This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC][PATCH 4/4] gdbserver: Add RISC-V/Linux support


>>>>> "Maciej" == Maciej W Rozycki <macro@wdc.com> writes:

Maciej> Implement RISC-V/Linux support for both RV64 and RV32 systems, including 
Maciej> XML target description handling based on features determined, GPR and 
Maciej> FPR regset support including dynamic sizing of the latter, and software 
Maciej> breakpoint handling.

I saw a couple of small nits here.

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

In case you haven't already, please report this to glibc.

Maciej> As the architecture names come from BFD it will clearly have to be there 
Maciej> that the problem is fixed, with the ISA suffixes driving the choice of 
Maciej> instructions handled by the disassembler and possibly other functions.

Seems reasonable enough.  I don't really know much about this area
though.

Maciej>  Also I think large parts of `riscv_linux_nat_target::read_description' in 
Maciej> RISC-V/Linux/native support could be factored out to a piece of code 
Maciej> shared with `riscv_arch_setup'.  This might be less important a problem, 
Maciej> but nevertheless I think it should be done sooner than later so as not to 
Maciej> make the two pieces diverge

Agreed.

Maciej> +/* Implementation of linux_target_ops method "regs_info".  */
Maciej> +
Maciej> +static const struct regs_info *
Maciej> +riscv_regs_info (void)

We stopped using "(void)" in new code, in favor of just "()".
This showed up in a few spots.

Tom


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]