Repository navigation
Major Release 7.0.0
Pull Requests Since 6.12.0
Unlabeled
- #251 - GpuAsyncCore V4
- #253 - re-sync rogue's driver file updates back to aes-stream-driver
- #250 - include/GpuAsyncRegs.h: Remove ReadDetectEn registers
- #252 - sync-ing rogue@PR#1117 back to original .h source files
- #248 - driver/gpu_async.c: Check for max datagpu version, error out if unsupported
- #254 - common/driver: Remove uses of GFP_DMA flag
Pull Request Details
driver/gpu_async.c: Check for max datagpu version, error out if unsupported
| Author: | Jeremy L. lorelli@slac.stanford.edu |
| Date: | Wed Feb 18 12:42:46 2026 -0800 |
| Pull: | #248 (13 additions, 3 deletions, 1 files changed) |
| Branch: | slaclab/pr-gpu-ver-check |
Notes:
Description
Avoid initializing the driver on devices with unsupported GpuAsync versions.
It will simply disable the GPU path if unsupported, allowing them to still be used as ordinary datadevs.
include/GpuAsyncRegs.h: Remove ReadDetectEn registers
| Author: | Jeremy L. lorelli@slac.stanford.edu |
| Date: | Wed Feb 18 15:55:48 2026 -0800 |
| Pull: | #250 (66 additions, 10 deletions, 3 files changed) |
| Branch: | slaclab/pr-remove-remote-read-en-regs |
Notes:
Description
These registers are completely redundant. The firmware detects writes to the RemoteReadSize registers and uses that to trigger the read. While this register "exists" in the firmware, it's not actually a distinct register from the software point of view.
GpuAsyncCore V4
| Author: | Larry Ruckman ruckman@slac.stanford.edu |
| Date: | Mon Mar 23 21:15:08 2026 -0700 |
| Pull: | #251 (839 additions, 286 deletions, 8 files changed) |
| Branch: | slaclab/pr-gpu-async-v4 |
Notes:
Description
WIP support for GpuAsyncCore v4. Per slaclab/axi-pcie-core#178
sync-ing rogue@PR#1117 back to original .h source files
| Author: | Larry Ruckman ruckman@slac.stanford.edu |
| Date: | Thu Feb 26 12:42:19 2026 -0800 |
| Pull: | #252 (17 additions, 11 deletions, 1 files changed) |
| Branch: | slaclab/doc-update-4 |
Notes:
Description
- As this is a copied file from aes-stream-drivers to rogue, it is important to make sure we keep these two files in-sync with each other (even if it documentation only changes)
re-sync rogue's driver file updates back to aes-stream-driver
| Author: | Larry Ruckman ruckman@slac.stanford.edu |
| Date: | Fri Feb 27 22:50:59 2026 -0800 |
| Pull: | #253 (265 additions, 206 deletions, 2 files changed) |
| Branch: | slaclab/rogue-header-resync |
Notes:
Description
- Copying the files from the rogue v6.8.3 tag release
- there is no functional changes. Only comment documentation changes
common/driver: Remove uses of GFP_DMA flag
| Author: | Jeremy L. lorelli@slac.stanford.edu |
| Date: | Fri Mar 13 11:55:24 2026 -0700 |
| Pull: | #254 (5 additions, 5 deletions, 2 files changed) |
| Branch: | slaclab/pr-remove-GFP_DMA |
Notes:
Description
This flag is mostly intended for legacy ISA devices, which can only handle addresses in the lowest 16MB of memory (part of ZONE_DMA).
I'm not sure that this flag was even doing anything here, since none of the addresses actually came from that pool. Nonetheless, this flag isn't necessary.
See: https://docs.kernel.org/core-api/mm-api.html#mm-api-gfp-flags
GFP_DMA exists for historical reasons and should be avoided where possible. The flags indicates that the caller requires that the lowest zone be used (ZONE_DMA or 16M on x86-64). Ideally, this would be removed but it would require careful auditing as some users really require it and others use the flag to avoid lowmem reserves in ZONE_DMA and treat the lowest zone as a type of emergency reserve.