From: Will Newton <will.newton@linaro.org>
To: Omair Javaid <omair.javaid@linaro.org>
Cc: "gdb-patches@sourceware.org" <gdb-patches@sourceware.org>,
Patch Tracking <patches@linaro.org>
Subject: Re: [PATCH v3 5/6] Implement support for recording vector data transfer instructions
Date: Wed, 13 Aug 2014 14:10:00 -0000 [thread overview]
Message-ID: <CANu=DmiADC1L8EAgJFq+2PPJQhHgFVriQXJPOBEi+jibSfHmaw@mail.gmail.com> (raw)
In-Reply-To: <1407935535-27978-6-git-send-email-omair.javaid@linaro.org>
On 13 August 2014 14:12, Omair Javaid <omair.javaid@linaro.org> wrote:
> gdb:
>
> 2014-08-13 Omair Javaid <omair.javaid@linaro.org>
>
> * arm-tdep.c (arm_record_vdata_transfer_insn): Added record handler for
> vector data transfer instructions.
> (arm_record_coproc_data_proc): Updated.
>
> ---
> gdb/arm-tdep.c | 98 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
> 1 file changed, 97 insertions(+), 1 deletion(-)
Looks ok to me.
> diff --git a/gdb/arm-tdep.c b/gdb/arm-tdep.c
> index 315b5b0..7f651bc 100644
> --- a/gdb/arm-tdep.c
> +++ b/gdb/arm-tdep.c
> @@ -11990,6 +11990,102 @@ arm_record_unsupported_insn (insn_decode_record *arm_insn_r)
> return -1;
> }
>
> +/* Record handler for vector data transfer instructions. */
> +
> +static int
> +arm_record_vdata_transfer_insn (insn_decode_record *arm_insn_r)
> +{
> + uint32_t bits_a, bit_c, bit_l, reg_t, reg_v;
> + uint32_t record_buf[4];
> +
> + const int num_regs = gdbarch_num_regs (arm_insn_r->gdbarch);
> + reg_t = bits (arm_insn_r->arm_insn, 12, 15);
> + reg_v = bits (arm_insn_r->arm_insn, 21, 23);
> + bits_a = bits (arm_insn_r->arm_insn, 21, 23);
> + bit_l = bit (arm_insn_r->arm_insn, 20);
> + bit_c = bit (arm_insn_r->arm_insn, 8);
> +
> + /* Handle VMOV instruction. */
> + if (bit_l && bit_c)
> + {
> + record_buf[0] = reg_t;
> + arm_insn_r->reg_rec_count = 1;
> + }
> + else if (bit_l && !bit_c)
> + {
> + /* Handle VMOV instruction. */
> + if (bits_a == 0x00)
> + {
> + if (bit (arm_insn_r->arm_insn, 20))
> + record_buf[0] = reg_t;
> + else
> + record_buf[0] = num_regs + (bit (arm_insn_r->arm_insn, 7) |
> + (reg_v << 1));
> +
> + arm_insn_r->reg_rec_count = 1;
> + }
> + /* Handle VMRS instruction. */
> + else if (bits_a == 0x07)
> + {
> + if (reg_t == 15)
> + reg_t = ARM_PS_REGNUM;
> +
> + record_buf[0] = reg_t;
> + arm_insn_r->reg_rec_count = 1;
> + }
> + }
> + else if (!bit_l && !bit_c)
> + {
> + /* Handle VMOV instruction. */
> + if (bits_a == 0x00)
> + {
> + if (bit (arm_insn_r->arm_insn, 20))
> + record_buf[0] = reg_t;
> + else
> + record_buf[0] = num_regs + (bit (arm_insn_r->arm_insn, 7) |
> + (reg_v << 1));
> +
> + arm_insn_r->reg_rec_count = 1;
> + }
> + /* Handle VMSR instruction. */
> + else if (bits_a == 0x07)
> + {
> + record_buf[0] = ARM_FPSCR_REGNUM;
> + arm_insn_r->reg_rec_count = 1;
> + }
> + }
> + else if (!bit_l && bit_c)
> + {
> + /* Handle VMOV instruction. */
> + if (!(bits_a & 0x04))
> + {
> + record_buf[0] = (reg_v | (bit (arm_insn_r->arm_insn, 7) << 4))
> + + ARM_D0_REGNUM;
> + arm_insn_r->reg_rec_count = 1;
> + }
> + /* Handle VDUP instruction. */
> + else
> + {
> + if (bit (arm_insn_r->arm_insn, 21))
> + {
> + reg_v = reg_v | (bit (arm_insn_r->arm_insn, 7) << 4);
> + record_buf[0] = reg_v + ARM_D0_REGNUM;
> + record_buf[1] = reg_v + ARM_D0_REGNUM + 1;
> + arm_insn_r->reg_rec_count = 2;
> + }
> + else
> + {
> + reg_v = reg_v | (bit (arm_insn_r->arm_insn, 7) << 4);
> + record_buf[0] = reg_v + ARM_D0_REGNUM;
> + arm_insn_r->reg_rec_count = 1;
> + }
> + }
> + }
> +
> + REG_ALLOC (arm_insn_r->arm_regs, arm_insn_r->reg_rec_count, record_buf);
> + return 0;
> +}
> +
> /* Record handler for extension register load/store instructions. */
>
> static int
> @@ -12467,7 +12563,7 @@ arm_record_coproc_data_proc (insn_decode_record *arm_insn_r)
>
> /* Advanced SIMD, VFP instructions. */
> if (!op1_sbit && op)
> - return arm_record_unsupported_insn (arm_insn_r);
> + return arm_record_vdata_transfer_insn (arm_insn_r);
> }
> else
> {
> --
> 1.9.1
>
--
Will Newton
Toolchain Working Group, Linaro
next prev parent reply other threads:[~2014-08-13 14:10 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-08-13 13:13 [PATCH v3 0/6] ARM process record/replay improvements Omair Javaid
2014-08-13 13:13 ` [PATCH v3 6/6] Fix reverse-step and reverse-next over undebuggable solib code Omair Javaid
2014-08-27 9:09 ` Omair Javaid
2014-08-27 10:34 ` Pedro Alves
2014-08-13 13:13 ` [PATCH v3 1/6] Implements support for recording arm/thumb mode coprocessor instructions Omair Javaid
2014-08-13 14:10 ` Will Newton
2014-08-27 9:07 ` Omair Javaid
2014-08-27 10:05 ` Pedro Alves
2014-08-28 9:50 ` [PATCH] Implement " Omair Javaid
2014-09-02 14:51 ` Will Newton
2014-08-13 13:13 ` [PATCH v3 4/6] Implement support for recording extension register ld/st insn Omair Javaid
2014-08-13 14:10 ` Will Newton
2014-08-27 9:21 ` Omair Javaid
2014-08-27 10:17 ` Pedro Alves
2014-08-28 12:56 ` Omair Javaid
2014-09-02 14:59 ` Will Newton
2014-09-10 11:29 ` Omair Javaid
2014-09-10 11:55 ` Will Newton
2014-08-13 13:13 ` [PATCH v3 2/6] Implements support for recording thumb2 ASIMD struct " Omair Javaid
2014-08-13 14:10 ` Will Newton
2014-08-27 9:08 ` Omair Javaid
2014-08-27 10:09 ` Pedro Alves
2014-08-28 10:56 ` [PATCH v3 2/6] Implement support for recording thumb2 ASIMD struct ld/st insns Omair Javaid
2014-09-02 14:55 ` Will Newton
2014-09-10 11:16 ` Omair Javaid
2014-09-10 11:54 ` Will Newton
2014-08-13 13:13 ` [PATCH v3 3/6] Implement support for recording VFP data processing instructions Omair Javaid
2014-08-13 14:10 ` Will Newton
2014-08-27 9:10 ` Omair Javaid
2014-08-27 10:11 ` Pedro Alves
2014-08-28 11:06 ` Omair Javaid
2014-08-13 13:13 ` [PATCH v3 5/6] Implement support for recording vector data transfer instructions Omair Javaid
2014-08-13 14:10 ` Will Newton [this message]
2014-08-27 9:09 ` Omair Javaid
2014-08-27 10:19 ` Pedro Alves
2014-08-28 13:07 ` Omair Javaid
2014-08-27 9:05 ` [PATCH v3 0/6] ARM process record/replay improvements Omair Javaid
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='CANu=DmiADC1L8EAgJFq+2PPJQhHgFVriQXJPOBEi+jibSfHmaw@mail.gmail.com' \
--to=will.newton@linaro.org \
--cc=gdb-patches@sourceware.org \
--cc=omair.javaid@linaro.org \
--cc=patches@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