Commits · de80bdd9f8e84f5129d2ba932195371e1dfeca88 · Projects / Simple PCIe FMC carrier SPEC · Open Hardware Repository

30 Jul, 2020 1 commit
- Merge branch 'feature/hdl/top_level_cleanup' into develop · de80bdd9
  Dimitris Lampridis authored Jul 30, 2020
  
  de80bdd9
28 Jul, 2020 9 commits
- hdl: introduce golden-100T and golden-150T · 9288616d
  Dimitris Lampridis authored Jul 28, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  9288616d
- hdl: remove golden_wr-150T, not used anymore. · a399b0fc
  Dimitris Lampridis authored Jul 28, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  a399b0fc
- hdl: rename golden_wr top to spec_base_wr_example and clean-up. · 907cc696
  Dimitris Lampridis authored Jul 28, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  907cc696
- hdl: rename dma_test testbench to golden. · 398dc356
  Dimitris Lampridis authored Jul 28, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  398dc356
- hdl: rename golden to golden-45T. · cb256101
  Dimitris Lampridis authored Jul 28, 2020
  
  cb256101
- hdl: remove "full" top-level, not used anymore. · 659f3d8f
  Dimitris Lampridis authored Jul 28, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  659f3d8f
- hdl: make dma_test testbench use the golden top-level. · bddf0f2c
  Dimitris Lampridis authored Jul 28, 2020
```
Also remove the dma_test top-level, not used anymore.
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  bddf0f2c
- hdl: re-define and clean-up golden to also include DDR3. · 7051bce9
  Dimitris Lampridis authored Jul 28, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  7051bce9
- hdl: add default value to app_wb_i port when not connected. · 13368245
  Dimitris Lampridis authored Jul 28, 2020
  
  13368245
27 Jul, 2020 3 commits
- Merge branch 'bugfix/dma_misalignment' into develop · 903f594d
  Dimitris Lampridis authored Jul 27, 2020
  
  903f594d
- hdl: fix DMA clock to 125MHz · f2dfe5de
  Dimitris Lampridis authored Jul 24, 2020
```
Signed-off-by: Dimitris Lampridis <dimitris.lampridis@cern.ch>
```
  f2dfe5de
- hdl: bump submodules to get latest releases · 682da324
  Dimitris Lampridis authored Jul 24, 2020
  
  682da324
24 Jul, 2020 4 commits
- hdl: Fix correct location for gen_sourceid.py script. · 15e93834
  Dimitris Lampridis authored Jul 13, 2020
```
Depending on whether spec is a submodule of another project or a standalone project.
```
  15e93834
- hdl: introduce simplified top-level for DMA testing and update testbench · 8fbf7438
  Dimitris Lampridis authored Jun 29, 2020
  
  8fbf7438
- syn: constraints path correction · 90be31a5
  Evangelia Gousiou authored Jun 15, 2020
  
  90be31a5
- hdl: introduce DDR DMA testbench · 025c2a5d
  Dimitris Lampridis authored Jun 25, 2020
  
  025c2a5d
13 Jul, 2020 1 commit
- tst: fix skipif statement · 0ff34a6f
  Federico Vaga authored Jul 13, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  0ff34a6f
08 Jul, 2020 4 commits
- tst: remove unclear comment · 3564deba
  Federico Vaga authored Jul 08, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  3564deba
- sw: enforce scatterlist using a module parameter · a509cbd7
  Federico Vaga authored Jul 08, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  a509cbd7
- sw|doc: better explaination about the 4KiB write split · 6bf976cb
  Federico Vaga authored Jul 08, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  6bf976cb
- tst: make --bitstream optional · 3164b2c7
  Federico Vaga authored Jul 08, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  3164b2c7
06 Jul, 2020 5 commits
- sw: add tools to measure throughput · e5879288
  Federico Vaga authored Jul 06, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  e5879288
- tst: DMA read throughput test · 5545419e
  Federico Vaga authored Jul 06, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  5545419e
- sw:drv: add parameter to customize DMA allocation · 01547b49
  Federico Vaga authored Jul 06, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  01547b49
- sw:drv: accept the maximum DMA write length · 2a10539c
  Federico Vaga authored Jul 06, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  2a10539c
- sw:drv: delay DMA WRITE completion to overcome a bug · 3fdabb7a
  Federico Vaga authored Jul 06, 2020
```
The HDL asserts the IRQ line before the end of a write to DDR transfer.
We can't fix the problem on HDL due to lack of resources, so we add a
5us (empirical tests) delay in software.
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  3fdabb7a
02 Jul, 2020 8 commits
- doc: add DMA chapter · 32659261
  Federico Vaga authored Jul 02, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  32659261
- sw:drv: fix DMA size limitations · 7f6fb421
  Federico Vaga authored Jul 02, 2020
```
No limites for read, but the implementation is not symmetric. For
DMA_MEM_TO_DEV the transfer size limitation is 4096.
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  7f6fb421
- fixup! tst: add regression test for 4Byte corruption · a6499126
  Federico Vaga authored Jul 02, 2020
  
  a6499126
- tst: program FPGA before running tests · 7f643fb0
  Federico Vaga authored Jul 02, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  7f643fb0
- fixup! sw:drv: fix bug in computing segments · 67d436d5
  Federico Vaga authored Jul 02, 2020
  
  67d436d5
- sw:py: add method to program the FPGA · ea51ad5e
  Federico Vaga authored Jul 02, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  ea51ad5e
- tst: add regression test for 4Byte corruption · a108596a
  Federico Vaga authored Jul 02, 2020
```
Problem: sometimes a 4Byte transfers corrupt the gennum chip
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  a108596a
- tst: add regression test for zero bytes · 6d9fec0e
  Federico Vaga authored Jul 02, 2020
```
Problem: after Xbytes the DMA read returns only zeros. This happens at
a power-of-two offset, typically below 1024
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  6d9fec0e
01 Jul, 2020 2 commits
- sw:drv: fix bug in computing segments · 6a86876a
  Federico Vaga authored Jul 01, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  6a86876a
- sw:drv: add necessary header on modern Linux · 64ee0f36
  Federico Vaga authored Jul 01, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  64ee0f36
30 Jun, 2020 3 commits
- Merge branch 'feature/debugfs-dma' into develop · 0d2c60fd
  Federico Vaga authored Jun 30, 2020
  
  0d2c60fd
- ignore clangd files · 282dfa29
  Federico Vaga authored Jun 30, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  282dfa29
- tst: simplify assert message and remove useless assert · 6cd4a336
  Federico Vaga authored Jun 30, 2020
```
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
```
  6cd4a336