From: Andrew Jones <ajones@ventanamicro.com>
To: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Cc: qemu-devel@nongnu.org, qemu-riscv@nongnu.org,
alistair.francis@wdc.com, bmeng@tinylab.org,
liweiwei@iscas.ac.cn, zhiwei_liu@linux.alibaba.com,
palmer@rivosinc.com
Subject: Re: [PATCH RESEND v8 17/20] target/riscv/cpu.c: introduce RISCVCPUMultiExtConfig
Date: Thu, 31 Aug 2023 15:43:57 +0200 [thread overview]
Message-ID: <20230831-b89b2b92f98ad55ab70bb431@orel> (raw)
In-Reply-To: <20230824221440.484675-18-dbarboza@ventanamicro.com>
On Thu, Aug 24, 2023 at 07:14:37PM -0300, Daniel Henrique Barboza wrote:
> If we want to make better decisions when auto-enabling extensions during
> realize() we need a way to tell if an user set an extension manually.
> The RISC-V KVM driver has its own solution via a KVMCPUConfig struct
> that has an 'user_set' flag that is set during the Property set()
> callback. The set() callback also does init() time validations based on
> the current KVM driver capabilities.
>
> For TCG we would want a 'user_set' mechanic too, but we would look
> ad-hoc via cpu_cfg_ext_auto_update() if a certain extension was user set
> or not. If we copy what was made in the KVM side we would look for
> 'user_set' for one into 60+ extension structs spreaded in 3 arrays
> (riscv_cpu_extensions, riscv_cpu_experimental_exts,
> riscv_cpu_vendor_exts).
>
> We'll still need an extension struct but we won't be using the
> 'user_set' flag:
>
> - 'RISCVCPUMultiExtConfig' will be our specialized structure, similar to what
> we're already doing with the MISA extensions in 'RISCVCPUMisaExtConfig'.
> DEFINE_PROP_BOOL() for all 3 extensions arrays were replaced by
> MULTI_EXT_CFG_BOOL(), a macro that will init our specialized struct;
>
> - the 'multi_ext_user_opts' hash will be used to store the offset of each
> extension that the user set via the set() callback, cpu_set_multi_ext_cfg().
> For now we're just initializing and populating it - next patch will use
> it to determine if a certain extension was user set;
>
> - cpu_add_multi_ext_prop() is a new helper that will replace the
> qdev_property_add_static() calls that our macros are doing to populate
> user properties. The macro was renamed to ADD_CPU_MULTIEXT_PROPS_ARRAY()
> for clarity. Note that the non-extension properties in
> riscv_cpu_options[] still need to be declared via qdev().
>
> Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
> Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
> ---
> target/riscv/cpu.c | 257 ++++++++++++++++++++++++++++-----------------
> 1 file changed, 158 insertions(+), 99 deletions(-)
>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
next prev parent reply other threads:[~2023-08-31 13:44 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-24 22:14 [PATCH RESEND v8 00/20] riscv: 'max' CPU, detect user choice in TCG Daniel Henrique Barboza
2023-08-24 22:14 ` [PATCH RESEND v8 01/20] target/riscv/cpu.c: split CPU options from riscv_cpu_extensions[] Daniel Henrique Barboza
2023-08-31 12:49 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 02/20] target/riscv/cpu.c: skip 'bool' check when filtering KVM props Daniel Henrique Barboza
2023-08-31 12:49 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 03/20] target/riscv/cpu.c: split kvm prop handling to its own helper Daniel Henrique Barboza
2023-08-31 12:51 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 04/20] target/riscv: add DEFINE_PROP_END_OF_LIST() to riscv_cpu_options[] Daniel Henrique Barboza
2023-08-31 12:54 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 05/20] target/riscv/cpu.c: split non-ratified exts from riscv_cpu_extensions[] Daniel Henrique Barboza
2023-08-31 12:56 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 06/20] target/riscv/cpu.c: split vendor " Daniel Henrique Barboza
2023-08-31 12:57 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 07/20] target/riscv/cpu.c: add riscv_cpu_add_qdev_prop_array() Daniel Henrique Barboza
2023-08-31 13:01 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 08/20] target/riscv/cpu.c: add riscv_cpu_add_kvm_unavail_prop_array() Daniel Henrique Barboza
2023-08-31 13:03 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 09/20] target/riscv/cpu.c: limit cfg->vext_spec log message Daniel Henrique Barboza
2023-08-24 22:14 ` [PATCH RESEND v8 10/20] target/riscv: add 'max' CPU type Daniel Henrique Barboza
2023-08-31 13:11 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 11/20] avocado, risc-v: add opensbi tests for 'max' CPU Daniel Henrique Barboza
2023-08-31 13:16 ` Andrew Jones
2023-08-31 14:20 ` Daniel Henrique Barboza
2023-08-24 22:14 ` [PATCH RESEND v8 12/20] target/riscv: deprecate the 'any' CPU type Daniel Henrique Barboza
2023-08-31 13:19 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 13/20] target/riscv/cpu.c: use offset in isa_ext_is_enabled/update_enabled Daniel Henrique Barboza
2023-08-31 13:20 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 14/20] target/riscv: make CPUCFG() macro public Daniel Henrique Barboza
2023-08-31 13:22 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 15/20] target/riscv/cpu.c: introduce cpu_cfg_ext_auto_update() Daniel Henrique Barboza
2023-08-31 13:33 ` Andrew Jones
2023-08-31 14:12 ` Daniel Henrique Barboza
2023-08-24 22:14 ` [PATCH RESEND v8 16/20] target/riscv/cpu.c: use cpu_cfg_ext_auto_update() during realize() Daniel Henrique Barboza
2023-08-31 13:37 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 17/20] target/riscv/cpu.c: introduce RISCVCPUMultiExtConfig Daniel Henrique Barboza
2023-08-31 13:43 ` Andrew Jones [this message]
2023-08-24 22:14 ` [PATCH RESEND v8 18/20] target/riscv: use isa_ext_update_enabled() in init_max_cpu_extensions() Daniel Henrique Barboza
2023-08-31 13:56 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 19/20] target/riscv/cpu.c: honor user choice in cpu_cfg_ext_auto_update() Daniel Henrique Barboza
2023-08-31 13:58 ` Andrew Jones
2023-08-24 22:14 ` [PATCH RESEND v8 20/20] target/riscv/cpu.c: consider user option with RVG Daniel Henrique Barboza
2023-08-31 14:02 ` Andrew Jones
2023-08-31 15:43 ` Daniel Henrique Barboza
2023-08-31 14:05 ` [PATCH RESEND v8 00/20] riscv: 'max' CPU, detect user choice in TCG Andrew Jones
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=20230831-b89b2b92f98ad55ab70bb431@orel \
--to=ajones@ventanamicro.com \
--cc=alistair.francis@wdc.com \
--cc=bmeng@tinylab.org \
--cc=dbarboza@ventanamicro.com \
--cc=liweiwei@iscas.ac.cn \
--cc=palmer@rivosinc.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).