Darren Reed wrote: >> Add a new "offset:" output field to pfiles (in verbose mode, that is, >> lacking the -n switch). The field would be printed for any file that >> responds to lseek() with a valid offset, appearing just after the >> pathname (if any).
> Is this field displayed with sockets? > > While sockets may not be seek'able, my experience with this kind of > utility and extracting information from the kernel is that the "offset" > for sockets is actually the number of bytes written (as it is for any > file opened with O_TRUNC) and that can be of value. This would seem to depend on the behavior of sockfs, not on pfiles. http://cvs.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/sockfs/sockcommon_vnops.c 460 /* 461 * Sockets are not seekable. 462 * (and there is a bug to fix STREAMS to make them fail this as well). 463 */ 464 /*ARGSUSED*/ 465 int 466 socket_vop_seek(struct vnode *vp, offset_t ooff, offset_t *noffp, 467 caller_context_t *ct) 468 { 469 return (ESPIPE); 470 } Not this case. - Bart -- Bart Smaalders Solaris Kernel Performance barts at cyber.eng.sun.com http://blogs.sun.com/barts "You will contribute more with mercurial than with thunderbird."
