On 10 April 2018 at 09:30, Michel Dänzer <michel at daenzer.net> wrote: > On 2018-04-04 04:29 PM, Emil Velikov wrote: >> From: Emil Velikov <emil.velikov at collabora.com> >> >> Signed-off-by: Emil Velikov <emil.l.velikov at gmail.com> >> --- >> todo | 9 +++++++++ >> 1 file changed, 9 insertions(+) >> create mode 100644 todo >> >> diff --git a/todo b/todo >> new file mode 100644 >> index 0000000..10c1ad5 >> --- /dev/null >> +++ b/todo >> @@ -0,0 +1,9 @@ >> + - on amdgpu_probe failure, the pScrn entry is leaked - missing api/examples? > > Might be similar to patch 11; does valgrind actually report a leak if > you force this? > > >> + - introduce xf86ConfigEntity and use it >> + - remove embedded AMDGPUInfoRec::pEnt >> + - consistently use gAMDGPUEntityIndex or getAMDGPUEntityIndex >> + - consistently use of pEnt/entity_num -> pScrn->list[], AMDPRIV >> + - kill off DRI_1_ DRICreatePCIBusID - demote again to DRI1 only in X codebase >> + - compose bus string early & strcmp instead of device_match? >> + - remove embedded AMDGPUInfoRec::PciInfo - reuse EntityInfoRec::chipset, GDevRec::chiIDi, amdgpu_gpu_info::asic_id or ... >> + - use odev to fetch render_node? > > I'm afraid I don't really see these as important enough to be tracked > like this. > Agreed - no reason to keep these in-tree. Idea was to gather feedback on the topics. One example: Do we need the getAMDGPUEntityIndex helper, considering ~half of the existing codebase uses it. Yet other half references gAMDGPUEntityIndex directly. Most of the above, seem to be a copy/paste from the radeon driver, which in turn is a copy from (?) and the original commit lacks any information :-\ -Emil
