pbuf(9) [freebsd man page]

PBUF(9) 						   BSD Kernel Developer's Manual						   PBUF(9)

NAME

     pbuf, getpbuf, trypbuf, relpbuf -- functions for managing physical buffers

SYNOPSIS

     #include <sys/param.h>
     #include <sys/systm.h>
     #include <sys/bio.h>
     #include <sys/buf.h>

     struct buf *
     getpbuf(int *pfreecnt);

     struct buf *
     trypbuf(int *pfreecnt);

     void
     relpbuf(struct buf *bp, int *pfreecnt);

DESCRIPTION

     These functions are used to allocate and release physical buffers.

     The physical buffers are allocated at system startup and are maintained in a separate pool from the main system buffers.  They are intended
     for use by subsystems that cannot or should not be reliant on the main pool of buffers (for example the swap pager).  The system allocates
     between 16 and 256 physical buffers depending on the amount of memory in the system.

     Each subsystem that allocates buffers via these calls is expected to manage its own percentage free counter.  If the value is initialized to
     -1 the number of buffers available to the subsystem is limited only by the number of physical buffers available.  The number of buffers is
     stored in nswbuf which is defined in <sys/buf.h> and initialized in cpu_startup().  A recommended initialization value is 1/2 nswbuf.

     The getpbuf() function returns the first available buffer to the user.  If there are no buffers available, getpbuf() will sleep waiting for
     one to become available.  If pfreecnt is zero, getpbuf() will sleep until it increases.  pfreecnt is decremented prior to returning.

     The trypbuf() function returns the first available buffer.  If there are no buffers available, NULL is returned.  As well, if pfreecnt is
     zero, NULL is returned.  pfreecnt is decremented prior to returning a valid buffer.  If NULL is returned, pfreecnt is not modified.

     The relpbuf() function releases the buffer back to the free list.	If the buffers b_rcred or b_wcred structures are not NULL, they are freed.
     See crfree(9).

     pfreecnt is incremented prior to returning.

RETURN VALUES

     getpbuf() and trypbuf() return a pointer to the buffer.  In the case of trypbuf(), NULL can also be returned indicating that there are no
     buffers available.

AUTHORS

     This manual page was written by Chad David <davidc@acns.ab.ca>.

BSD
								   July 9, 2001 							       BSD

Check Out this Related Man Page

VINVALBUF(9)						   BSD Kernel Developer's Manual					      VINVALBUF(9)

NAME

     vinvalbuf -- flushes and invalidates all buffers associated with a vnode

SYNOPSIS

     #include <sys/param.h>
     #include <sys/vnode.h>

     int
     vinvalbuf(struct vnode *vp, int flags, struct ucred *cred, int slpflag, int slptimeo);

DESCRIPTION

     The vinvalbuf() function invalidates all of the buffers associated with the given vnode.  This includes buffers on the clean list and the
     dirty list.  If the V_SAVE flag is specified then the buffers on the dirty list are synced prior to being released.  If there is a VM Object
     associated with the vnode, it is removed.

     Its arguments are:

     vp        A pointer to the vnode whose buffers will be invalidated.

     flags     The only supported flag is V_SAVE and it indicates that dirty buffers should be synced with the disk.

     cred      The user credentials that are used to VOP_FSYNC(9) buffers if V_SAVE is set.

     slpflag   The slp flag that will be used in the priority of any sleeps in the function.

     slptimeo  The timeout for any sleeps in the function.

LOCKS

     The vnode is assumed to be locked prior to the call and remains locked upon return.

     Giant must be held by prior to the call and remains locked upon return.

RETURN VALUES

     A 0 value is returned on success.

PSEUDOCODE

	   vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
	   error = vinvalbuf(devvp, V_SAVE, cred, 0, 0);
	   VOP_UNLOCK(devvp, 0);
	   if (error)
		   return (error);

ERRORS

     [ENOSPC]		The file system is full.  (With V_SAVE)

     [EDQUOT]		Disc quota exceeded.  (With V_SAVE)

     [EWOULDBLOCK]	Sleep operation timed out.  (See slptimeo)

     [ERESTART] 	A signal needs to be delivered and the system call should be restarted.  (With PCATCH set in slpflag)

     [EINTR]		The system has been interrupted by a signal.  (With PCATCH set in slpflag)

SEE ALSO

     tsleep(9), VOP_FSYNC(9)

AUTHORS

     This manual page was written by Chad David <davidc@acns.ab.ca>.

BSD
								 October 20, 2008							       BSD

Linux and UNIX Man Pages

pbuf(9) [freebsd man page]

Check Out this Related Man Page