From: Nelson Chu <nelson@rivosinc.com>
To: Jan Beulich <jbeulich@suse.com>
Cc: Binutils <binutils@sourceware.org>,
Palmer Dabbelt <palmer@dabbelt.com>,
Andrew Waterman <andrew@sifive.com>,
Jim Wilson <jim.wilson.gcc@gmail.com>
Subject: Re: [PATCH v2] RISC-V: make C-extension JAL available again for (32-bit) assembly
Date: Tue, 31 Jan 2023 09:11:40 +0800 [thread overview]
Message-ID: <CAPpQWtDiTspFK6L-=UMgEJt7aDRKjiVu_ODdOQRb=R7Ke9SwsQ@mail.gmail.com> (raw)
In-Reply-To: <b61869e7-8fd3-2bec-da2b-046066edc3f3@suse.com>
Thanks for helping to follow these. Looks reasonable, so please commit.
Nelson
On Mon, Jan 30, 2023 at 10:07 PM Jan Beulich <jbeulich@suse.com> wrote:
>
> Along with the normal JAL alias, the C-extension one should have been
> moved as well by 839189bc932e ("RISC-V: re-arrange opcode table for
> consistent alias handling"), for the assembler to actually be able to
> use it where/when possible.
>
> Since neither this nor any other compressed branch insn was being tested
> so far, take the opportunity and introduce a new testcase covering those.
> ---
> Pretty clearly this should also go on the 2.40 branch.
> ---
> v2: Avoid re-introducing PR/gas 29940.
>
> --- a/gas/config/tc-riscv.c
> +++ b/gas/config/tc-riscv.c
> @@ -2764,6 +2764,8 @@ riscv_ip (char *str, struct riscv_cl_ins
> case 'p':
> goto branch;
> case 'a':
> + if (oparg == insn->args + 1)
> + goto jump_check_gpr;
> goto jump;
> case 'S': /* Floating-point RS1 x8-x15. */
> if (!reg_lookup (&asarg, RCLASS_FPR, ®no)
> @@ -3273,6 +3275,7 @@ riscv_ip (char *str, struct riscv_cl_ins
> but the 2nd (with 2 operands) might. */
> if (oparg == insn->args)
> {
> + jump_check_gpr:
> asargStart = asarg;
> if (reg_lookup (&asarg, RCLASS_GPR, NULL)
> && (*asarg == ',' || (ISSPACE (*asarg) && asarg[1] == ',')))
> --- a/opcodes/riscv-opc.c
> +++ b/opcodes/riscv-opc.c
> @@ -340,9 +340,9 @@ const struct riscv_opcode riscv_opcodes[
> {"jalr", 0, INSN_CLASS_I, "d,s,j", MATCH_JALR, MASK_JALR, match_opcode, INSN_JSR },
> {"j", 0, INSN_CLASS_C, "Ca", MATCH_C_J, MASK_C_J, match_opcode, INSN_ALIAS|INSN_BRANCH },
> {"j", 0, INSN_CLASS_I, "a", MATCH_JAL, MASK_JAL|MASK_RD, match_opcode, INSN_ALIAS|INSN_BRANCH },
> +{"jal", 32, INSN_CLASS_C, "Ca", MATCH_C_JAL, MASK_C_JAL, match_opcode, INSN_ALIAS|INSN_JSR },
> {"jal", 0, INSN_CLASS_I, "a", MATCH_JAL|(X_RA << OP_SH_RD), MASK_JAL|MASK_RD, match_opcode, INSN_ALIAS|INSN_JSR },
> {"jal", 0, INSN_CLASS_I, "d,a", MATCH_JAL, MASK_JAL, match_opcode, INSN_JSR },
> -{"jal", 32, INSN_CLASS_C, "Ca", MATCH_C_JAL, MASK_C_JAL, match_opcode, INSN_ALIAS|INSN_JSR },
> {"call", 0, INSN_CLASS_I, "d,c", (X_T1 << OP_SH_RS1), (int) M_CALL, match_never, INSN_MACRO },
> {"call", 0, INSN_CLASS_I, "c", (X_RA << OP_SH_RS1)|(X_RA << OP_SH_RD), (int) M_CALL, match_never, INSN_MACRO },
> {"tail", 0, INSN_CLASS_I, "c", (X_T1 << OP_SH_RS1), (int) M_CALL, match_never, INSN_MACRO },
> --- /dev/null
> +++ b/gas/testsuite/gas/riscv/c-branch-na.d
> @@ -0,0 +1,20 @@
> +#as: -march=rv32ic
> +#source: c-branch.s
> +#objdump: -drw -Mno-aliases
> +
> +.*:[ ]+file format .*
> +
> +
> +Disassembly of section .text:
> +
> +0+ <target>:
> +[ ]+[0-9a-f]+:[ ]+c001[ ]+c\.beqz[ ]+s0,0 <target>[ ]+0: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+dcfd[ ]+c\.beqz[ ]+s1,0 <target>[ ]+2: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+fc75[ ]+c\.bnez[ ]+s0,0 <target>[ ]+4: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+fced[ ]+c\.bnez[ ]+s1,0 <target>[ ]+6: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+bfe5[ ]+c\.j[ ]+0 <target>[ ]+8: R_RISCV_RVC_JUMP .*
> +[ ]+[0-9a-f]+:[ ]+3fdd[ ]+c\.jal[ ]+0 <target>[ ]+a: R_RISCV_RVC_JUMP .*
> +[ ]+[0-9a-f]+:[ ]+9302[ ]+c\.jalr[ ]+t1
> +[ ]+[0-9a-f]+:[ ]+8382[ ]+c\.jr[ ]+t2
> +[ ]+[0-9a-f]+:[ ]+8082[ ]+c\.jr[ ]+ra
> +#...
> --- /dev/null
> +++ b/gas/testsuite/gas/riscv/c-branch.d
> @@ -0,0 +1,19 @@
> +#as: -march=rv64ic
> +#objdump: -drw
> +
> +.*:[ ]+file format .*
> +
> +
> +Disassembly of section .text:
> +
> +0+ <target>:
> +[ ]+[0-9a-f]+:[ ]+c001[ ]+beqz[ ]+s0,0 <target>[ ]+0: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+dcfd[ ]+beqz[ ]+s1,0 <target>[ ]+2: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+fc75[ ]+bnez[ ]+s0,0 <target>[ ]+4: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+fced[ ]+bnez[ ]+s1,0 <target>[ ]+6: R_RISCV_RVC_BRANCH .*
> +[ ]+[0-9a-f]+:[ ]+bfe5[ ]+j[ ]+0 <target>[ ]+8: R_RISCV_RVC_JUMP .*
> +[ ]+[0-9a-f]+:[ ]+ff7ff0ef[ ]+jal[ ]+0 <target>[ ]+a: R_RISCV_JAL .*
> +[ ]+[0-9a-f]+:[ ]+9302[ ]+jalr[ ]+t1
> +[ ]+[0-9a-f]+:[ ]+8382[ ]+jr[ ]+t2
> +[ ]+[0-9a-f]+:[ ]+8082[ ]+ret
> +#...
> --- /dev/null
> +++ b/gas/testsuite/gas/riscv/c-branch.s
> @@ -0,0 +1,11 @@
> + .text
> +target:
> + beq x8, x0, target
> + beqz x9, target
> + bne x8, x0, target
> + bnez x9, target
> + j target
> + jal target
> + jalr x6
> + jr x7
> + ret
next prev parent reply other threads:[~2023-01-31 1:11 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-30 14:07 Jan Beulich
2023-01-31 1:11 ` Nelson Chu [this message]
2023-01-31 8:52 ` Jan Beulich
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='CAPpQWtDiTspFK6L-=UMgEJt7aDRKjiVu_ODdOQRb=R7Ke9SwsQ@mail.gmail.com' \
--to=nelson@rivosinc.com \
--cc=andrew@sifive.com \
--cc=binutils@sourceware.org \
--cc=jbeulich@suse.com \
--cc=jim.wilson.gcc@gmail.com \
--cc=palmer@dabbelt.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 read-only IMAP folder(s) and NNTP newsgroup(s).