[PATCH] gdb/riscv: Improve flen length determiniation

[PATCH] gdb/riscv: Improve flen length determiniation

[Simon Cook]

This solves an assertion failure when a remote provides a target
description which only refers to floating point registers by their
hardware name (e.g. f0), rather than their ABI name (e.g. ft0). GDB
assumed that should the floating point register feature be presented,
it would contain a register called ft0.

The floating point length is now instead determined by searching for
the same register, but looking for any of its aliases.

gdb/ChangeLog:

	* riscv-tdep.c (riscv_gdbarch_init): Support determining flen from
	target descriptions using exclusively floating point register name
	aliases.
---
 gdb/ChangeLog    |  6 ++++++
 gdb/riscv-tdep.c | 17 ++++++++++++++++-
 2 files changed, 22 insertions(+), 1 deletion(-)

diff --git a/gdb/riscv-tdep.c b/gdb/riscv-tdep.c
index 4fe07ef437..072c8e3720 100644
--- a/gdb/riscv-tdep.c
+++ b/gdb/riscv-tdep.c
@@ -3094,7 +3094,22 @@ riscv_gdbarch_init (struct gdbarch_info info,
       valid_p &= riscv_check_tdesc_feature (tdesc_data, feature_fpu,
                                             &riscv_freg_feature);
 
-      int bitsize = tdesc_register_bitsize (feature_fpu, "ft0");
+      /* Search for the first floating point register (by any alias), to
+         determine the bitsize.  */
+      int bitsize = -1;
+      const auto &fp0 = riscv_freg_feature.registers[0];
+
+      for (const char *name : fp0.names)
+	{
+	  if (tdesc_unnumbered_register (feature_fpu, name))
+	    {
+	      bitsize = tdesc_register_bitsize (feature_fpu, name);
+	      break;
+	    }
+	}
+
+      gdb_assert (bitsize != -1);
+
       features.flen = (bitsize / 8);
 
       if (riscv_debug_gdbarch)
-- 
2.17.1

Re: [PATCH] gdb/riscv: Improve flen length determiniation

[Andrew Burgess]

* Simon Cook <simon.cook@embecosm.com> [2019-05-22 12:37:45 +0100]:

> This solves an assertion failure when a remote provides a target
> description which only refers to floating point registers by their
> hardware name (e.g. f0), rather than their ABI name (e.g. ft0). GDB
> assumed that should the floating point register feature be presented,
> it would contain a register called ft0.
> 
> The floating point length is now instead determined by searching for
> the same register, but looking for any of its aliases.
> 
> gdb/ChangeLog:
> 
> 	* riscv-tdep.c (riscv_gdbarch_init): Support determining flen from
> 	target descriptions using exclusively floating point register name
> 	aliases.

Thanks for looking at this.

I've pushed your patch to master.

Thanks,
Andrew


> ---
>  gdb/ChangeLog    |  6 ++++++
>  gdb/riscv-tdep.c | 17 ++++++++++++++++-
>  2 files changed, 22 insertions(+), 1 deletion(-)
> 
> diff --git a/gdb/riscv-tdep.c b/gdb/riscv-tdep.c
> index 4fe07ef437..072c8e3720 100644
> --- a/gdb/riscv-tdep.c
> +++ b/gdb/riscv-tdep.c
> @@ -3094,7 +3094,22 @@ riscv_gdbarch_init (struct gdbarch_info info,
>        valid_p &= riscv_check_tdesc_feature (tdesc_data, feature_fpu,
>                                              &riscv_freg_feature);
>  
> -      int bitsize = tdesc_register_bitsize (feature_fpu, "ft0");
> +      /* Search for the first floating point register (by any alias), to
> +         determine the bitsize.  */
> +      int bitsize = -1;
> +      const auto &fp0 = riscv_freg_feature.registers[0];
> +
> +      for (const char *name : fp0.names)
> +	{
> +	  if (tdesc_unnumbered_register (feature_fpu, name))
> +	    {
> +	      bitsize = tdesc_register_bitsize (feature_fpu, name);
> +	      break;
> +	    }
> +	}
> +
> +      gdb_assert (bitsize != -1);
> +
>        features.flen = (bitsize / 8);
>  
>        if (riscv_debug_gdbarch)
> -- 
> 2.17.1
>