[PATCH] aarch64/gdbserver: fix floating point registers display
catalin.udma@freescale.com
catalin.udma@freescale.com
Wed Aug 13 12:43:00 GMT 2014
Hi, all,
I'll add some details, just to have a better view about the cpsr size changes:
- CPSR register size was initially 32 bits (features/aarch64-core.xml), but the
register size has been changed in commit a4d9ba85e (2013-12-18):
features/aarch64-core.xml (cpsr): Changed to be 64 bits
- When Changing .xml file, there should be regenerated accordingly aarch64.c and
regformats/aarch64.dat, to match the new size definition in the XML file.
- features/aarch64.c has been regenerated in the above mentioned commit, while the
file regformats/aarch64.dat wasn't regenerated, thus using the previous cpsr size of 32 bits
The regformats/aarch64.dat update in my commit fixes the de-synchronization problem
where gdb is seeing the cpsr size 64-bits while gdbserver 32-bits.
Regards,
Catalin
> -----Original Message-----
> From: Philippe Waroquiers [mailto:philippe.waroquiers@skynet.be]
> Sent: Wednesday, August 13, 2014 3:26 PM
> To: Yao Qi
> Cc: Richard Earnshaw; Udma Catalin-Dan-B32721; gdb-patches@sourceware.org
> Subject: Re: [PATCH] aarch64/gdbserver: fix floating point registers
> display
>
> On Tue, 2014-08-12 at 18:25 +0800, Yao Qi wrote:
> > On 08/12/2014 05:43 PM, Richard Earnshaw wrote:
> > > This doesn't seem right to me. The CPSR is a 32-bit register, not a
> > > 64-bit one.
> >
> > cpsr is 64-bit in the target description, see
> > gdb/features/aarch64-core.xml,
> >
> > <reg name="cpsr" bitsize="64"/>
> >
> > We need to fix it.
> The 'it' in 'fix it' is ambiguous to me.
> Does the 'it' mean:
> fix aarch64-core.xml to change cpsr to 32 bits ?
> or does that confirm the initial proposal i.e.
> fix e.g. aarch64.dat to change cpsr to 64 bits ?
>
> Philippe
> (following the discussion, to update Valgrind aarch64 gdbserver
> accordingly)
More information about the Gdb-patches
mailing list