Re: [PATCH] ptrace.2: document struct ptrace_syscall_info

["Dmitry V. Levin" <ldv@altlinux.org>]

On Fri, Sep 13, 2019 at 08:48:21PM +0200, Michael Kerrisk (man-pages) wrote:
> On 9/13/19 5:40 PM, Dmitry V. Levin wrote:
> > On Thu, Sep 12, 2019 at 11:17:52AM +0200, Michael Kerrisk (man-pages) wrote:
> >> Hello Dmitry,
> >>
> >> On 9/11/19 7:38 PM, Dmitry V. Levin wrote:
> >>> Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
> >>> ---
> >>>  man2/ptrace.2 | 51 ++++++++++++++++++++++++++++++++++++++++++++++++++-
> >>>  1 file changed, 50 insertions(+), 1 deletion(-)
> >>>
> >>> diff --git a/man2/ptrace.2 b/man2/ptrace.2
> >>> index ff897bb19..b030247e6 100644
> >>> --- a/man2/ptrace.2
> >>> +++ b/man2/ptrace.2
> >>> @@ -1026,7 +1026,56 @@ If the size of the data to be written by the kernel exceeds the size
> >>>  specified by the
> >>>  .I addr
> >>>  argument, the output data is truncated.
> >>> -.\" FIXME Document 'struct ptrace_syscall_info'
> >>> +.IP
> >>> +The
> >>> +.I ptrace_syscall_info
> >>> +structure contains the following fields:
> >>> +.IP
> >>> +.in
> >>> +.EX
> >>> +struct ptrace_syscall_info {
> >>> +    __u8 op;                    /* PTRACE_SYSCALL_INFO_* value
> >>> +                                   describing the kind
> >>> +                                   of system call stop,
> >>> +                                   see <linux/ptrace.h> */
> >>> +    __u32 arch;                 /* AUDIT_ARCH_* value,
> >>> +                                   see seccomp(2) */
> >>> +    __u64 instruction_pointer;  /* CPU instruction pointer */
> >>> +    __u64 stack_pointer;        /* CPU stack pointer */
> >>> +    union {
> >>> +        struct {
> >>> +            __u64 nr;           /* System call number */
> >>> +            __u64 args[6];      /* System call arguments */
> >>> +        } entry;                /* Information specific to
> >>> +                                   system call entry stops */
> >>> +        struct {
> >>> +            __s64 rval;         /* System call return value */
> >>> +            __u8 is_error;      /* System call error flag */
> >>> +        } exit;                 /* Information specific to
> >>> +                                   system call exit stops */
> >>> +        struct {
> >>> +            __u64 nr;           /* System call number */
> >>> +            __u64 args[6];      /* System call arguments */
> >>> +            __u32 ret_data;     /* The SECCOMP_RET_DATA portion
> >>> +                                   of SECCOMP_RET_TRACE return
> >>> +                                   value */
> >>> +        } seccomp;              /* Information specific to
> >>> +                                   PTRACE_EVENT_SECCOMP stops */
> >>> +    };
> >>> +};
> >>> +.EE
> >>> +.in
> >>> +.IP
> >>> +.IR op ,
> >>> +.IR arch ,
> >>> +.IR instruction_pointer ,
> >>> +and
> >>> +.I stack_pointer
> >>> +fields are defined for all kinds of ptrace system call stops.
> >>> +The rest of the structure is a union, one should read only those fields
> >>> +that are meaningful for the kind of system call stop specified by the
> >>> +.IR op
> >>> +field.
> >>>  .\"
> >>>  .SS Death under ptrace
> >>>  When a (possibly multithreaded) process receives a killing signal
> >>
> >> Thanks. I've applied, and done some light editing.
> >>
> >> I also added some pieces. Could you please take a look at commits
> >> f04534d22c69 [1] and 227a3682312 [2] let me know if they are okay.
> > 
> > Thanks.  The only thing that needs correction is
> > "Boolean: is rval != 0?" comment about is_error field.
> > 
> > In fact, is_error is a boolean that tells whether rval field
> > contains a non-error return value (e.g. file descriptor) or
> > an error value (e.g. -ERRORCODE).
> 
> Thanks for checking this Dmitry. I see what I missed before.
> I've now changed the text here to be:
> 
>             __u8 is_error;     /* System call error flag;
>                                   Boolean: does rval contain
>                                   an error value (\-ERRCODE) or
>                                   a nonerror return value? */

Yes, that's better, thanks.


-- 
ldv