From: "Philippe Mathieu-Daudé" <f4bug@amsat.org>
To: Richard Henderson <richard.henderson@linaro.org>, qemu-devel@nongnu.org
Subject: Re: [PATCH v7 15/27] tcg/tci: Change encoding to uint32_t units
Date: Sat, 19 Jun 2021 19:48:14 +0200 [thread overview]
Message-ID: <ca5f8c8e-2e92-75c7-5e09-e82ada0f122b@amsat.org> (raw)
In-Reply-To: <20210601150106.12761-16-richard.henderson@linaro.org>
On 6/1/21 5:00 PM, Richard Henderson wrote:
> This removes all of the problems with unaligned accesses
> to the bytecode stream.
>
> With an 8-bit opcode at the bottom, we have 24 bits remaining,
> which are generally split into 6 4-bit slots. This fits well
> with the maximum length opcodes, e.g. INDEX_op_add2_i32, which
> have 6 register operands.
>
> We have, in previous patches, rearranged things such that there
> are no operations with a label which have more than one other
> operand. Which leaves us with a 20-bit field in which to encode
> a label, giving us a maximum TB size of 512k -- easily large.
>
> Change the INDEX_op_tci_movi_{i32,i64} opcodes to tci_mov[il].
> The former puts the immediate in the upper 20 bits of the insn,
> like we do for the label displacement. The later uses a label
> to reference an entry in the constant pool. Thus, in the worst
> case we still have a single memory reference for any constant,
> but now the constants are out-of-line of the bytecode and can
> be shared between different moves saving space.
>
> Change INDEX_op_call to use a label to reference a pair of
> pointers in the constant pool. This removes the only slightly
> dodgy link with the layout of struct TCGHelperInfo.
>
> The re-encode cannot be done in pieces.
>
> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
> ---
> include/tcg/tcg-opc.h | 4 +-
> tcg/tci/tcg-target.h | 3 +-
> tcg/tci.c | 541 +++++++++++++++------------------------
> tcg/tci/tcg-target.c.inc | 379 ++++++++++++---------------
> tcg/tci/README | 20 +-
> 5 files changed, 384 insertions(+), 563 deletions(-)
> @@ -1082,87 +961,69 @@ static const char *str_c(TCGCond c)
> /* Disassemble TCI bytecode. */
> int print_insn_tci(bfd_vma addr, disassemble_info *info)
> {
> case INDEX_op_setcond_i32:
> case INDEX_op_setcond_i64:
> - tci_args_rrrc(&tb_ptr, &r0, &r1, &r2, &c);
> + tci_args_rrrc(insn, &r0, &r1, &r2, &c);
> info->fprintf_func(info->stream, "%-12s %s, %s, %s, %s",
> op_name, str_r(r0), str_r(r1), str_r(r2), str_c(c));
> break;
>
> - case INDEX_op_tci_movi_i32:
> - tci_args_ri(&tb_ptr, &r0, &i1);
> - info->fprintf_func(info->stream, "%-12s %s, 0x%" TCG_PRIlx,
> + case INDEX_op_tci_movi:
> + tci_args_ri(insn, &r0, &i1);
> + info->fprintf_func(info->stream, "%-12s %s,0x%" TCG_PRIlx "",
Missing space in format: "%s, 0x%"
> op_name, str_r(r0), i1);
> break;
>
> -#if TCG_TARGET_REG_BITS == 64
> - case INDEX_op_tci_movi_i64:
> - tci_args_rI(&tb_ptr, &r0, &i1);
> - info->fprintf_func(info->stream, "%-12s %s, 0x%" TCG_PRIlx,
> - op_name, str_r(r0), i1);
> + case INDEX_op_tci_movl:
> + tci_args_rl(insn, tb_ptr, &r0, &ptr);
> + info->fprintf_func(info->stream, "%-12s %s, %p",
> + op_name, str_r(r0), ptr);
> break;
> -#endif
next prev parent reply other threads:[~2021-06-19 17:49 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-01 15:00 [PATCH v7 00/27] TCI fixes and cleanups Richard Henderson
2021-06-01 15:00 ` [PATCH v7 01/27] tcg: Combine dh_is_64bit and dh_is_signed to dh_typecode Richard Henderson
2021-06-01 15:00 ` [PATCH v7 02/27] tcg: Add tcg_call_flags Richard Henderson
2021-06-01 15:00 ` [PATCH v7 03/27] accel/tcg/plugin-gen: Drop inline markers Richard Henderson
2021-06-01 15:00 ` [PATCH v7 04/27] plugins: Drop tcg_flags from struct qemu_plugin_dyn_cb Richard Henderson
2021-06-02 9:22 ` Alex Bennée
2021-06-02 16:02 ` Richard Henderson
2021-06-01 15:00 ` [PATCH v7 05/27] accel/tcg: Add tcg call flags to plugins helpers Richard Henderson
2021-06-01 15:00 ` [PATCH v7 06/27] tcg: Store the TCGHelperInfo in the TCGOp for call Richard Henderson
2021-06-01 15:00 ` [PATCH v7 07/27] tcg: Add tcg_call_func Richard Henderson
2021-06-01 15:00 ` [PATCH v7 08/27] tcg: Build ffi data structures for helpers Richard Henderson
2021-06-01 15:00 ` [PATCH v7 09/27] tcg/tci: Improve tcg_target_call_clobber_regs Richard Henderson
2021-06-02 17:59 ` Philippe Mathieu-Daudé
2021-06-01 15:00 ` [PATCH v7 10/27] tcg/tci: Move call-return regs to end of tcg_target_reg_alloc_order Richard Henderson
2021-06-01 15:00 ` [PATCH v7 11/27] tcg/tci: Use ffi for calls Richard Henderson
2021-06-02 10:31 ` Alex Bennée
2021-06-01 15:00 ` [PATCH v7 12/27] tcg/tci: Reserve r13 for a temporary Richard Henderson
2021-06-19 15:50 ` Philippe Mathieu-Daudé
2021-06-01 15:00 ` [PATCH v7 13/27] tcg/tci: Emit setcond before brcond Richard Henderson
2021-06-19 15:50 ` Philippe Mathieu-Daudé
2021-06-01 15:00 ` [PATCH v7 14/27] tcg/tci: Remove tci_write_reg Richard Henderson
2021-06-01 15:00 ` [PATCH v7 15/27] tcg/tci: Change encoding to uint32_t units Richard Henderson
2021-06-12 10:21 ` Philippe Mathieu-Daudé
2021-06-12 15:40 ` Richard Henderson
2021-06-19 17:48 ` Philippe Mathieu-Daudé [this message]
2021-06-19 18:05 ` Richard Henderson
2021-06-01 15:00 ` [PATCH v7 16/27] tcg/tci: Implement goto_ptr Richard Henderson
2021-06-12 9:45 ` Philippe Mathieu-Daudé
2021-06-01 15:00 ` [PATCH v7 17/27] tcg/tci: Implement movcond Richard Henderson
2021-06-01 15:00 ` [PATCH v7 18/27] tcg/tci: Implement andc, orc, eqv, nand, nor Richard Henderson
2021-06-01 15:00 ` [PATCH v7 19/27] tcg/tci: Implement extract, sextract Richard Henderson
2021-06-01 15:00 ` [PATCH v7 20/27] tcg/tci: Implement clz, ctz, ctpop Richard Henderson
2021-06-01 15:01 ` [PATCH v7 21/27] tcg/tci: Implement mulu2, muls2 Richard Henderson
2021-06-01 15:01 ` [PATCH v7 22/27] tcg/tci: Implement add2, sub2 Richard Henderson
2021-06-12 9:36 ` Philippe Mathieu-Daudé
2021-06-01 15:01 ` [PATCH v7 23/27] tcg/tci: Split out tci_qemu_ld, tci_qemu_st Richard Henderson
2021-06-02 18:01 ` Philippe Mathieu-Daudé
2021-06-01 15:01 ` [PATCH v7 24/27] Revert "tcg/tci: Use exec/cpu_ldst.h interfaces" Richard Henderson
2021-06-12 9:24 ` Philippe Mathieu-Daudé
2021-06-01 15:01 ` [PATCH v7 25/27] tcg/tci: Remove the qemu_ld/st_type macros Richard Henderson
2021-06-02 18:03 ` Philippe Mathieu-Daudé
2021-06-01 15:01 ` [PATCH v7 26/27] tcg/tci: Use {set,clear}_helper_retaddr Richard Henderson
2021-06-12 9:24 ` Philippe Mathieu-Daudé
2021-06-01 15:01 ` [PATCH v7 27/27] tests/tcg: Increase timeout for TCI Richard Henderson
2021-06-02 9:26 ` Alex Bennée
2021-06-01 15:30 ` [PATCH v7 00/27] TCI fixes and cleanups no-reply
2021-06-19 17:42 ` Philippe Mathieu-Daudé
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=ca5f8c8e-2e92-75c7-5e09-e82ada0f122b@amsat.org \
--to=f4bug@amsat.org \
--cc=qemu-devel@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.