From: "Richard Earnshaw (lists)" <Richard.Earnshaw@arm.com>
To: Victor Do Nascimento <victor.donascimento@arm.com>,
binutils@sourceware.org
Cc: nickc@redhat.com
Subject: Re: [PATCH v2 1/2] aarch64: system register aliasing detection
Date: Tue, 3 Oct 2023 14:40:14 +0100 [thread overview]
Message-ID: <9882e14b-e59a-7868-d41c-36f4fae2fc76@arm.com> (raw)
In-Reply-To: <20231003105338.1812768-2-victor.donascimento@arm.com>
On 03/10/2023 11:51, Victor Do Nascimento via Binutils wrote:
> This patch adds a mechanism for system register name alias detection
> to register matching mechanisms.
>
> A new `F_REG_ALIAS' flag is added to the set of register flags and
> used to label which entries in aarch64_sys_regs[] correspond to
> aliases (and thus which CPENC values are non-unique in this array).
>
> Where this is used is, for example, in `aarch64_print_operand' where,
> in the case of system register decoding, the aarch64_sys_regs[] array
> is iterated through until a match in CPENC value is made and the first
> match accepted. If insufficient care is given in the ordering of
> system registers in this array, the alias is encountered before the
> "real" register and used incorrectly as the register name in the
> disassembled output.
>
> With this flag and the new `aarch64_sys_reg_alias_p' test, search
> candidates corresponding to aliases can be conveniently skipped over.
>
> One concrete example of where this is useful is with the
> `trcextinselr0' system register. It was initially placed in the
> system register list before `trcextinselr', in contrast to a more
> natural alphabetical order.
>
> include/ChangeLog:
> * opcode/aarch64.h: add `aarch64_sys_reg_alias_p' prototype.
>
> opcodes/ChangeLog:
> * aarch64-opc.c (aarch64_sys_reg_alias_p): New.
> (aarch64_print_operand): add aarch64_sys_reg_alias_p check.
> (aarch64_sys_regs): Add F_REG_ALIAS flag to "trcextinselr"
> entry.
> * aarch64-opc.h (F_REG_ALIAS): New.
OK
Reviewed-by: Richard.Earnshaw@arm.com
R.
> ---
> include/opcode/aarch64.h | 1 +
> opcodes/aarch64-opc.c | 9 ++++++++-
> opcodes/aarch64-opc.h | 3 +++
> 3 files changed, 12 insertions(+), 1 deletion(-)
>
> diff --git a/include/opcode/aarch64.h b/include/opcode/aarch64.h
> index ab42acac8c2..d4ec60c1354 100644
> --- a/include/opcode/aarch64.h
> +++ b/include/opcode/aarch64.h
> @@ -1240,6 +1240,7 @@ typedef struct
> extern const aarch64_sys_reg aarch64_sys_regs [];
> extern const aarch64_sys_reg aarch64_pstatefields [];
> extern bool aarch64_sys_reg_deprecated_p (const uint32_t);
> +extern bool aarch64_sys_reg_alias_p (const uint32_t);
> extern bool aarch64_pstatefield_supported_p (const aarch64_feature_set,
> const aarch64_sys_reg *);
>
> diff --git a/opcodes/aarch64-opc.c b/opcodes/aarch64-opc.c
> index df1b5113641..c901d3d587b 100644
> --- a/opcodes/aarch64-opc.c
> +++ b/opcodes/aarch64-opc.c
> @@ -4507,6 +4507,7 @@ aarch64_print_operand (char *buf, size_t size, bfd_vma pc,
> partial match that was found. */
> if (aarch64_sys_regs[i].value == opnd->sysreg.value
> && ! aarch64_sys_reg_deprecated_p (aarch64_sys_regs[i].flags)
> + && ! aarch64_sys_reg_alias_p (aarch64_sys_regs[i].flags)
> && (name == NULL || exact_match))
> {
> name = aarch64_sys_regs[i].name;
> @@ -5320,7 +5321,7 @@ const aarch64_sys_reg aarch64_sys_regs [] =
> SR_CORE ("trceventctl0r", CPENC (2,1,C0,C8,0), 0),
> SR_CORE ("trceventctl1r", CPENC (2,1,C0,C9,0), 0),
> SR_CORE ("trcextinselr0", CPENC (2,1,C0,C8,4), 0),
> - SR_CORE ("trcextinselr", CPENC (2,1,C0,C8,4), 0),
> + SR_CORE ("trcextinselr", CPENC (2,1,C0,C8,4), F_REG_ALIAS),
> SR_CORE ("trcextinselr1", CPENC (2,1,C0,C9,4), 0),
> SR_CORE ("trcextinselr2", CPENC (2,1,C0,C10,4), 0),
> SR_CORE ("trcextinselr3", CPENC (2,1,C0,C11,4), 0),
> @@ -5742,6 +5743,12 @@ aarch64_sys_reg_deprecated_p (const uint32_t reg_flags)
> return (reg_flags & F_DEPRECATED) != 0;
> }
>
> +bool
> +aarch64_sys_reg_alias_p (const uint32_t reg_flags)
> +{
> + return (reg_flags & F_REG_ALIAS) != 0;
> +}
> +
> /* The CPENC below is fairly misleading, the fields
> here are not in CPENC form. They are in op2op1 form. The fields are encoded
> by ins_pstatefield, which just shifts the value by the width of the fields
> diff --git a/opcodes/aarch64-opc.h b/opcodes/aarch64-opc.h
> index 32e4da2bbb6..fe1f882c20e 100644
> --- a/opcodes/aarch64-opc.h
> +++ b/opcodes/aarch64-opc.h
> @@ -292,6 +292,9 @@ verify_constraints (const struct aarch64_inst *, const aarch64_insn, bfd_vma,
> #undef F_REG_IN_CRM
> #define F_REG_IN_CRM (1 << 5) /* Register extra encoding in CRm. */
>
> +#undef F_REG_ALIAS
> +#define F_REG_ALIAS (1 << 6) /* Register name aliases another. */
> +
> /* PSTATE field name for the MSR instruction this is encoded in "op1:op2:CRm".
> Part of CRm can be used to encode <pstatefield>. E.g. CRm[3:1] for SME.
> In order to set/get full PSTATE field name use flag F_REG_IN_CRM and below
next prev parent reply other threads:[~2023-10-03 13:40 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-03 10:51 [PATCH v2 0/2] aarch64: standardize system register representation Victor Do Nascimento
2023-10-03 10:51 ` [PATCH v2 1/2] aarch64: system register aliasing detection Victor Do Nascimento
2023-10-03 13:40 ` Richard Earnshaw (lists) [this message]
2023-10-03 10:51 ` [PATCH v2 2/2] aarch64: Refactor system register data Victor Do Nascimento
2023-10-03 13:46 ` Richard Earnshaw (lists)
2023-10-04 10:15 ` Victor Do Nascimento
2023-10-04 10:22 ` Victor Do Nascimento
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=9882e14b-e59a-7868-d41c-36f4fae2fc76@arm.com \
--to=richard.earnshaw@arm.com \
--cc=binutils@sourceware.org \
--cc=nickc@redhat.com \
--cc=victor.donascimento@arm.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).