From: Atish Patra <atishp@atishpatra.org> To: Bin Meng <bmeng.cn@gmail.com> Cc: "open list:RISC-V" <qemu-riscv@nongnu.org>, Bin Meng <bin.meng@windriver.com>, "qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>, Atish Patra <atish.patra@wdc.com>, Alistair Francis <alistair.francis@wdc.com>, Palmer Dabbelt <palmer@dabbelt.com> Subject: Re: [ RFC v2 1/9] target/riscv: Fix PMU CSR predicate function Date: Thu, 16 Sep 2021 11:39:11 -0700 [thread overview] Message-ID: <CAOnJCUK+8trfTaB3R3-awE4_iEzq9m+uq6jDu+iFgfjJDegBmA@mail.gmail.com> (raw) In-Reply-To: <CAEUhbmXEiLiskioH=ZZa7W=naZHWC8288GAkbeakTfC21MEAUg@mail.gmail.com> On Wed, Sep 15, 2021 at 7:51 AM Bin Meng <bmeng.cn@gmail.com> wrote: > > On Fri, Sep 10, 2021 at 4:27 AM Atish Patra <atish.patra@wdc.com> wrote: > > > > Currently, the predicate function for PMU related CSRs only works if > > virtualization is enabled. Ideally, they should check the mcountern > > bits before cycle/minstret/hpmcounterx access. The predicate function > > also calculates the counter index incorrectly for hpmcounterx. > > > > Signed-off-by: Atish Patra <atish.patra@wdc.com> > > --- > > target/riscv/csr.c | 62 +++++++++++++++++++++++++++++++++++++++++++--- > > 1 file changed, 58 insertions(+), 4 deletions(-) > > > > diff --git a/target/riscv/csr.c b/target/riscv/csr.c > > index 9a4ed18ac597..0515d851b948 100644 > > --- a/target/riscv/csr.c > > +++ b/target/riscv/csr.c > > @@ -62,12 +62,64 @@ static RISCVException ctr(CPURISCVState *env, int csrno) > > #if !defined(CONFIG_USER_ONLY) > > CPUState *cs = env_cpu(env); > > RISCVCPU *cpu = RISCV_CPU(cs); > > + int ctr_index; > > > > if (!cpu->cfg.ext_counters) { > > /* The Counters extensions is not enabled */ > > return RISCV_EXCP_ILLEGAL_INST; > > } > > > > + if (env->priv == PRV_S) { > > + switch (csrno) { > > + case CSR_CYCLE: > > + if (!get_field(env->mcounteren, HCOUNTEREN_CY)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_TIME: > > + if (!get_field(env->mcounteren, HCOUNTEREN_TM)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_INSTRET: > > + if (!get_field(env->mcounteren, HCOUNTEREN_IR)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_HPMCOUNTER3...CSR_HPMCOUNTER31: > > + ctr_index = csrno - CSR_HPMCOUNTER3 + 3; > > + if (!get_field(env->mcounteren, 1 << ctr_index)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + } > > + if (riscv_cpu_is_32bit(env)) { > > + switch (csrno) { > > + case CSR_CYCLEH: > > + if (!get_field(env->mcounteren, HCOUNTEREN_CY)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_TIMEH: > > + if (!get_field(env->mcounteren, HCOUNTEREN_TM)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_INSTRETH: > > + if (!get_field(env->mcounteren, HCOUNTEREN_IR)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_HPMCOUNTER3H...CSR_HPMCOUNTER31H: > > + ctr_index = csrno - CSR_HPMCOUNTER3H + 3; > > + if (!get_field(env->mcounteren, 1 << ctr_index)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + } > > + } > > + } > > + > > if (riscv_cpu_virt_enabled(env)) { > > switch (csrno) { > > case CSR_CYCLE: > > @@ -89,8 +141,9 @@ static RISCVException ctr(CPURISCVState *env, int csrno) > > } > > break; > > case CSR_HPMCOUNTER3...CSR_HPMCOUNTER31: > > - if (!get_field(env->hcounteren, 1 << (csrno - CSR_HPMCOUNTER3)) && > > - get_field(env->mcounteren, 1 << (csrno - CSR_HPMCOUNTER3))) { > > + ctr_index = csrno - CSR_HPMCOUNTER3 + 3; > > ctr_index = csrno - CSR_CYCLE; Will update it. > > > + if (!get_field(env->hcounteren, 1 << ctr_index) && > > + get_field(env->mcounteren, 1 << ctr_index)) { > > This fix (along with the H part below) should be put in a separate patch. Sure. > > > return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; > > } > > break; > > @@ -116,8 +169,9 @@ static RISCVException ctr(CPURISCVState *env, int csrno) > > } > > break; > > case CSR_HPMCOUNTER3H...CSR_HPMCOUNTER31H: > > - if (!get_field(env->hcounteren, 1 << (csrno - CSR_HPMCOUNTER3H)) && > > - get_field(env->mcounteren, 1 << (csrno - CSR_HPMCOUNTER3H))) { > > + ctr_index = csrno - CSR_HPMCOUNTER3H + 3; > > ctr_index = csrno - CSR_CYCLEH; > > > + if (!get_field(env->hcounteren, 1 << ctr_index) && > > + get_field(env->mcounteren, 1 << ctr_index)) { > > return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; > > } > > break; > > -- > > You may need to rebase the patch on: > http://patchwork.ozlabs.org/project/qemu-devel/patch/20210915084601.24304-1-bmeng.cn@gmail.com/ > Sure. > Regards, > Bin > -- Regards, Atish
WARNING: multiple messages have this Message-ID (diff)
From: Atish Patra <atishp@atishpatra.org> To: Bin Meng <bmeng.cn@gmail.com> Cc: Atish Patra <atish.patra@wdc.com>, Palmer Dabbelt <palmer@dabbelt.com>, Bin Meng <bin.meng@windriver.com>, Alistair Francis <alistair.francis@wdc.com>, "qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>, "open list:RISC-V" <qemu-riscv@nongnu.org> Subject: Re: [ RFC v2 1/9] target/riscv: Fix PMU CSR predicate function Date: Thu, 16 Sep 2021 11:39:11 -0700 [thread overview] Message-ID: <CAOnJCUK+8trfTaB3R3-awE4_iEzq9m+uq6jDu+iFgfjJDegBmA@mail.gmail.com> (raw) In-Reply-To: <CAEUhbmXEiLiskioH=ZZa7W=naZHWC8288GAkbeakTfC21MEAUg@mail.gmail.com> On Wed, Sep 15, 2021 at 7:51 AM Bin Meng <bmeng.cn@gmail.com> wrote: > > On Fri, Sep 10, 2021 at 4:27 AM Atish Patra <atish.patra@wdc.com> wrote: > > > > Currently, the predicate function for PMU related CSRs only works if > > virtualization is enabled. Ideally, they should check the mcountern > > bits before cycle/minstret/hpmcounterx access. The predicate function > > also calculates the counter index incorrectly for hpmcounterx. > > > > Signed-off-by: Atish Patra <atish.patra@wdc.com> > > --- > > target/riscv/csr.c | 62 +++++++++++++++++++++++++++++++++++++++++++--- > > 1 file changed, 58 insertions(+), 4 deletions(-) > > > > diff --git a/target/riscv/csr.c b/target/riscv/csr.c > > index 9a4ed18ac597..0515d851b948 100644 > > --- a/target/riscv/csr.c > > +++ b/target/riscv/csr.c > > @@ -62,12 +62,64 @@ static RISCVException ctr(CPURISCVState *env, int csrno) > > #if !defined(CONFIG_USER_ONLY) > > CPUState *cs = env_cpu(env); > > RISCVCPU *cpu = RISCV_CPU(cs); > > + int ctr_index; > > > > if (!cpu->cfg.ext_counters) { > > /* The Counters extensions is not enabled */ > > return RISCV_EXCP_ILLEGAL_INST; > > } > > > > + if (env->priv == PRV_S) { > > + switch (csrno) { > > + case CSR_CYCLE: > > + if (!get_field(env->mcounteren, HCOUNTEREN_CY)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_TIME: > > + if (!get_field(env->mcounteren, HCOUNTEREN_TM)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_INSTRET: > > + if (!get_field(env->mcounteren, HCOUNTEREN_IR)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_HPMCOUNTER3...CSR_HPMCOUNTER31: > > + ctr_index = csrno - CSR_HPMCOUNTER3 + 3; > > + if (!get_field(env->mcounteren, 1 << ctr_index)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + } > > + if (riscv_cpu_is_32bit(env)) { > > + switch (csrno) { > > + case CSR_CYCLEH: > > + if (!get_field(env->mcounteren, HCOUNTEREN_CY)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_TIMEH: > > + if (!get_field(env->mcounteren, HCOUNTEREN_TM)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_INSTRETH: > > + if (!get_field(env->mcounteren, HCOUNTEREN_IR)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + case CSR_HPMCOUNTER3H...CSR_HPMCOUNTER31H: > > + ctr_index = csrno - CSR_HPMCOUNTER3H + 3; > > + if (!get_field(env->mcounteren, 1 << ctr_index)) { > > + return RISCV_EXCP_ILLEGAL_INST; > > + } > > + break; > > + } > > + } > > + } > > + > > if (riscv_cpu_virt_enabled(env)) { > > switch (csrno) { > > case CSR_CYCLE: > > @@ -89,8 +141,9 @@ static RISCVException ctr(CPURISCVState *env, int csrno) > > } > > break; > > case CSR_HPMCOUNTER3...CSR_HPMCOUNTER31: > > - if (!get_field(env->hcounteren, 1 << (csrno - CSR_HPMCOUNTER3)) && > > - get_field(env->mcounteren, 1 << (csrno - CSR_HPMCOUNTER3))) { > > + ctr_index = csrno - CSR_HPMCOUNTER3 + 3; > > ctr_index = csrno - CSR_CYCLE; Will update it. > > > + if (!get_field(env->hcounteren, 1 << ctr_index) && > > + get_field(env->mcounteren, 1 << ctr_index)) { > > This fix (along with the H part below) should be put in a separate patch. Sure. > > > return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; > > } > > break; > > @@ -116,8 +169,9 @@ static RISCVException ctr(CPURISCVState *env, int csrno) > > } > > break; > > case CSR_HPMCOUNTER3H...CSR_HPMCOUNTER31H: > > - if (!get_field(env->hcounteren, 1 << (csrno - CSR_HPMCOUNTER3H)) && > > - get_field(env->mcounteren, 1 << (csrno - CSR_HPMCOUNTER3H))) { > > + ctr_index = csrno - CSR_HPMCOUNTER3H + 3; > > ctr_index = csrno - CSR_CYCLEH; > > > + if (!get_field(env->hcounteren, 1 << ctr_index) && > > + get_field(env->mcounteren, 1 << ctr_index)) { > > return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; > > } > > break; > > -- > > You may need to rebase the patch on: > http://patchwork.ozlabs.org/project/qemu-devel/patch/20210915084601.24304-1-bmeng.cn@gmail.com/ > Sure. > Regards, > Bin > -- Regards, Atish
next prev parent reply other threads:[~2021-09-16 18:43 UTC|newest] Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-09-09 20:26 [ RFC v2 0/9] Improve PMU support Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 1/9] target/riscv: Fix PMU CSR predicate function Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-15 14:49 ` Bin Meng 2021-09-15 14:49 ` Bin Meng 2021-09-16 18:39 ` Atish Patra [this message] 2021-09-16 18:39 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 2/9] target/riscv: pmu: Rename the counters extension to pmu Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-15 14:49 ` Bin Meng 2021-09-15 14:49 ` Bin Meng 2021-09-16 4:49 ` Alistair Francis 2021-09-16 4:49 ` Alistair Francis 2021-09-09 20:26 ` [ RFC v2 3/9] target/riscv: pmu: Make number of counters configurable Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-15 14:49 ` Bin Meng 2021-09-15 14:49 ` Bin Meng 2021-09-16 18:52 ` Atish Patra 2021-09-16 18:52 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 4/9] target/riscv: Implement mcountinhibit CSR Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-15 14:49 ` Bin Meng 2021-09-15 14:49 ` Bin Meng 2021-09-16 18:47 ` Atish Patra 2021-09-16 18:47 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 5/9] target/riscv: Add support for hpmcounters/hpmevents Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 6/9] target/riscv: Support mcycle/minstret write operation Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 7/9] target/riscv: Add sscofpmf extension support Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 8/9] target/riscv: Add few cache related PMU events Atish Patra 2021-09-09 20:26 ` Atish Patra 2021-09-09 20:26 ` [ RFC v2 9/9] hw/riscv: virt: Add PMU DT node to the device tree Atish Patra 2021-09-09 20:26 ` Atish Patra
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=CAOnJCUK+8trfTaB3R3-awE4_iEzq9m+uq6jDu+iFgfjJDegBmA@mail.gmail.com \ --to=atishp@atishpatra.org \ --cc=alistair.francis@wdc.com \ --cc=atish.patra@wdc.com \ --cc=bin.meng@windriver.com \ --cc=bmeng.cn@gmail.com \ --cc=palmer@dabbelt.com \ --cc=qemu-devel@nongnu.org \ --cc=qemu-riscv@nongnu.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.