Repository navigation
Gen8 uses incorrect MUX configuration for compute extended #13
Description
I noticed this the other day while poking around, but for some reason forgot to post an issue, so while I still remember...
The generated configuration code for BDW in i915_oa_bdw.c:select_compute_extended_config incorrectly initialises the MUX registers for INTEL_INFO(dev_priv)->subslice_mask & 0x20.
I guess there is also the issue now of whether or not the mux_config_compute_extended is in fact needed at all...or maybe this should just go in the else?
Activity
perf tools: tracepoint_error() can receive e=NULL, robustify it
mm: fix devm_memremap_pages crash, use mem_hotplug_{begin, done}
dccp/tcp: fix routing redirect race
libnvdimm: fix reconfig_mutex, mmap_sem, and jbd2_handle lockdep splat
xfs: fix up xfs_swap_extent_forks inline extent handling
PCI: Enable access to non-standard VPD for Chelsio devices (cxgb3)
mm: fix devm_memremap_pages crash, use mem_hotplug_{begin, done}
dccp/tcp: fix routing redirect race
libnvdimm: fix reconfig_mutex, mmap_sem, and jbd2_handle lockdep splat
KVM: x86: Fix potential preemption when get the current kvmclock time…
dccp/tcp: fix routing redirect race
libnvdimm: fix reconfig_mutex, mmap_sem, and jbd2_handle lockdep splat
nvme-pci: fix multiple ctrl removal scheduling
fscache: fix race between enablement and dropping of object
crypto: authenc - fix parsing key with misaligned rta_len
crypto: authencesn - Avoid twice completion call in decrypt path
s390/smp: Fix calling smp_call_ipl_cpu() from ipl CPU
mtd: powernv_flash: Fix device registration error
perf top: Delete the evlist before perf_session, fixing heap-use-afte…
perf evsel: Free evsel->counts in perf_evsel__exit()
perf tests: Fix a memory leak of cpu_map object in the openat_syscall…
perf tests: Fix a memory leak in test__perf_evsel__tp_sched_test()
perf tools: Fix map reference counting