From: "Philippe Mathieu-Daudé" <philmd@linaro.org>
To: Daniel Henrique Barboza <dbarboza@ventanamicro.com>,
Andrew Jones <ajones@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, Michael Tokarev <mjt@tls.msk.ru>
Subject: Re: [PATCH v2 10/19] target/riscv: remove kvm-stub.c
Date: Tue, 12 Sep 2023 13:15:08 +0200 [thread overview]
Message-ID: <e095221f-9d15-df10-3aa3-31310dad57f9@linaro.org> (raw)
In-Reply-To: <0aa9885e-f56f-d72d-1d41-773661974746@ventanamicro.com>
On 12/9/23 12:48, Daniel Henrique Barboza wrote:
> On 9/11/23 09:23, Daniel Henrique Barboza wrote:
>> On 9/11/23 06:04, Andrew Jones wrote:
>>> On Mon, Sep 11, 2023 at 09:49:06AM +0200, Andrew Jones wrote:
>>>> On Wed, Sep 06, 2023 at 12:23:19PM +0200, Philippe Mathieu-Daudé wrote:
>>>>> On 6/9/23 11:16, Daniel Henrique Barboza wrote:
>>>>>> This file is not needed for some time now. All the stubs
>>>>>> implemented in
>>>>>> it (kvm_riscv_reset_vcpu() and kvm_riscv_set_irq()) are wrapped in
>>>>>> 'if
>>>>>> kvm_enabled()' blocks that the compiler will rip it out in non-KVM
>>>>>> builds.
>>>>>>
>>>>>> We'll also add non-KVM stubs for all functions declared in
>>>>>> kvm_riscv.h.
>>>>>> All stubs are implemented as g_asserted_not_reached(), meaning
>>>>>> that we
>>>>>> won't support them in non-KVM builds. This is done by other kvm
>>>>>> headers
>>>>>> like kvm_arm.h and kvm_ppc.h.
>>>>>
>>>>> Aren't them also protected by kvm_enabled()? Otherwise shouldn't they?
>>>>
>>>> Yes, I think your earlier suggestion that we always invoke kvm
>>>> functions
>>>> from non-kvm files with a kvm_enabled() guard makes sense.
>>>>
>>>>>
>>>>>> Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
>>>>>> ---
>>>>>> target/riscv/kvm-stub.c | 30 ------------------------------
>>>>>> target/riscv/kvm_riscv.h | 31 +++++++++++++++++++++++++++++++
>>>>>> target/riscv/meson.build | 2 +-
>>>>>> 3 files changed, 32 insertions(+), 31 deletions(-)
>>>>>> delete mode 100644 target/riscv/kvm-stub.c
>>>>>
>>>>>
>>>>>> diff --git a/target/riscv/kvm_riscv.h b/target/riscv/kvm_riscv.h
>>>>>> index f6501e68e2..c9ecd9a967 100644
>>>>>> --- a/target/riscv/kvm_riscv.h
>>>>>> +++ b/target/riscv/kvm_riscv.h
>>>>>> @@ -19,6 +19,7 @@
>>>>>> #ifndef QEMU_KVM_RISCV_H
>>>>>> #define QEMU_KVM_RISCV_H
>>>>>> +#ifdef CONFIG_KVM
>>>>>> void kvm_riscv_cpu_add_kvm_properties(Object *obj);
>>>>>
>>>>> At a glance kvm_riscv_cpu_add_kvm_properties() is.
>>>>> Keep the prototype declared (before #ifdef CONFIG_KVM) is enough
>>>>> for the
>>>>> compiler to elide it.
>>>>
>>>> Yes, when building without CONFIG_KVM enabled it's actually better
>>>> to not
>>>> have the stubs, since the compiler will catch an unguarded kvm function
>>>> call (assuming the kvm function is defined in a file which is only
>>>> built
>>>> with CONFIG_KVM).
>>>>
>>>> Unfortunately we don't have anything to protect developers from
>>>> forgetting
>>>> the kvm_enabled() guard when building a QEMU which supports both TCG
>>>> and
>>>> KVM. We could try to remember to put 'assert(kvm_enabled())' at the
>>>> start
>>>> of each of these types of functions. It looks like mips does that for a
>>>> couple functions.
>>>
>>> Eh, ignore this suggestion. We don't need asserts, because we have
>>> CI. As
>>> long as our CI does a CONFIG_KVM=n build and all KVM functions are in
>>> kvm-
>>> only files, then we'll always catch calls of KVM functions which are
>>> missing their kvm_enabled() guards.
>>
>> So, to sum up, get rid of the 'g_assert_not_reached()' stubs since
>> they should
>> be cropped by well placed kvm_enabled() calls in the code and, if
>> that's not
>> the case, a non-KVM build failure will let us know that there's
>> something wrong.
>>
>> I can live with that. It's less code to deal with in the header as
>> well. Thanks,
>
>
> I changed my mind after talking with Michael in this thread:
>
> https://lists.gnu.org/archive/html/qemu-devel/2023-09/msg02567.html
>
>
> There are differences in how gcc and clang (and probably other
> toolchains) behave
> when choosing to crop code gated with 'if (kvm_enabled && something)'.
> It is
> better to have stubs for all KVM functions to avoid relying on compiler
> behavior.
>
> We should keep the stubs in this patch as is. Thanks,
Don't get stuck at this; get your series merged, we'll have a look
at that later, this is not important.
Regards,
Phil.
next prev parent reply other threads:[~2023-09-12 11:16 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-06 9:16 [PATCH v2 00/19] riscv: split TCG/KVM accelerators from cpu.c Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 01/19] target/riscv: introduce TCG AccelCPUClass Daniel Henrique Barboza
2023-09-19 9:51 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 02/19] target/riscv: move riscv_cpu_realize_tcg() to TCG::cpu_realizefn() Daniel Henrique Barboza
2023-09-19 9:54 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 03/19] target/riscv: move riscv_cpu_validate_set_extensions() to tcg-cpu.c Daniel Henrique Barboza
2023-09-19 10:59 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 04/19] target/riscv: move riscv_tcg_ops " Daniel Henrique Barboza
2023-09-19 11:08 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 05/19] target/riscv/cpu.c: add .instance_post_init() Daniel Henrique Barboza
2023-09-11 7:08 ` Andrew Jones
2023-09-19 9:16 ` LIU Zhiwei
2023-09-19 16:07 ` Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 06/19] target/riscv: move 'host' CPU declaration to kvm.c Daniel Henrique Barboza
2023-09-19 11:11 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 07/19] target/riscv/cpu.c: mark extensions arrays as 'const' Daniel Henrique Barboza
2023-09-19 11:19 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 08/19] target/riscv: move riscv_cpu_add_kvm_properties() to kvm.c Daniel Henrique Barboza
2023-09-19 11:25 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 09/19] target/riscv: make riscv_add_satp_mode_properties() public Daniel Henrique Barboza
2023-09-11 7:09 ` Andrew Jones
2023-09-19 11:26 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 10/19] target/riscv: remove kvm-stub.c Daniel Henrique Barboza
2023-09-06 10:23 ` Philippe Mathieu-Daudé
2023-09-11 7:49 ` Andrew Jones
2023-09-11 9:04 ` Andrew Jones
2023-09-11 12:23 ` Daniel Henrique Barboza
2023-09-12 10:48 ` Daniel Henrique Barboza
2023-09-12 11:15 ` Philippe Mathieu-Daudé [this message]
2023-09-12 12:03 ` Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 11/19] target/riscv: introduce KVM AccelCPUClass Daniel Henrique Barboza
2023-09-11 7:52 ` Andrew Jones
2023-09-19 11:37 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 12/19] target/riscv: move KVM only files to kvm subdir Daniel Henrique Barboza
2023-09-19 11:37 ` LIU Zhiwei
2023-09-06 9:16 ` [PATCH v2 13/19] target/riscv/kvm: do not use riscv_cpu_add_misa_properties() Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 14/19] target/riscv/cpu.c: export set_misa() Daniel Henrique Barboza
2023-09-06 10:18 ` Philippe Mathieu-Daudé
2023-09-11 7:54 ` Andrew Jones
2023-09-06 9:16 ` [PATCH v2 15/19] target/riscv/tcg: introduce tcg_cpu_instance_init() Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 16/19] target/riscv/cpu.c: make misa_ext_cfgs[] 'const' Daniel Henrique Barboza
2023-09-11 7:56 ` Andrew Jones
2023-09-06 9:16 ` [PATCH v2 17/19] target/riscv/tcg: move riscv_cpu_add_misa_properties() to tcg-cpu.c Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 18/19] target/riscv/cpu.c: export isa_edata_arr[] Daniel Henrique Barboza
2023-09-06 9:16 ` [PATCH v2 19/19] target/riscv/cpu: move priv spec functions to tcg-cpu.c Daniel Henrique Barboza
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=e095221f-9d15-df10-3aa3-31310dad57f9@linaro.org \
--to=philmd@linaro.org \
--cc=ajones@ventanamicro.com \
--cc=alistair.francis@wdc.com \
--cc=bmeng@tinylab.org \
--cc=dbarboza@ventanamicro.com \
--cc=liweiwei@iscas.ac.cn \
--cc=mjt@tls.msk.ru \
--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).