From: Thomas Huth <thuth@redhat.com>
To: Ivan Klokov <ivan.klokov@syntacore.com>, qemu-devel@nongnu.org
Cc: qemu-riscv@nongnu.org, palmer@dabbelt.com, bmeng.cn@gmail.com,
liwei1518@gmail.com, dbarboza@ventanamicro.com,
zhiwei_liu@linux.alibaba.com, lvivier@redhat.com,
pbonzini@redhat.com
Subject: Re: [RFC PATCH v2 1/2] Add RISC-V CSR qtest support
Date: Tue, 18 Jun 2024 09:47:15 +0200 [thread overview]
Message-ID: <cfb98ee4-00e8-4d2e-9a08-be63164b93b2@redhat.com> (raw)
In-Reply-To: <20240618064443.6474-2-ivan.klokov@syntacore.com>
On 18/06/2024 08.44, Ivan Klokov wrote:
> The RISC-V architecture supports the creation of custom
> CSR-mapped devices. It would be convenient to test them in the same way
> as MMIO-mapped devices. To do this, a new call has been added
> to read/write CSR registers.
>
> Signed-off-by: Ivan Klokov <ivan.klokov@syntacore.com>
> ---
...
> diff --git a/target/riscv/csr.c b/target/riscv/csr.c
> index 58ef7079dc..82540ae5dc 100644
> --- a/target/riscv/csr.c
> +++ b/target/riscv/csr.c
> @@ -29,7 +29,7 @@
> #include "sysemu/cpu-timers.h"
> #include "qemu/guest-random.h"
> #include "qapi/error.h"
> -
> +#include "tests/qtest/libqtest.h"
>
> /* CSR function table public API */
> void riscv_get_csr_ops(int csrno, riscv_csr_operations *ops)
> @@ -4549,6 +4549,53 @@ static RISCVException write_jvt(CPURISCVState *env, int csrno,
> return RISCV_EXCP_NONE;
> }
>
> +static uint64_t csr_call(char *cmd, uint64_t cpu_num, int csrno,
> + uint64_t *val)
> +{
> + RISCVCPU *cpu = RISCV_CPU(cpu_by_arch_id(cpu_num));
> + CPURISCVState *env = &cpu->env;
> +
> + int ret = RISCV_EXCP_NONE;
> + if (strcmp(cmd, "get_csr") == 0) {
> + ret = riscv_csrrw(env, csrno, (target_ulong *)val, 0, 0);
> +
> + } else if (strcmp(cmd, "set_csr") == 0) {
> + ret = riscv_csrrw(env, csrno, NULL, *(target_ulong *)val, MAKE_64BIT_MASK(0, TARGET_LONG_BITS));
> + }
> +
> + if (ret == RISCV_EXCP_NONE) {
> + ret = 0;
> + }
Is there a reason for ignoring errors here? If not, I'd rather replace that
final if-statement with:
else {
g_assert_not_reached();
}
to make sure that mistakes in setting the right sub-command don't get
ignored without any error message.
> + return ret;
> +}
> +
> +bool csr_qtest_callback(CharBackend *chr, gchar **words)
> +{
> + if (strcmp(words[0], "csr") == 0) {
> +
> + uint64_t res, cpu;
> +
> + uint64_t val;
> + int rc, csr;
> +
> + rc = qemu_strtou64(words[2], NULL, 0, &cpu);
> + g_assert(rc == 0);
> + rc = qemu_strtoi(words[3], NULL, 0, &csr);
> + g_assert(rc == 0);
> + rc = qemu_strtou64(words[4], NULL, 0, &val);
> + g_assert(rc == 0);
> + res = csr_call(words[1], cpu, csr, &val);
> +
> + qtest_send_prefix(chr);
> + qtest_sendf(chr, "OK %"PRIx64" "TARGET_FMT_lx"\n", res, (target_ulong)val);
> +
> + return true;
> + }
> +
> + return false;
> +}
> +
> /*
> * Control and Status Register function table
> * riscv_csr_operations::predicate() must be provided for an implemented CSR
> diff --git a/tests/qtest/libqos/csr.c b/tests/qtest/libqos/csr.c
> new file mode 100644
> index 0000000000..2dc52fc442
> --- /dev/null
> +++ b/tests/qtest/libqos/csr.c
> @@ -0,0 +1,42 @@
> +/*
> + * QTest RISC-V CSR driver
> + *
> + * Copyright (c) 2024 Syntacore
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "qemu/osdep.h"
> +#include "../libqtest.h"
> +#include "csr.h"
> +
> +static uint64_t qcsr_call(QTestState *qts, const char *name, uint64_t cpu,
> + int csrno, uint64_t *val)
> +{
> + uint64_t res = 0;
> +
> + res = qtest_csr_call(qts, name, cpu, csrno, val);
> +
> + return res;
> +}
> +
> +int qcsr_get_csr(QTestState *qts, uint64_t cpu,
> + int csrno, uint64_t *val)
> +{
> + int res;
> +
> + res = qcsr_call(qts, "get_csr", cpu, csrno, val);
> +
> + return res;
> +}
> +
> +int qcsr_set_csr(QTestState *qts, uint64_t cpu,
> + int csrno, uint64_t *val)
> +{
> + int res;
> +
> + res = qcsr_call(qts, "set_csr", cpu, csrno, val);
> +
> + return res;
> +}
Technically, there does not seem to be anything related to libqos in your
patch set. libqos is a framework for executing tests on various buses, e.g.
to test PCI devices on various host PCI bus implementations. All that is
triggered via qos-test.c. Your CSR test does not seem to fit into that
catogory, so please put that code rather directly in your riscv-csr-test.c
file instead. (unless you want to use it in a lot of other tests in the
future, too, then maybe you could move them as static inlines into the csr.h
header instead).
> diff --git a/tests/qtest/libqos/csr.h b/tests/qtest/libqos/csr.h
> new file mode 100644
> index 0000000000..d953735fe8
> --- /dev/null
> +++ b/tests/qtest/libqos/csr.h
Again, not related to libqos, please move it up to the qtest folder itself.
> @@ -0,0 +1,16 @@
> +/*
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#ifndef LIBQOS_CSR_H
> +#define LIBQOS_CSR_H
> +
> +int qcsr_get_csr(QTestState *qts, uint64_t cpu,
> + int csrno, uint64_t *val);
> +
> +int qcsr_set_csr(QTestState *qts, uint64_t cpu,
> + int csrno, uint64_t *val);
> +
> +
> +#endif /* LIBQOS_CSR_H */
> diff --git a/tests/qtest/libqos/meson.build b/tests/qtest/libqos/meson.build
> index 558eb4c24b..a944febbd8 100644
> --- a/tests/qtest/libqos/meson.build
> +++ b/tests/qtest/libqos/meson.build
> @@ -25,6 +25,9 @@ libqos_srcs = files(
> # usb
> 'usb.c',
>
> + #riscv csr
> + 'csr.c',
> +
> # qgraph devices:
> 'e1000e.c',
> 'i2c.c',
> diff --git a/tests/qtest/libqtest.c b/tests/qtest/libqtest.c
> index 18e2f7f282..4667d8d873 100644
> --- a/tests/qtest/libqtest.c
> +++ b/tests/qtest/libqtest.c
> @@ -1200,6 +1200,33 @@ uint64_t qtest_rtas_call(QTestState *s, const char *name,
> return 0;
> }
>
> +static void qtest_rsp_csr(QTestState *s, uint64_t *val)
> +{
> + gchar **args;
> + uint64_t ret;
> + int rc;
> +
> + args = qtest_rsp_args(s, 3);
> +
> + rc = qemu_strtou64(args[1], NULL, 16, &ret);
> + g_assert(rc == 0);
> + rc = qemu_strtou64(args[2], NULL, 16, val);
> + g_assert(rc == 0);
> +
> + g_strfreev(args);
> +}
> +
> +uint64_t qtest_csr_call(QTestState *s, const char *name,
> + uint64_t cpu, int csr,
> + uint64_t *val)
> +{
> + qtest_sendf(s, "csr %s 0x%"PRIx64" %d 0x%"PRIx64"\n",
> + name, cpu, csr, *val);
> +
> + qtest_rsp_csr(s, val);
Just a matter of taste, but I think I'd rather inline the contents of
qtest_rsp_csr() here since both functions are not very big yet.
(unless you need qtest_rsp_csr() in another function later, then it's of
course better to keep it separate)
> + return 0;
> +}
Thomas
next prev parent reply other threads:[~2024-06-18 7:48 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-18 6:44 [RFC PATCH v2 0/2] Support RISC-V CSR read/write in Qtest environment Ivan Klokov
2024-06-18 6:44 ` [RFC PATCH v2 1/2] Add RISC-V CSR qtest support Ivan Klokov
2024-06-18 7:47 ` Thomas Huth [this message]
2024-06-18 6:44 ` [RFC PATCH v2 2/2] QTest example for RISC-V CSR register Ivan Klokov
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=cfb98ee4-00e8-4d2e-9a08-be63164b93b2@redhat.com \
--to=thuth@redhat.com \
--cc=bmeng.cn@gmail.com \
--cc=dbarboza@ventanamicro.com \
--cc=ivan.klokov@syntacore.com \
--cc=liwei1518@gmail.com \
--cc=lvivier@redhat.com \
--cc=palmer@dabbelt.com \
--cc=pbonzini@redhat.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).