qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Philippe Mathieu-Daudé" <philmd@linaro.org>
To: Richard Henderson <richard.henderson@linaro.org>, qemu-devel@nongnu.org
Cc: Daniel Henrique Barboza <dbarboza@ventanamicro.com>,
	qemu-riscv <qemu-riscv@nongnu.org>
Subject: Re: [PULL 33/42] tcg/riscv: Require TCG_TARGET_REG_BITS == 64
Date: Mon, 24 Feb 2025 10:24:38 +0100	[thread overview]
Message-ID: <af650a49-8917-4d19-9cd4-e29a1538b86b@linaro.org> (raw)
In-Reply-To: <2a2026eb-b92e-450f-9bcf-211620bdf450@linaro.org>

On 22/2/25 19:17, Richard Henderson wrote:
> On 2/20/25 15:27, Philippe Mathieu-Daudé wrote:
>> On 5/5/23 23:24, Richard Henderson wrote:
>>> The port currently does not support "oversize" guests, which
>>> means riscv32 can only target 32-bit guests.  We will soon be
>>> building TCG once for all guests.  This implies that we can
>>> only support riscv64.
>>>
>>> Since all Linux distributions target riscv64 not riscv32,
>>> this is not much of a restriction and simplifies the code.
>>>
>>> The brcond2 and setcond2 opcodes are exclusive to 32-bit hosts,
>>> so we can and should remove the stubs.
>>>
>>> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
>>> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
>>> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
>>> ---
>>>   tcg/riscv/tcg-target-con-set.h |   8 --
>>>   tcg/riscv/tcg-target.h         |  22 ++--
>>>   tcg/riscv/tcg-target.c.inc     | 232 +++++++++------------------------
>>>   3 files changed, 72 insertions(+), 190 deletions(-)
>>
>>
>>> diff --git a/tcg/riscv/tcg-target.h b/tcg/riscv/tcg-target.h
>>> index 0deb33701f..dddf2486c1 100644
>>> --- a/tcg/riscv/tcg-target.h
>>> +++ b/tcg/riscv/tcg-target.h
>>> @@ -25,11 +25,14 @@
>>>   #ifndef RISCV_TCG_TARGET_H
>>>   #define RISCV_TCG_TARGET_H
>>> -#if __riscv_xlen == 32
>>> -# define TCG_TARGET_REG_BITS 32
>>> -#elif __riscv_xlen == 64
>>> -# define TCG_TARGET_REG_BITS 64
>>> +/*
>>> + * We don't support oversize guests.
>>> + * Since we will only build tcg once, this in turn requires a 64-bit 
>>> host.
>>> + */
>>> +#if __riscv_xlen != 64
>>> +#error "unsupported code generation mode"
>>>   #endif
>>> +#define TCG_TARGET_REG_BITS 64
>>>   #define TCG_TARGET_INSN_UNIT_SIZE 4
>>>   #define TCG_TARGET_TLB_DISPLACEMENT_BITS 20
>>> @@ -83,13 +86,8 @@ typedef enum {
>>>   #define TCG_TARGET_STACK_ALIGN          16
>>>   #define TCG_TARGET_CALL_STACK_OFFSET    0
>>>   #define TCG_TARGET_CALL_ARG_I32         TCG_CALL_ARG_NORMAL
>>> -#if TCG_TARGET_REG_BITS == 32
>>> -#define TCG_TARGET_CALL_ARG_I64         TCG_CALL_ARG_EVEN
>>> -#define TCG_TARGET_CALL_ARG_I128        TCG_CALL_ARG_EVEN
>>> -#else
>>>   #define TCG_TARGET_CALL_ARG_I64         TCG_CALL_ARG_NORMAL
>>>   #define TCG_TARGET_CALL_ARG_I128        TCG_CALL_ARG_NORMAL
>>> -#endif
>>>   #define TCG_TARGET_CALL_RET_I128        TCG_CALL_RET_NORMAL
>>>   /* optional instructions */
>>> @@ -106,8 +104,8 @@ typedef enum {
>>>   #define TCG_TARGET_HAS_sub2_i32         1
>>>   #define TCG_TARGET_HAS_mulu2_i32        0
>>>   #define TCG_TARGET_HAS_muls2_i32        0
>>> -#define TCG_TARGET_HAS_muluh_i32        (TCG_TARGET_REG_BITS == 32)
>>> -#define TCG_TARGET_HAS_mulsh_i32        (TCG_TARGET_REG_BITS == 32)
>>> +#define TCG_TARGET_HAS_muluh_i32        0
>>> +#define TCG_TARGET_HAS_mulsh_i32        0
>>
>> Should have we squashed the following with these changes?
> 
> Yes, mulsh_i32 is not reachable anymore.  At this point I'll just leave 
> this cleanup to conversion of mulsh to TCGOutOpBinary.

Sure, I was not planing to ask for another patch, but trying to see if
I was correctly understanding your conversion.

Thanks,

Phil.


  reply	other threads:[~2025-02-24  9:25 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-05 21:24 [PULL 00/42] tcg patch queue Richard Henderson
2023-05-05 21:24 ` [PULL 01/42] softfloat: Fix the incorrect computation in float32_exp2 Richard Henderson
2023-05-05 21:24 ` [PULL 02/42] target/avr: Finish conversion to tcg_gen_qemu_{ld,st}_* Richard Henderson
2023-05-05 21:24 ` [PULL 03/42] target/cris: Finish conversion to tcg_gen_qemu_{ld, st}_* Richard Henderson
2023-05-05 21:24 ` [PULL 04/42] target/Hexagon: " Richard Henderson
2023-05-05 21:24 ` [PULL 05/42] target/m68k: " Richard Henderson
2023-05-08 11:44   ` Laurent Vivier
2023-05-08 13:11     ` Richard Henderson
2023-05-05 21:24 ` [PULL 06/42] target/mips: " Richard Henderson
2023-05-05 21:24 ` [PULL 07/42] target/s390x: " Richard Henderson
2023-05-05 21:24 ` [PULL 08/42] target/sparc: " Richard Henderson
2023-05-05 21:24 ` [PULL 09/42] target/xtensa: " Richard Henderson
2023-05-05 21:24 ` [PULL 10/42] tcg: Remove compatability helpers for qemu ld/st Richard Henderson
2023-05-05 21:24 ` [PULL 11/42] target/alpha: Use MO_ALIGN for system UNALIGN() Richard Henderson
2023-05-05 21:24 ` [PULL 12/42] target/alpha: Use MO_ALIGN where required Richard Henderson
2023-05-05 21:24 ` [PULL 13/42] target/alpha: Remove TARGET_ALIGNED_ONLY Richard Henderson
2023-05-05 21:24 ` [PULL 14/42] target/hppa: Use MO_ALIGN for system UNALIGN() Richard Henderson
2023-05-05 21:24 ` [PULL 15/42] target/hppa: Remove TARGET_ALIGNED_ONLY Richard Henderson
2023-05-05 21:24 ` [PULL 16/42] target/sparc: Use MO_ALIGN where required Richard Henderson
2023-05-05 21:24 ` [PULL 17/42] target/sparc: Use cpu_ld*_code_mmu Richard Henderson
2023-05-05 21:24 ` [PULL 18/42] target/sparc: Remove TARGET_ALIGNED_ONLY Richard Henderson
2023-05-05 21:24 ` [PULL 19/42] tcg/i386: Rationalize args to tcg_out_qemu_{ld,st} Richard Henderson
2023-05-05 21:24 ` [PULL 20/42] tcg/i386: Generalize multi-part load overlap test Richard Henderson
2023-05-05 21:24 ` [PULL 21/42] tcg/i386: Introduce HostAddress Richard Henderson
2023-05-05 21:24 ` [PULL 22/42] tcg/i386: Drop r0+r1 local variables from tcg_out_tlb_load Richard Henderson
2023-05-05 21:24 ` [PULL 23/42] tcg/i386: Introduce tcg_out_testi Richard Henderson
2023-05-05 21:24 ` [PULL 24/42] tcg/aarch64: Rationalize args to tcg_out_qemu_{ld,st} Richard Henderson
2023-05-05 21:24 ` [PULL 25/42] tcg/aarch64: Introduce HostAddress Richard Henderson
2023-05-05 21:24 ` [PULL 26/42] tcg/arm: Rationalize args to tcg_out_qemu_{ld,st} Richard Henderson
2023-05-05 21:24 ` [PULL 27/42] tcg/arm: Introduce HostAddress Richard Henderson
2023-05-05 21:24 ` [PULL 28/42] tcg/loongarch64: Rationalize args to tcg_out_qemu_{ld, st} Richard Henderson
2023-05-05 21:24 ` [PULL 29/42] tcg/loongarch64: Introduce HostAddress Richard Henderson
2023-05-05 21:24 ` [PULL 30/42] tcg/mips: Rationalize args to tcg_out_qemu_{ld,st} Richard Henderson
2023-05-05 21:24 ` [PULL 31/42] tcg/ppc: " Richard Henderson
2023-05-05 21:24 ` [PULL 32/42] tcg/ppc: Introduce HostAddress Richard Henderson
2023-05-05 21:24 ` [PULL 33/42] tcg/riscv: Require TCG_TARGET_REG_BITS == 64 Richard Henderson
2025-02-20 23:27   ` Philippe Mathieu-Daudé
2025-02-22 18:17     ` Richard Henderson
2025-02-24  9:24       ` Philippe Mathieu-Daudé [this message]
2023-05-05 21:24 ` [PULL 34/42] tcg/riscv: Rationalize args to tcg_out_qemu_{ld,st} Richard Henderson
2023-05-05 21:24 ` [PULL 35/42] tcg/s390x: Pass TCGType " Richard Henderson
2023-05-05 21:24 ` [PULL 36/42] tcg/s390x: Introduce HostAddress Richard Henderson
2023-05-05 21:24 ` [PULL 37/42] tcg/sparc64: Drop is_64 test from tcg_out_qemu_ld data return Richard Henderson
2023-05-05 21:24 ` [PULL 38/42] tcg/sparc64: Pass TCGType to tcg_out_qemu_{ld,st} Richard Henderson
2023-05-05 21:24 ` [PULL 39/42] tcg: Move TCGLabelQemuLdst to tcg.c Richard Henderson
2023-05-05 21:24 ` [PULL 40/42] tcg: Replace REG_P with arg_loc_reg_p Richard Henderson
2023-05-05 21:24 ` [PULL 41/42] tcg: Introduce arg_slot_stk_ofs Richard Henderson
2023-05-05 21:24 ` [PULL 42/42] tcg: Widen helper_*_st[bw]_mmu val arguments Richard Henderson
2023-05-06  7:11 ` [PULL 00/42] tcg patch queue Richard Henderson

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=af650a49-8917-4d19-9cd4-e29a1538b86b@linaro.org \
    --to=philmd@linaro.org \
    --cc=dbarboza@ventanamicro.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-riscv@nongnu.org \
    --cc=richard.henderson@linaro.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: 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).