drv: add helper to configure gennum · 5dc76e3e
  Federico Vaga authored Aug 15, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  5dc76e3e
- drv: add some ZIO helpers · d7bb4319
  Federico Vaga authored Aug 15, 2018
```
They should be implemented in ZIO
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  d7bb4319
- drv: move out HW_BUSY flag set|get and use ZIO API · e9567174
  Federico Vaga authored Aug 15, 2018
```
the HW_BUSY flag must be cleared before doing data_done in order to hold
only during DMA.
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  e9567174
- drv: relax CPU on active wait · 31d98b13
  Federico Vaga authored Aug 15, 2018
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  31d98b13