Message ID | 20220513015522.910856-1-atishp@rivosinc.com (mailing list archive) |
---|---|
Headers | show |
Series | Miscellaneous fixes for PMU driver | expand |
On Thu, 12 May 2022 18:55:19 PDT (-0700), Atish Patra wrote: > This series fixes issues PMU driver code. > PATCH 1 & 3 are fixes for rv32 while PATCH 2 fixes a redundant > user page update issue during counter start. > > Changes from v1->v2: > 1. Add proper compile time rv32 checks. > > Atish Patra (3): > RISC-V: Fix counter restart during overflow for RV32 > RISC-V: Update user page mapping only once during start > RISC-V: Fix SBI PMU calls for RV32 > > drivers/perf/riscv_pmu.c | 1 - > drivers/perf/riscv_pmu_sbi.c | 16 ++++++++++++++++ > 2 files changed, 16 insertions(+), 1 deletion(-) +Will and Mark. Will recently took some stuff for drivers/perf/riscv_* into his for-next/perf tree (which I didn't even know about until then, oops), but the previous stuff I took through the RISC-V tree. Always happy to get stuff out of my tree, just LMK what you want me to do here. I did some minor cleanups to the commit text and put them over at palmer/riscv-pmu_fixes, it's all passing my tests. These generally look fine to me and they're all stable backports, so no big rush on the merge window (which I still have stuff for).
Hey Palmer, On Wed, Jun 01, 2022 at 10:58:08PM -0700, Palmer Dabbelt wrote: > On Thu, 12 May 2022 18:55:19 PDT (-0700), Atish Patra wrote: > > This series fixes issues PMU driver code. > > PATCH 1 & 3 are fixes for rv32 while PATCH 2 fixes a redundant > > user page update issue during counter start. > > > > Changes from v1->v2: > > 1. Add proper compile time rv32 checks. > > > > Atish Patra (3): > > RISC-V: Fix counter restart during overflow for RV32 > > RISC-V: Update user page mapping only once during start > > RISC-V: Fix SBI PMU calls for RV32 > > > > drivers/perf/riscv_pmu.c | 1 - > > drivers/perf/riscv_pmu_sbi.c | 16 ++++++++++++++++ > > 2 files changed, 16 insertions(+), 1 deletion(-) > > +Will and Mark. > > Will recently took some stuff for drivers/perf/riscv_* into his > for-next/perf tree (which I didn't even know about until then, oops), but > the previous stuff I took through the RISC-V tree. Always happy to get > stuff out of my tree, just LMK what you want me to do here. > > I did some minor cleanups to the commit text and put them over at > palmer/riscv-pmu_fixes, it's all passing my tests. These generally look > fine to me and they're all stable backports, so no big rush on the merge > window (which I still have stuff for). I don't have any perf driver fixes queued at the moment, so please just include these changes along with any other riscv fixes you have. Thanks, Will