On 3/3/10 4:17 PM, Richard Henderson wrote:
+ float64 sc_fpregs[2]; /* room for two fp registers */
...
- int f_fpcntl[3];
- int f_fpregs[8*3];
+ uint32_t f_fpcntl[3];
+ float64 f_fpregs[8];
Surely these float64 uses are incorrect. The kernel uses 3*int
at both of these places, which matches up with the 96-bit values
in the regular m68k fpu.
ColdFire FPU registers are 64-bit wide, this is one of the difference
between ColdFire and classic m68k.
The kernel patch to sigcontext.h to support ColdFire was merged several
days ago at
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=arch/m68k/include/asm/sigcontext.h;h=1320eaa4cc2aab4b531a565c57ab62afb30bd0ec;hb=HEAD
.
And thanks to your comment I noticed that an old version of the
sigcontext patch was merged into the kernel in the above commit. Will
need to fix it.
Regards,
--
Maxim Kuvyrkov
CodeSourcery
ma...@codesourcery.com
(650) 331-3385 x724
