Is this on a 32-bit x86 kernel? zap_cursor_serialize() returns a
uin64_t, but DTrace is only capable of tracing the high-order portion of
this value (at least for the return probe). Looking at your output
below, it looks like you're only seeing 4 bytes of data, not 8.
- Eric
On Wed, Jun 21, 2006 at 08:35:41PM +0800, Jeremy Teo wrote:
> Hello,
>
> according to the comments, zap_cursor_serialize returns a cookie which
> always has it's lower 4 bits set to 0.
>
> However, while dtracing the cookie retunred by zap_cursor_serialize,
> this does not appear to be the case. ie. I get values such as
> ------------------------------
> bash-3.00# dtrace -n 'fbt::zap_cursor_serialize:return {printf("0x%x",
> arg1)}'
> dtrace: description 'fbt::zap_cursor_serialize:return ' matched 1 probe
> CPU ID FUNCTION:NAME
> 0 57277 zap_cursor_serialize:return 0x10038984
> 0 57277 zap_cursor_serialize:return 0x102ff2f3
> 0 57277 zap_cursor_serialize:return 0x10383ee0
> 0 57277 zap_cursor_serialize:return 0x109fbf15
> 0 57277 zap_cursor_serialize:return 0x10fcdb72
> 0 57277 zap_cursor_serialize:return 0x11292137
> 0 57277 zap_cursor_serialize:return 0x1157a001
> 0 57277 zap_cursor_serialize:return 0x118b126a
> 0 57277 zap_cursor_serialize:return 0x11bfc289
> 0 57277 zap_cursor_serialize:return 0x1206d830
> 0 57277 zap_cursor_serialize:return 0x1278e719
> 0 57277 zap_cursor_serialize:return 0x12808b52
> 0 57277 zap_cursor_serialize:return 0x1373e220
> 0 57277 zap_cursor_serialize:return 0x142cc5d7
> 0 57277 zap_cursor_serialize:return 0x142f0362
> 0 57277 zap_cursor_serialize:return 0x145dea2f
> 0 57277 zap_cursor_serialize:return 0x14757afa
> 0 57277 zap_cursor_serialize:return 0x14975e0c
> 0 57277 zap_cursor_serialize:return 0x14977749
> 0 57277 zap_cursor_serialize:return 0x14a4e0c6
> 0 57277 zap_cursor_serialize:return 0x14ec25df
> ------------------------------
>
> If the lower 4 bits are always set to 0, then the last hex digit
> should always be zero.
>
> Did I mess up somewhere, or are the comments wrong?
> --
> Regards,
> Jeremy
> _______________________________________________
> zfs-code mailing list
> zfs-code at opensolaris.org
> http://opensolaris.org/mailman/listinfo/zfs-code
--
Eric Schrock, Solaris Kernel Development http://blogs.sun.com/eschrock
