[v5,1/3] RISC-V/Linux/native: Determine FLEN dynamically

  Fix RISC-V native Linux support to handle a 64-bit FPU (FLEN == 64) with 
both RV32 and RV64 systems, which is a part of the current Linux ABI for 
hard-float systems, rather than assuming that (FLEN == XLEN) in target 
description determination and that (FLEN == 64) in register access.

We can do better however and not rely on any particular value of FLEN 
and probe for it dynamically, by observing that the PTRACE_GETREGSET 
ptrace(2) call will only accept an exact regset size, and that will 
reflect FLEN.  Therefore iterate over the call in target description 
determination with a geometrically increasing regset size until a match 
is marked by a successful ptrace(2) call completion or we run beyond the 
maximum size we can support.

Update register accessors accordingly, using FLEN determined to size the 
buffer used for NT_PRSTATUS requests and then to exchange data with the 
regcache.

Also handle a glibc bug where ELF_NFPREG is defined in terms of NFPREG, 
however NFPREG is nowhere defined.

	gdb/
	* riscv-linux-nat.c [!NFPREG] (NFPREG): New macro.
	(supply_fpregset_regnum, fill_fpregset): Handle regset buffer 
	offsets according to FLEN determined.
	(riscv_linux_nat_target::read_description): Determine FLEN 
	dynamically.
	(riscv_linux_nat_target::fetch_registers): Size regset buffer 
	according to FLEN determined.
	(riscv_linux_nat_target::store_registers): Likewise.
---
No changes from v4.

No changes from v3.

Changes from v2:

- Avoid an ambiguous FGR reference in 
  `riscv_linux_nat_target::read_description'.

- Advance through the buffer accessed in `supply_fpregset_regnum' and 
  `fill_fpregset' so as to avoid doing a variable multiplication in a 
  loop; also precalculate the buffer offset for single register accesses 
  to avoid excessive line wrapping and improve code readability.

- Move the `i' variable declaration to the beginning of `fill_fpregset', 
  for consistency with `supply_fpregset_regnum' (since the loop is being 
  rewritten anyway).

Changes from v1:

- Also set the size of the regset buffer dynamically in 
  `riscv_linux_nat_target::fetch_registers' and 
  `riscv_linux_nat_target::store_registers', and update `fill_fpregset' 
  and `supply_fpregset_regnum' accordingly.
---
 gdb/riscv-linux-nat.c |  110 ++++++++++++++++++++++++++++++++++++++------------
 1 file changed, 84 insertions(+), 26 deletions(-)

gdb-riscv-linux-nat-flen.diff
  

* Maciej W. Rozycki <macro@wdc.com> [2020-02-01 20:20:46 +0000]:

> Fix RISC-V native Linux support to handle a 64-bit FPU (FLEN == 64) with 
> both RV32 and RV64 systems, which is a part of the current Linux ABI for 
> hard-float systems, rather than assuming that (FLEN == XLEN) in target 
> description determination and that (FLEN == 64) in register access.
> 
> We can do better however and not rely on any particular value of FLEN 
> and probe for it dynamically, by observing that the PTRACE_GETREGSET 
> ptrace(2) call will only accept an exact regset size, and that will 
> reflect FLEN.  Therefore iterate over the call in target description 
> determination with a geometrically increasing regset size until a match 
> is marked by a successful ptrace(2) call completion or we run beyond the 
> maximum size we can support.
> 
> Update register accessors accordingly, using FLEN determined to size the 
> buffer used for NT_PRSTATUS requests and then to exchange data with the 
> regcache.
> 
> Also handle a glibc bug where ELF_NFPREG is defined in terms of NFPREG, 
> however NFPREG is nowhere defined.
> 
> 	gdb/
> 	* riscv-linux-nat.c [!NFPREG] (NFPREG): New macro.
> 	(supply_fpregset_regnum, fill_fpregset): Handle regset buffer 
> 	offsets according to FLEN determined.
> 	(riscv_linux_nat_target::read_description): Determine FLEN 
> 	dynamically.
> 	(riscv_linux_nat_target::fetch_registers): Size regset buffer 
> 	according to FLEN determined.
> 	(riscv_linux_nat_target::store_registers): Likewise.

Thanks for working on this.

I'm happy for this patch to go in, with a couple of minor adjustments
that I've noted inline below.

Before you merge I just wanted to double check; given this patch is
from a wdc.com email, do you have a copyright assignment in place from
wdc?

Thanks,
Andrew

> ---
> No changes from v4.
> 
> No changes from v3.
> 
> Changes from v2:
> 
> - Avoid an ambiguous FGR reference in 
>   `riscv_linux_nat_target::read_description'.
> 
> - Advance through the buffer accessed in `supply_fpregset_regnum' and 
>   `fill_fpregset' so as to avoid doing a variable multiplication in a 
>   loop; also precalculate the buffer offset for single register accesses 
>   to avoid excessive line wrapping and improve code readability.
> 
> - Move the `i' variable declaration to the beginning of `fill_fpregset', 
>   for consistency with `supply_fpregset_regnum' (since the loop is being 
>   rewritten anyway).
> 
> Changes from v1:
> 
> - Also set the size of the regset buffer dynamically in 
>   `riscv_linux_nat_target::fetch_registers' and 
>   `riscv_linux_nat_target::store_registers', and update `fill_fpregset' 
>   and `supply_fpregset_regnum' accordingly.
> ---
>  gdb/riscv-linux-nat.c |  110 ++++++++++++++++++++++++++++++++++++++------------
>  1 file changed, 84 insertions(+), 26 deletions(-)
> 
> gdb-riscv-linux-nat-flen.diff
> Index: binutils-gdb/gdb/riscv-linux-nat.c
> ===================================================================
> --- binutils-gdb.orig/gdb/riscv-linux-nat.c
> +++ binutils-gdb/gdb/riscv-linux-nat.c
> @@ -28,6 +28,11 @@
>  
>  #include <sys/ptrace.h>
>  
> +/* Work around glibc header breakage causing ELF_NFPREG not to be usable.  */
> +#ifndef NFPREG
> +# define NFPREG 33
> +#endif
> +
>  /* RISC-V Linux native additions to the default linux support.  */
>  
>  class riscv_linux_nat_target final : public linux_nat_target
> @@ -88,21 +93,35 @@ static void
>  supply_fpregset_regnum (struct regcache *regcache, const prfpregset_t *fpregs,
>  			int regnum)
>  {
> +  int flen = register_size (regcache->arch (), RISCV_FIRST_FP_REGNUM);
> +  union
> +    {
> +      const prfpregset_t *fpregs;
> +      const gdb_byte *buf;
> +    }
> +  fpbuf = { .fpregs = fpregs };
>    int i;
>  
>    if (regnum == -1)
>      {
>        /* We only support the FP registers and FCSR here.  */
> -      for (i = RISCV_FIRST_FP_REGNUM; i <= RISCV_LAST_FP_REGNUM; i++)
> -	regcache->raw_supply (i, &fpregs->__d.__f[i - RISCV_FIRST_FP_REGNUM]);
> +      for (i = RISCV_FIRST_FP_REGNUM;
> +	   i <= RISCV_LAST_FP_REGNUM;
> +	   i++, fpbuf.buf += flen)
> +	regcache->raw_supply (i, fpbuf.buf);
>  
> -      regcache->raw_supply (RISCV_CSR_FCSR_REGNUM, &fpregs->__d.__fcsr);
> +      regcache->raw_supply (RISCV_CSR_FCSR_REGNUM, fpbuf.buf);
>      }
>    else if (regnum >= RISCV_FIRST_FP_REGNUM && regnum <= RISCV_LAST_FP_REGNUM)
> -    regcache->raw_supply (regnum,
> -			  &fpregs->__d.__f[regnum - RISCV_FIRST_FP_REGNUM]);
> +    {
> +      fpbuf.buf += flen * (regnum - RISCV_FIRST_FP_REGNUM);
> +      regcache->raw_supply (regnum, fpbuf.buf);
> +    }
>    else if (regnum == RISCV_CSR_FCSR_REGNUM)
> -    regcache->raw_supply (RISCV_CSR_FCSR_REGNUM, &fpregs->__d.__fcsr);
> +    {
> +      fpbuf.buf += flen * (RISCV_LAST_FP_REGNUM - RISCV_FIRST_FP_REGNUM + 1);
> +      regcache->raw_supply (RISCV_CSR_FCSR_REGNUM, fpbuf.buf);
> +    }
>  }
>  
>  /* Copy all floating point registers from regset FPREGS into REGCACHE.  */
> @@ -145,19 +164,35 @@ void
>  fill_fpregset (const struct regcache *regcache, prfpregset_t *fpregs,
>  	       int regnum)
>  {
> +  int flen = register_size (regcache->arch (), RISCV_FIRST_FP_REGNUM);
> +  union
> +    {
> +      prfpregset_t *fpregs;
> +      gdb_byte *buf;
> +    }
> +  fpbuf = { .fpregs = fpregs };
> +  int i;
> +
>    if (regnum == -1)
>      {
>        /* We only support the FP registers and FCSR here.  */
> -      for (int i = RISCV_FIRST_FP_REGNUM; i <= RISCV_LAST_FP_REGNUM; i++)
> -	regcache->raw_collect (i, &fpregs->__d.__f[i - RISCV_FIRST_FP_REGNUM]);
> +      for (i = RISCV_FIRST_FP_REGNUM;
> +	   i <= RISCV_LAST_FP_REGNUM;
> +	   i++, fpbuf.buf += flen)
> +	regcache->raw_collect (i, fpbuf.buf);
>  
> -      regcache->raw_collect (RISCV_CSR_FCSR_REGNUM, &fpregs->__d.__fcsr);
> +      regcache->raw_collect (RISCV_CSR_FCSR_REGNUM, fpbuf.buf);
>      }
>    else if (regnum >= RISCV_FIRST_FP_REGNUM && regnum <= RISCV_LAST_FP_REGNUM)
> -    regcache->raw_collect (regnum,
> -			   &fpregs->__d.__f[regnum - RISCV_FIRST_FP_REGNUM]);
> +    {
> +      fpbuf.buf += flen * (regnum - RISCV_FIRST_FP_REGNUM);
> +      regcache->raw_collect (regnum, fpbuf.buf);
> +    }
>    else if (regnum == RISCV_CSR_FCSR_REGNUM)
> -    regcache->raw_collect (RISCV_CSR_FCSR_REGNUM, &fpregs->__d.__fcsr);
> +    {
> +      fpbuf.buf += flen * (RISCV_LAST_FP_REGNUM - RISCV_FIRST_FP_REGNUM + 1);
> +      regcache->raw_collect (RISCV_CSR_FCSR_REGNUM, fpbuf.buf);
> +    }
>  }
>  
>  /* Return a target description for the current target.  */
> @@ -166,8 +201,8 @@ const struct target_desc *
>  riscv_linux_nat_target::read_description ()
>  {
>    struct riscv_gdbarch_features features;
> -  struct iovec iov;
>    elf_fpregset_t regs;
> +  int flen;
>    int tid;
>  
>    /* Figuring out xlen is easy.  */
> @@ -175,19 +210,40 @@ riscv_linux_nat_target::read_description
>  
>    tid = inferior_ptid.lwp ();
>  
> -  iov.iov_base = &regs;
> -  iov.iov_len = sizeof (regs);
> +  /* Start with no f-registers.  */
> +  features.flen = 0;
>  
> -  /* Can we fetch the f-registers?  */
> -  if (ptrace (PTRACE_GETREGSET, tid, NT_FPREGSET,
> -	      (PTRACE_TYPE_ARG3) &iov) == -1)
> -    features.flen = 0;		/* No f-registers.  */
> -  else
> +  /* How much worth of f-registers can we fetch if any?  */
> +  for (flen = sizeof (regs.__f.__f[0]); ; flen *= 2)
>      {
> -      /* TODO: We need a way to figure out the actual length of the
> -	 f-registers.  We could have 64-bit x-registers, with 32-bit
> -	 f-registers.  For now, just assumed xlen and flen match.  */
> -      features.flen = features.xlen;
> +      size_t regset_size;
> +      struct iovec iov;
> +
> +      /* Regsets have a uniform slot size, so we count FSCR like
> +	 an FP data register.  */
> +      regset_size = ELF_NFPREG * flen;
> +      if (regset_size > sizeof (regs))
> +	break;
> +
> +      iov.iov_base = &regs;
> +      iov.iov_len = regset_size;
> +      if (ptrace (PTRACE_GETREGSET, tid, NT_FPREGSET,
> +		  (PTRACE_TYPE_ARG3) &iov) == -1)
> +	{
> +	  switch (errno)
> +	    {
> +	    case EINVAL:
> +	      continue;
> +	    case EIO:
> +	      break;
> +	    default:
> +	      perror_with_name (_("Couldn't get registers"));
> +	      break;
> +	    }
> +	}
> +      else
> +	features.flen = flen;
> +      break;
>      }
>  
>    return riscv_create_target_description (features);
> @@ -228,7 +284,8 @@ riscv_linux_nat_target::fetch_registers
>        elf_fpregset_t regs;
>  
>        iov.iov_base = &regs;
> -      iov.iov_len = sizeof (regs);
> +      iov.iov_len = ELF_NFPREG * register_size (regcache->arch (),
> +						RISCV_FIRST_FP_REGNUM);

I think that this would be made clearer, and more robust if we added
this assertion:

  gdb_assert (iov.iov_len <= sizeof (regs));

>  
>        if (ptrace (PTRACE_GETREGSET, tid, NT_FPREGSET,
>  		  (PTRACE_TYPE_ARG3) &iov) == -1)
> @@ -289,7 +346,8 @@ riscv_linux_nat_target::store_registers
>        elf_fpregset_t regs;
>  
>        iov.iov_base = &regs;
> -      iov.iov_len = sizeof (regs);
> +      iov.iov_len = ELF_NFPREG * register_size (regcache->arch (),
> +						RISCV_FIRST_FP_REGNUM);

And I think we should add the assertion here too.

With these assertions, and assuming you have a suitable copyright
assignment, then feel free to merge this.

Thanks,
Andrew


>  
>        if (ptrace (PTRACE_GETREGSET, tid, NT_FPREGSET,
>  		  (PTRACE_TYPE_ARG3) &iov) == -1)