Thanks Timo. We could argue for introducing mlxreg in an SRU as being hardware enablement (SRU team would be the judge), as long as we could clearly demonstrate a negligible regression risk in doing so. I took a stab at a backport to gauge that feasibility. Unfortunately, I found that this is non-trivial. I identified the following set of commits as touching mlxreg, or other code that mlxreg introduced (e.g. adb_parser):
e425ca0 Porting mlxreg - with bug fixes 844101e Fix build issues with expat and boost - bugs 1732270 and 1732354 349d693 Update Fix build issues with expat and boost - bugs 1732270 and 1732354 00fe460 Update Makefile.am 81ab091 Fix format security issue for ununtu 27762b9 Fix issues in catching exceptios and add mstreg in specs 5f78cec Fix issues in adb_parser and fw utilities - 1751157 and 1751189 015de1f Title: Porting from MFT to mstflint bbc94a6 Title: mlxreg and fwtrace porting fix 60da93e Title: Removed openssl dependency from adb generic tools. 0b0e962 Porting from MFT to mstflint 4470808 Fixed mlxreg behaviour with autogenerated subnodes without access attr d27d866 Fix mstreg access type field display 1d2cc0c Globally disable RoCE through mst Some of these commits touch code all over the place w/o providing a technical explanation other than "porting from MFT". Some also add build and runtime dependencies (e.g. boost libraries) that would need to be vetted. While I'm not at a point where I would say "not possible", it would certainly require some engineering. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1864475 Title: Mstflint package - add RoCE disable support in Ubuntu 18.04 To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mstflint/+bug/1864475/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
