* [PATCH][SH] info registers
@ 2005-11-16 16:48 Andrew STUBBS
2005-11-16 16:54 ` Mark Kettenis
0 siblings, 1 reply; 5+ messages in thread
From: Andrew STUBBS @ 2005-11-16 16:48 UTC (permalink / raw)
To: gdb-patches
[-- Attachment #1: Type: text/plain, Size: 5913 bytes --]
Hi,
sh-tdep.c contains quite a lot of code to implement the 'info registers'
command (and friends).
It has a number of issues:
- The double and vector registers are only in hex.
- Invalid registers are still printed for some types (not that GDB
knows which registers are invalid - that's my next patch).
- The double registers are incorrectly named dr0..dr7, and the vector
registers incorrectly named fv0..fv3. They should be dr0,dr2,dr4... and
fv0,fv4,...
I have discovered that the default code (when the architecture may
override) does everything the SH code does right and yet does not do
everything it does wrong. Specifically all the above problems are
addressed.
I therefore propose removing all this garbage entirely.
Before
(gdb) info all-registers
r0 0x0 0
r1 0x0 0
r2 0x0 0
r3 0x0 0
r4 0x0 0
r5 0x0 0
r6 0x0 0
r7 0x0 0
r8 0x0 0
r9 0x0 0
r10 0x0 0
r11 0x0 0
r12 0x0 0
r13 0x0 0
r14 0x0 0
r15 0x0 0
pc 0x0 0
pr 0x0 0
gbr 0x0 0
vbr 0x0 0
mach 0x0 0
macl 0x0 0
sr 0x0 0
fpul 0 (raw 0x00000000)
fpscr 0x0 0
fr0 0 (raw 0x00000000)
fr1 0 (raw 0x00000000)
fr2 0 (raw 0x00000000)
fr3 0 (raw 0x00000000)
fr4 0 (raw 0x00000000)
fr5 0 (raw 0x00000000)
fr6 0 (raw 0x00000000)
fr7 0 (raw 0x00000000)
fr8 0 (raw 0x00000000)
fr9 0 (raw 0x00000000)
fr10 0 (raw 0x00000000)
fr11 0 (raw 0x00000000)
fr12 0 (raw 0x00000000)
fr13 0 (raw 0x00000000)
fr14 0 (raw 0x00000000)
fr15 0 (raw 0x00000000)
ssr 0x0 0
spc 0x0 0
r0b0 0x0 0
r1b0 0x0 0
r2b0 0x0 0
r3b0 0x0 0
r4b0 0x0 0
r5b0 0x0 0
r6b0 0x0 0
r7b0 0x0 0
r0b1 0x0 0
r1b1 0x0 0
r2b1 0x0 0
r3b1 0x0 0
r4b1 0x0 0
r5b1 0x0 0
r6b1 0x0 0
r7b1 0x0 0
bank 0
dr0 0x0000000000000000
dr1 0x0000000000000000
dr2 0x0000000000000000
dr3 0x0000000000000000
dr4 0x0000000000000000
dr5 0x0000000000000000
dr6 0x0000000000000000
dr7 0x0000000000000000
fv0 0x00000000 0x00000000 0x00000000 0x00000000
fv1 0x00000000 0x00000000 0x00000000 0x00000000
fv2 0x00000000 0x00000000 0x00000000 0x00000000
fv3 0x00000000 0x00000000 0x00000000 0x00000000
After
(gdb) info all-registers
r0 0x0 0
r1 0x0 0
r2 0x0 0
r3 0x0 0
r4 0x0 0
r5 0x0 0
r6 0x0 0
r7 0x0 0
r8 0x0 0
r9 0x0 0
r10 0x0 0
r11 0x0 0
r12 0x0 0
r13 0x0 0
r14 0x0 0
r15 0x0 0
pc 0x0 0
pr 0x0 0
gbr 0x0 0
vbr 0x0 0
mach 0x0 0
macl 0x0 0
sr 0x0 0
fpul 0 (raw 0x00000000)
fpscr 0x0 0
fr0 0 (raw 0x00000000)
fr1 0 (raw 0x00000000)
fr2 0 (raw 0x00000000)
fr3 0 (raw 0x00000000)
fr4 0 (raw 0x00000000)
fr5 0 (raw 0x00000000)
fr6 0 (raw 0x00000000)
fr7 0 (raw 0x00000000)
fr8 0 (raw 0x00000000)
fr9 0 (raw 0x00000000)
fr10 0 (raw 0x00000000)
fr11 0 (raw 0x00000000)
fr12 0 (raw 0x00000000)
fr13 0 (raw 0x00000000)
fr14 0 (raw 0x00000000)
fr15 0 (raw 0x00000000)
ssr 0x0 0
spc 0x0 0
r0b0 0x0 0
r1b0 0x0 0
r2b0 0x0 0
r3b0 0x0 0
r4b0 0x0 0
r5b0 0x0 0
r6b0 0x0 0
r7b0 0x0 0
r0b1 0x0 0
r1b1 0x0 0
r2b1 0x0 0
r3b1 0x0 0
r4b1 0x0 0
r5b1 0x0 0
r6b1 0x0 0
r7b1 0x0 0
dr0 0 (raw 0x0000000000000000)
dr2 0 (raw 0x0000000000000000)
dr4 0 (raw 0x0000000000000000)
dr6 0 (raw 0x0000000000000000)
dr8 0 (raw 0x0000000000000000)
dr10 0 (raw 0x0000000000000000)
dr12 0 (raw 0x0000000000000000)
dr14 0 (raw 0x0000000000000000)
fv0 {0x0, 0x0, 0x0, 0x0} {0, 0, 0, 0}
fv4 {0x0, 0x0, 0x0, 0x0} {0, 0, 0, 0}
fv8 {0x0, 0x0, 0x0, 0x0} {0, 0, 0, 0}
fv12 {0x0, 0x0, 0x0, 0x0} {0, 0, 0, 0}
Thanks
Andrew Stubbs
[-- Attachment #2: info_register.patch --]
[-- Type: text/plain, Size: 6901 bytes --]
2005-11-16 Andrew Stubbs <andrew.stubbs@st.com>
* sh-tdep.c (do_fv_register_info, do_dr_register_info): Delete function.
(do_bank_register_info, do_bank_register_info): Likewise.
(sh_print_pseudo_register, sh_do_fp_register): Likewise.
(sh_do_register, sh_print_register, sh_print_registers_info): Likewise.
(sh_gdbarch_init): Remove call to set_gdbarch_print_registers_info().
Index: src/gdb/sh-tdep.c
===================================================================
--- src.orig/gdb/sh-tdep.c 2005-11-04 14:20:18.000000000 +0000
+++ src/gdb/sh-tdep.c 2005-11-04 16:57:00.000000000 +0000
@@ -2071,184 +2071,6 @@ sh_pseudo_register_write (struct gdbarch
}
}
-/* Floating point vector of 4 float registers. */
-static void
-do_fv_register_info (struct gdbarch *gdbarch, struct ui_file *file,
- int fv_regnum)
-{
- int first_fp_reg_num = fv_reg_base_num (fv_regnum);
- fprintf_filtered (file, "fv%d\t0x%08x\t0x%08x\t0x%08x\t0x%08x\n",
- fv_regnum - FV0_REGNUM,
- (int) read_register (first_fp_reg_num),
- (int) read_register (first_fp_reg_num + 1),
- (int) read_register (first_fp_reg_num + 2),
- (int) read_register (first_fp_reg_num + 3));
-}
-
-/* Double precision registers. */
-static void
-do_dr_register_info (struct gdbarch *gdbarch, struct ui_file *file,
- int dr_regnum)
-{
- int first_fp_reg_num = dr_reg_base_num (dr_regnum);
-
- fprintf_filtered (file, "dr%d\t0x%08x%08x\n",
- dr_regnum - DR0_REGNUM,
- (int) read_register (first_fp_reg_num),
- (int) read_register (first_fp_reg_num + 1));
-}
-static void
-do_bank_register_info (struct gdbarch *gdbarch, struct ui_file *file)
-{
- fprintf_filtered (file, "bank %d\n",
- (int) read_register (BANK_REGNUM));
-}
-
-static void
-sh_print_pseudo_register (struct gdbarch *gdbarch, struct ui_file *file,
- int regnum)
-{
- if (regnum < NUM_REGS || regnum >= NUM_REGS + NUM_PSEUDO_REGS)
- internal_error (__FILE__, __LINE__,
- _("Invalid pseudo register number %d\n"), regnum);
- else if (regnum == PSEUDO_BANK_REGNUM)
- do_bank_register_info (gdbarch, file);
- else if (regnum >= DR0_REGNUM && regnum <= DR_LAST_REGNUM)
- do_dr_register_info (gdbarch, file, regnum);
- else if (regnum >= FV0_REGNUM && regnum <= FV_LAST_REGNUM)
- do_fv_register_info (gdbarch, file, regnum);
-}
-
-static void
-sh_do_fp_register (struct gdbarch *gdbarch, struct ui_file *file, int regnum)
-{ /* do values for FP (float) regs */
- char *raw_buffer;
- double flt; /* double extracted from raw hex data */
- int inv;
- int j;
-
- /* Allocate space for the float. */
- raw_buffer = (char *) alloca (register_size (gdbarch, FP0_REGNUM));
-
- /* Get the data in raw format. */
- if (!frame_register_read (get_selected_frame (NULL), regnum, raw_buffer))
- error (_("can't read register %d (%s)"), regnum, REGISTER_NAME (regnum));
-
- /* Get the register as a number */
- flt = unpack_double (builtin_type_float, raw_buffer, &inv);
-
- /* Print the name and some spaces. */
- fputs_filtered (REGISTER_NAME (regnum), file);
- print_spaces_filtered (15 - strlen (REGISTER_NAME (regnum)), file);
-
- /* Print the value. */
- if (inv)
- fprintf_filtered (file, "<invalid float>");
- else
- fprintf_filtered (file, "%-10.9g", flt);
-
- /* Print the fp register as hex. */
- fprintf_filtered (file, "\t(raw 0x");
- for (j = 0; j < register_size (gdbarch, regnum); j++)
- {
- int idx = (TARGET_BYTE_ORDER == BFD_ENDIAN_BIG
- ? j
- : register_size (gdbarch, regnum) - 1 - j);
- fprintf_filtered (file, "%02x", (unsigned char) raw_buffer[idx]);
- }
- fprintf_filtered (file, ")");
- fprintf_filtered (file, "\n");
-}
-
-static void
-sh_do_register (struct gdbarch *gdbarch, struct ui_file *file, int regnum)
-{
- char raw_buffer[MAX_REGISTER_SIZE];
-
- fputs_filtered (REGISTER_NAME (regnum), file);
- print_spaces_filtered (15 - strlen (REGISTER_NAME (regnum)), file);
-
- /* Get the data in raw format. */
- if (!frame_register_read (get_selected_frame (NULL), regnum, raw_buffer))
- fprintf_filtered (file, "*value not available*\n");
-
- val_print (gdbarch_register_type (gdbarch, regnum), raw_buffer, 0, 0,
- file, 'x', 1, 0, Val_pretty_default);
- fprintf_filtered (file, "\t");
- val_print (gdbarch_register_type (gdbarch, regnum), raw_buffer, 0, 0,
- file, 0, 1, 0, Val_pretty_default);
- fprintf_filtered (file, "\n");
-}
-
-static void
-sh_print_register (struct gdbarch *gdbarch, struct ui_file *file, int regnum)
-{
- if (regnum < 0 || regnum >= NUM_REGS + NUM_PSEUDO_REGS)
- internal_error (__FILE__, __LINE__,
- _("Invalid register number %d\n"), regnum);
-
- else if (regnum >= 0 && regnum < NUM_REGS)
- {
- if (TYPE_CODE (gdbarch_register_type (gdbarch, regnum)) ==
- TYPE_CODE_FLT)
- sh_do_fp_register (gdbarch, file, regnum); /* FP regs */
- else
- sh_do_register (gdbarch, file, regnum); /* All other regs */
- }
-
- else if (regnum < NUM_REGS + NUM_PSEUDO_REGS)
- {
- sh_print_pseudo_register (gdbarch, file, regnum);
- }
-}
-
-static void
-sh_print_registers_info (struct gdbarch *gdbarch, struct ui_file *file,
- struct frame_info *frame, int regnum, int fpregs)
-{
- if (regnum != -1) /* do one specified register */
- {
- if (*(REGISTER_NAME (regnum)) == '\0')
- error (_("Not a valid register for the current processor type"));
-
- sh_print_register (gdbarch, file, regnum);
- }
- else
- /* do all (or most) registers */
- {
- for (regnum = 0; regnum < NUM_REGS; ++regnum)
- {
- /* If the register name is empty, it is undefined for this
- processor, so don't display anything. */
- if (REGISTER_NAME (regnum) == NULL
- || *(REGISTER_NAME (regnum)) == '\0')
- continue;
-
- if (TYPE_CODE (gdbarch_register_type (gdbarch, regnum)) ==
- TYPE_CODE_FLT)
- {
- /* true for "INFO ALL-REGISTERS" command */
- if (fpregs)
- sh_do_fp_register (gdbarch, file, regnum); /* FP regs */
- }
- else
- sh_do_register (gdbarch, file, regnum); /* All other regs */
- }
-
- if (regnum == PSEUDO_BANK_REGNUM
- && REGISTER_NAME (regnum)
- && *REGISTER_NAME (regnum))
- sh_print_pseudo_register (gdbarch, file, regnum++);
-
- if (fpregs)
- while (regnum < NUM_REGS + NUM_PSEUDO_REGS)
- {
- sh_print_pseudo_register (gdbarch, file, regnum);
- regnum++;
- }
- }
-}
-
static int
sh_dsp_register_sim_regno (int nr)
{
@@ -2654,8 +2476,6 @@ sh_gdbarch_init (struct gdbarch_info inf
set_gdbarch_register_type (gdbarch, sh_default_register_type);
set_gdbarch_register_reggroup_p (gdbarch, sh_register_reggroup_p);
- set_gdbarch_print_registers_info (gdbarch, sh_print_registers_info);
-
set_gdbarch_breakpoint_from_pc (gdbarch, sh_breakpoint_from_pc);
set_gdbarch_print_insn (gdbarch, gdb_print_insn_sh);
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: [PATCH][SH] info registers
2005-11-16 16:48 [PATCH][SH] info registers Andrew STUBBS
@ 2005-11-16 16:54 ` Mark Kettenis
2005-11-23 18:39 ` Andrew STUBBS
0 siblings, 1 reply; 5+ messages in thread
From: Mark Kettenis @ 2005-11-16 16:54 UTC (permalink / raw)
To: andrew.stubbs; +Cc: gdb-patches
> Date: Wed, 16 Nov 2005 15:51:09 +0000
> From: Andrew STUBBS <andrew.stubbs@st.com>
>
> Hi,
>
> sh-tdep.c contains quite a lot of code to implement the 'info registers'
> command (and friends).
>
> It has a number of issues:
> - The double and vector registers are only in hex.
> - Invalid registers are still printed for some types (not that GDB
> knows which registers are invalid - that's my next patch).
> - The double registers are incorrectly named dr0..dr7, and the vector
> registers incorrectly named fv0..fv3. They should be dr0,dr2,dr4... and
> fv0,fv4,...
>
> I have discovered that the default code (when the architecture may
> override) does everything the SH code does right and yet does not do
> everything it does wrong. Specifically all the above problems are
> addressed.
>
> I therefore propose removing all this garbage entirely.
I'm very much in favour of this, but please wait a week or so with
checking this in to give people who actually use gdb for the SH the
opportunity to object.
Mark
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH][SH] info registers
2005-11-16 16:54 ` Mark Kettenis
@ 2005-11-23 18:39 ` Andrew STUBBS
2005-11-23 18:39 ` Daniel Jacobowitz
0 siblings, 1 reply; 5+ messages in thread
From: Andrew STUBBS @ 2005-11-23 18:39 UTC (permalink / raw)
To: Mark Kettenis; +Cc: gdb-patches
Mark Kettenis wrote:
> I'm very much in favour of this, but please wait a week or so with
> checking this in to give people who actually use gdb for the SH the
> opportunity to object.
>
> Mark
>
No objections after a week.
May I commit this patch?
Thanks
Andrew Stubbs
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH][SH] info registers
2005-11-23 18:39 ` Andrew STUBBS
@ 2005-11-23 18:39 ` Daniel Jacobowitz
2005-11-23 19:08 ` Andrew STUBBS
0 siblings, 1 reply; 5+ messages in thread
From: Daniel Jacobowitz @ 2005-11-23 18:39 UTC (permalink / raw)
To: Andrew STUBBS; +Cc: Mark Kettenis, gdb-patches
On Wed, Nov 23, 2005 at 01:02:54PM +0000, Andrew STUBBS wrote:
> Mark Kettenis wrote:
> >I'm very much in favour of this, but please wait a week or so with
> >checking this in to give people who actually use gdb for the SH the
> >opportunity to object.
> >
> >Mark
> >
>
> No objections after a week.
>
> May I commit this patch?
Yes, you may.
--
Daniel Jacobowitz
CodeSourcery, LLC
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2005-11-23 16:51 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-11-16 16:48 [PATCH][SH] info registers Andrew STUBBS
2005-11-16 16:54 ` Mark Kettenis
2005-11-23 18:39 ` Andrew STUBBS
2005-11-23 18:39 ` Daniel Jacobowitz
2005-11-23 19:08 ` Andrew STUBBS
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox