[Bug translator/11871] New: cast failure

[Bug translator/11871] New: cast failure

[Menny_Hamburger at dell dot com]

I am using systemtap over a 32 bit Samba running in a 64 bit machine.
I added the folllowing type declarations to /tmp/mytpe.h:
%{
typedef struct exa_dent_name {
        size_t len;
        char *name;
} exa_dent_name;


struct xattr_stat {
        struct stat stat;
        uint32_t createtime_sec;
        uint32_t createtime_nsec;
};
typedef struct exanet_dir {
        int magic;
        int path_len;
        char *path;
        const connection_struct *conn;
        int dirfd;
        void *buf;
        void *bufp;
        int buf_id;
        int buf_size;
        uint64_t cookie;
        int eod;
        exa_dent_name prev_name;
        struct xattr_stat *prev_stat;
        dev_t st_dev;
        blksize_t st_blksize;
        char nfs_fh[32];
        void* prev_dir_pos;
        short int override_seekdir;

} exanet_dir;
%}

I then used the following probe:
probe process(/usr/local/samba/lib/vfs/fs.so").function("fs_readdir")
{
    printf("%s: dir magic is %d\n", 
           @cast($dir, "exanet_dir", "</tmp/mytype.h>")->magic)
}

I got the following error:
semantic error: type definition 'exanet_dir' not found: identifier '@cast' 
at ./smbd_guru1.stp:9:3
        source:                 @cast($dir, "exanet_dir", "</tmp/mytype.h>")-
>magic)

This does not seem a 32 over 64 bit issue but since I am attaching the binary, 
It is important to mention this.

-- 
           Summary: cast failure
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: translator
        AssignedTo: systemtap at sources dot redhat dot com
        ReportedBy: Menny_Hamburger at dell dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[Menny_Hamburger at dell dot com]

------- Additional Comments From Menny_Hamburger at dell dot com  2010-08-03 10:52 -------
Created an attachment (id=4903)
 --> (http://sourceware.org/bugzilla/attachment.cgi?id=4903&action=view)
binary file on which the probe failed


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2010-08-03 11:53 -------
(In reply to comment #2)
> Manny, [...]
Sorry, Menny.  (What an awesome name.)


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2010-08-03 11:52 -------
Manny, could you attach the header in question?

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[Menny_Hamburger at dell dot com]

------- Additional Comments From Menny_Hamburger at dell dot com  2010-08-03 11:58 -------
The header is just a copy/paste version of the declarations I sent in the 
previous message into /tmp/mytype.h.
Unknown types declarations inside exanet_dir, such as connection_struct can be 
changed to void * in order to eliminate necessity for additional headers.

M.

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2010-08-03 12:07 -------
You don't need the %{ %} stuff around the /tmp/mytpe.h file.
Maybe run stap -vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv to see
any gcc error messages while trying to process your header.
No, make that stap -vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv.
Just to be sure.


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[Menny_Hamburger at dell dot com]

------- Additional Comments From Menny_Hamburger at dell dot com  2010-08-08 09:24 -------
As I said before, the %{ %} are not included in myfile.h - this decaration 
used to be in the script itself and I did copy and past by mistake here.

I tested this with systemtap 1.3-2 from fedora - does not work.
No gcc errors - just seems /tmp/mytype.h is not parsed correctly although I 
see it in the command line (-i /tmp/mytype.h).



-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[Menny_Hamburger at dell dot com]

------- Additional Comments From Menny_Hamburger at dell dot com  2010-08-08 10:54 -------
A second inspection shows that gcc does show errors.

Thnx

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2010-08-11 14:18 -------
OK, can we close this as a NOTABUG then?

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug translator/11871] cast failure

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2010-08-31 21:12 -------
there appears not to be a stap bug here

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |RESOLVED
         Resolution|                            |WORKSFORME


http://sourceware.org/bugzilla/show_bug.cgi?id=11871

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.