From: "CHEN Yi" <chenyi2000@zju.edu.cn>
To: "Alistair Francis" <alistair23@gmail.com>
Cc: qemu-devel@nongnu.org, "Weiwei Li" <liweiwei@iscas.ac.cn>,
"Palmer Dabbelt" <palmer@dabbelt.com>,
"Alistair Francis" <alistair.francis@wdc.com>,
"Bin Meng" <bin.meng@windriver.com>,
"Daniel Henrique Barboza" <dbarboza@ventanamicro.com>,
"Liu Zhiwei" <zhiwei_liu@linux.alibaba.com>,
"open list:RISC-V TCG CPUs" <qemu-riscv@nongnu.org>
Subject: Re: Re: [PATCH v4] target/riscv: fix H extension TVM trap
Date: Thu, 6 Apr 2023 18:24:32 +0800 (GMT+08:00) [thread overview]
Message-ID: <7519b147.2e6e7.1875618b4ab.Coremail.chenyi2000@zju.edu.cn> (raw)
In-Reply-To: <CAKmqyKMC=8oKaduPE3Zt44qLfuQH9b1ckQmqFTog8QTwJB3GVA@mail.gmail.com>
> -----Original Messages-----
> From: "Alistair Francis" <alistair23@gmail.com>
> Sent Time: 2023-04-06 09:56:58 (Thursday)
> To: "Yi Chen" <chenyi2000@zju.edu.cn>
> Cc: qemu-devel@nongnu.org, "Weiwei Li" <liweiwei@iscas.ac.cn>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Alistair Francis" <alistair.francis@wdc.com>, "Bin Meng" <bin.meng@windriver.com>, "Daniel Henrique Barboza" <dbarboza@ventanamicro.com>, "Liu Zhiwei" <zhiwei_liu@linux.alibaba.com>, "open list:RISC-V TCG CPUs" <qemu-riscv@nongnu.org>
> Subject: Re: [PATCH v4] target/riscv: fix H extension TVM trap
>
> On Sun, Mar 12, 2023 at 10:07 PM Yi Chen <chenyi2000@zju.edu.cn> wrote:
> >
> > - Trap satp/hgatp accesses from HS-mode when MSTATUS.TVM is enabled.
> > - Trap satp accesses from VS-mode when HSTATUS.VTVM is enabled.
> > - Raise RISCV_EXCP_ILLEGAL_INST when U-mode executes SFENCE.VMA/SINVAL.VMA.
> > - Raise RISCV_EXCP_VIRT_INSTRUCTION_FAULT when VU-mode executes
> > SFENCE.VMA/SINVAL.VMA or VS-mode executes SFENCE.VMA/SINVAL.VMA with
> > HSTATUS.VTVM enabled.
> > - Raise RISCV_EXCP_VIRT_INSTRUCTION_FAULT when VU-mode executes
> > HFENCE.GVMA/HFENCE.VVMA/HINVAL.GVMA/HINVAL.VVMA.
>
> Thanks for the patch!
>
> It looks like this patch needs to be rebased. Do you mind rebasing it
> on https://github.com/alistair23/qemu/tree/riscv-to-apply.next and
> then re-sending?
Sure. I sent it just now.
> Also, when you are fixing a range of issues it's best to split the
> fixes into patches that fix each individual issue (where that is
> possible). This makes it easier to review but also makes it easier to
> track changes and regressions if any problems arise.
>
> In this case you don't need to split them up for a v5, but in future
> it's something to keep in mind
I see. I will keep that in mind in the future. Thanks for your kind note.
Best,
Yi
> The changes look good otherwise though :)
>
> Alistair
>
> >
> > Signed-off-by: Yi Chen <chenyi2000@zju.edu.cn>
> > Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn>
> > ---
> > Add reviewed-by
> > Replace "env->priv <= PRV_S && riscv_cpu_virt_enabled(env)" with "riscv_cpu_virt_enabled(env)"
> > target/riscv/csr.c | 56 +++++++++++++++++++++++++---------------
> > target/riscv/op_helper.c | 12 ++++-----
> > 2 files changed, 41 insertions(+), 27 deletions(-)
> >
> > diff --git a/target/riscv/csr.c b/target/riscv/csr.c
> > index d522efc0b6..26a02e57bd 100644
> > --- a/target/riscv/csr.c
> > +++ b/target/riscv/csr.c
> > @@ -443,6 +443,30 @@ static RISCVException sstc_32(CPURISCVState *env, int csrno)
> > return sstc(env, csrno);
> > }
> >
> > +static RISCVException satp(CPURISCVState *env, int csrno)
> > +{
> > + if (env->priv == PRV_S && !riscv_cpu_virt_enabled(env) &&
> > + get_field(env->mstatus, MSTATUS_TVM)) {
> > + return RISCV_EXCP_ILLEGAL_INST;
> > + }
> > + if (env->priv == PRV_S && riscv_cpu_virt_enabled(env) &&
> > + get_field(env->hstatus, HSTATUS_VTVM)) {
> > + return RISCV_EXCP_VIRT_INSTRUCTION_FAULT;
> > + }
> > +
> > + return smode(env, csrno);
> > +}
> > +
> > +static RISCVException hgatp(CPURISCVState *env, int csrno)
> > +{
> > + if (env->priv == PRV_S && !riscv_cpu_virt_enabled(env) &&
> > + get_field(env->mstatus, MSTATUS_TVM)) {
> > + return RISCV_EXCP_ILLEGAL_INST;
> > + }
> > +
> > + return hmode(env, csrno);
> > +}
> > +
> > /* Checks if PointerMasking registers could be accessed */
> > static RISCVException pointer_masking(CPURISCVState *env, int csrno)
> > {
> > @@ -2655,13 +2679,7 @@ static RISCVException read_satp(CPURISCVState *env, int csrno,
> > *val = 0;
> > return RISCV_EXCP_NONE;
> > }
> > -
> > - if (env->priv == PRV_S && get_field(env->mstatus, MSTATUS_TVM)) {
> > - return RISCV_EXCP_ILLEGAL_INST;
> > - } else {
> > - *val = env->satp;
> > - }
> > -
> > + *val = env->satp;
> > return RISCV_EXCP_NONE;
> > }
> >
> > @@ -2684,18 +2702,14 @@ static RISCVException write_satp(CPURISCVState *env, int csrno,
> > }
> >
> > if (vm && mask) {
> > - if (env->priv == PRV_S && get_field(env->mstatus, MSTATUS_TVM)) {
> > - return RISCV_EXCP_ILLEGAL_INST;
> > - } else {
> > - /*
> > - * The ISA defines SATP.MODE=Bare as "no translation", but we still
> > - * pass these through QEMU's TLB emulation as it improves
> > - * performance. Flushing the TLB on SATP writes with paging
> > - * enabled avoids leaking those invalid cached mappings.
> > - */
> > - tlb_flush(env_cpu(env));
> > - env->satp = val;
> > - }
> > + /*
> > + * The ISA defines SATP.MODE=Bare as "no translation", but we still
> > + * pass these through QEMU's TLB emulation as it improves
> > + * performance. Flushing the TLB on SATP writes with paging
> > + * enabled avoids leaking those invalid cached mappings.
> > + */
> > + tlb_flush(env_cpu(env));
> > + env->satp = val;
> > }
> > return RISCV_EXCP_NONE;
> > }
> > @@ -4180,7 +4194,7 @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = {
> > .min_priv_ver = PRIV_VERSION_1_12_0 },
> >
> > /* Supervisor Protection and Translation */
> > - [CSR_SATP] = { "satp", smode, read_satp, write_satp },
> > + [CSR_SATP] = { "satp", satp, read_satp, write_satp },
> >
> > /* Supervisor-Level Window to Indirectly Accessed Registers (AIA) */
> > [CSR_SISELECT] = { "siselect", aia_smode, NULL, NULL, rmw_xiselect },
> > @@ -4217,7 +4231,7 @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = {
> > .min_priv_ver = PRIV_VERSION_1_12_0 },
> > [CSR_HGEIP] = { "hgeip", hmode, read_hgeip,
> > .min_priv_ver = PRIV_VERSION_1_12_0 },
> > - [CSR_HGATP] = { "hgatp", hmode, read_hgatp, write_hgatp,
> > + [CSR_HGATP] = { "hgatp", hgatp, read_hgatp, write_hgatp,
> > .min_priv_ver = PRIV_VERSION_1_12_0 },
> > [CSR_HTIMEDELTA] = { "htimedelta", hmode, read_htimedelta,
> > write_htimedelta,
> > diff --git a/target/riscv/op_helper.c b/target/riscv/op_helper.c
> > index 84ee018f7d..8e16020f8d 100644
> > --- a/target/riscv/op_helper.c
> > +++ b/target/riscv/op_helper.c
> > @@ -381,12 +381,12 @@ void helper_wfi(CPURISCVState *env)
> > void helper_tlb_flush(CPURISCVState *env)
> > {
> > CPUState *cs = env_cpu(env);
> > - if (!(env->priv >= PRV_S) ||
> > - (env->priv == PRV_S &&
> > - get_field(env->mstatus, MSTATUS_TVM))) {
> > + if (!riscv_cpu_virt_enabled(env) &&
> > + (env->priv == PRV_U ||
> > + (env->priv == PRV_S && get_field(env->mstatus, MSTATUS_TVM)))) {
> > riscv_raise_exception(env, RISCV_EXCP_ILLEGAL_INST, GETPC());
> > - } else if (riscv_has_ext(env, RVH) && riscv_cpu_virt_enabled(env) &&
> > - get_field(env->hstatus, HSTATUS_VTVM)) {
> > + } else if (riscv_cpu_virt_enabled(env) &&
> > + (env->priv == PRV_U || get_field(env->hstatus, HSTATUS_VTVM))) {
> > riscv_raise_exception(env, RISCV_EXCP_VIRT_INSTRUCTION_FAULT, GETPC());
> > } else {
> > tlb_flush(cs);
> > @@ -403,7 +403,7 @@ void helper_hyp_tlb_flush(CPURISCVState *env)
> > {
> > CPUState *cs = env_cpu(env);
> >
> > - if (env->priv == PRV_S && riscv_cpu_virt_enabled(env)) {
> > + if (riscv_cpu_virt_enabled(env)) {
> > riscv_raise_exception(env, RISCV_EXCP_VIRT_INSTRUCTION_FAULT, GETPC());
> > }
> >
> > --
> > 2.39.2
> >
> >
prev parent reply other threads:[~2023-04-06 10:25 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-12 12:05 [PATCH v4] target/riscv: fix H extension TVM trap Yi Chen
2023-03-17 8:45 ` LIU Zhiwei
2023-04-06 1:56 ` Alistair Francis
2023-04-06 10:24 ` CHEN Yi [this message]
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=7519b147.2e6e7.1875618b4ab.Coremail.chenyi2000@zju.edu.cn \
--to=chenyi2000@zju.edu.cn \
--cc=alistair.francis@wdc.com \
--cc=alistair23@gmail.com \
--cc=bin.meng@windriver.com \
--cc=dbarboza@ventanamicro.com \
--cc=liweiwei@iscas.ac.cn \
--cc=palmer@dabbelt.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-riscv@nongnu.org \
--cc=zhiwei_liu@linux.alibaba.com \
/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: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).