I was just trying to provide insights that I happened to face with the firmwares from upstream commit 2e92a49f90f9, using kernel version 6.2 and somtehing similar of a TU117 gfx silicon, and that I am still facing very similar issues. Perhaps for the very reason that explicit TU117 firmware was not in that upstream commit. So just based on this, it might not even be a fix for this particular issue.
Not to mention the fact that I, too, had no issues with 5.19 kernel, but since 6.2 I have had plenty, regardless of FW updates (which for TU117 are none?)
But by the looks of it, Ubuntu updates changes tracking is quite confusing, nor am I an expert on which combination of NVIDIA firmware blobs are used with particular gfx silicon, I am out of contributions.
Sorry for attempted hijack.
I was just trying to provide insights that I happened to face with the firmwares from upstream commit 2e92a49f90f9, using kernel version 6.2 and somtehing similar of a TU117 gfx silicon, and that I am still facing very similar issues. Perhaps for the very reason that explicit TU117 firmware was not in that upstream commit. So just based on this, it might not even be a fix for this particular issue.
Not to mention the fact that I, too, had no issues with 5.19 kernel, but since 6.2 I have had plenty, regardless of FW updates (which for TU117 are none?)
But by the looks of it, Ubuntu updates changes tracking is quite confusing, nor am I an expert on which combination of NVIDIA firmware blobs are used with particular gfx silicon, I am out of contributions.