On 4/10/15 05:31, Peter Maydell wrote:
> On 27 March 2015 at 10:49, Chen Gang <xili_gchen_5...@hotmail.com> wrote:
[...]
>> +static inline void cpu_clone_regs(CPUTLGState *env, target_ulong newsp)
>> +{
>> + if (newsp) {
>> + env->regs[TILEGX_R_SP] = newsp;
>> + }
>> + env->regs[TILEGX_R_RE] = 0;
>
> This is slightly confusing, because the kernel code we're matching here
> doesn't call this register RE, it just uses 0:
> childregs->regs[0] = 0; /* return value is zero */
>
TILEGX_R_RE is just 0, I define a macro instead the hardcode number in
cpu.h with the comment for it.
[...]
>> +
>> +/* this struct defines a stack used during syscall handling */
>> +
>> +typedef struct target_sigaltstack {
>> + abi_ulong ss_sp;
>> + abi_ulong ss_size;
>> + abi_long ss_flags;
>> +} target_stack_t;
>
> Where does this come from? It doesn't match the kernel's
> generic-headers struct layout.
>
Oh, sorry, originally, I guess, I only copied it from microblaze, did
not check kernel.
I shall use generic-headers which tilegx will use (the result will like
alpha has done):
typedef struct target_sigaltstack {
abi_ulong ss_sp;
int32_t ss_flags;
int32_t dummy;
abi_ulong ss_size;
} target_stack_t;
[...]
>> +
>> +struct target_ipc_perm {
>> + abi_int __key; /* Key. */
>> + abi_uint uid; /* Owner's user ID. */
>> + abi_uint gid; /* Owner's group ID. */
>> + abi_uint cuid; /* Creator's user ID. */
>> + abi_uint cgid; /* Creator's group ID. */
>> + abi_uint mode; /* Read/write permission. */
>> + abi_ushort __seq; /* Sequence number. */
>> + abi_ushort __pad2;
>> + abi_ulong __unused1;
>> + abi_ulong __unused2;
>> +};
>
> Again, doesn't seem to match kernel?
>
For me, it matches kernel. mode is abi_uint (__kernel_mode_t is 32-bit).
Thanks.
--
Chen Gang
Open, share, and attitude like air, water, and life which God blessed
