Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Ronald Klop 

On Tue, 19 Nov 2013 22:08:21 +0100, Luigi Rizzo  wrote:


On Mon, Nov 18, 2013 at 10:58:14PM +, George V. Neville-Neil wrote:

Author: gnn
Date: Mon Nov 18 22:58:14 2013
New Revision: 258328
URL: http://svnweb.freebsd.org/changeset/base/258328

Log:
  Allow ethernet drivers to pass in packets connected via the nextpkt  
pointer.
  Handling packets in this way allows drivers to amortize work during  
packet reception.


yes.

This is only a first step and eventually we should pass the entire
batch to the netisr handler to further reduce overhead.

Some of the followup emails suggested to change the argument from
struct mbuf * to something else.
I do think we should change it, but what we need is a struct with
head and tail pointers _and_ a counter, because sometimes the code
downstream may have to append the mbuf/batch to a queue,
and these extra fields would save iterating through the chain.

Related to this: at some point we should also address batching
in the transmit path, and for that we will eventually need to
introduce a 'more packets to come' flag to the API/mbuf so that
intermediate functions in the path will build batches before passing
them down.

cheers
luigi


The Apache webserver works like this also. It has a 'bucket'-list which  
chains all data of a request and can also contain flush-commands in  
between. Probably nice to look what they do and in what ways.


Ronald.
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258365 - in head/sys: kern vm

2013-11-20 Thread Konstantin Belousov 
Author: kib
Date: Wed Nov 20 08:45:26 2013
New Revision: 258365
URL: http://svnweb.freebsd.org/changeset/base/258365

Log:
  Revert back to use int for the page counts.  In vn_io_fault(), the i/o
  is chunked to pieces limited by integer io_hold_cnt tunable, while
  vm_fault_quick_hold_pages() takes integer max_count as the upper bound.
  
  Rearrange the checks to correctly handle overflowing address arithmetic.
  
  Submitted by: bde
  Tested by:pho
  Discussed with:   alc
  MFC after:1 week

Modified:
  head/sys/kern/vfs_vnops.c
  head/sys/vm/vm_fault.c

Modified: head/sys/kern/vfs_vnops.c
==
--- head/sys/kern/vfs_vnops.c   Wed Nov 20 02:20:27 2013(r258364)
+++ head/sys/kern/vfs_vnops.c   Wed Nov 20 08:45:26 2013(r258365)
@@ -933,9 +933,8 @@ vn_io_fault(struct file *fp, struct uio 
void *rl_cookie;
struct mount *mp;
vm_page_t *prev_td_ma;
-   int error, save, saveheld, prev_td_ma_cnt;
+   int error, cnt, save, saveheld, prev_td_ma_cnt;
vm_offset_t addr, end;
-   vm_size_t cnt;
vm_prot_t prot;
size_t len, resid;
ssize_t adv;
@@ -1008,21 +1007,20 @@ vn_io_fault(struct file *fp, struct uio 
uio_clone->uio_iovcnt--;
continue;
}
-
-   addr = (vm_offset_t)uio_clone->uio_iov->iov_base;
+   if (len > io_hold_cnt * PAGE_SIZE)
+   len = io_hold_cnt * PAGE_SIZE;
+   addr = (uintptr_t)uio_clone->uio_iov->iov_base;
end = round_page(addr + len);
-   cnt = howmany(end - trunc_page(addr), PAGE_SIZE);
+   if (end < addr) {
+   error = EFAULT;
+   break;
+   }
+   cnt = atop(end - trunc_page(addr));
/*
 * A perfectly misaligned address and length could cause
 * both the start and the end of the chunk to use partial
 * page.  +2 accounts for such a situation.
 */
-   if (cnt > io_hold_cnt + 2) {
-   len = io_hold_cnt * PAGE_SIZE;
-   KASSERT(howmany(round_page(addr + len) -
-   trunc_page(addr), PAGE_SIZE) <= io_hold_cnt + 2,
-   ("cnt overflow"));
-   }
cnt = vm_fault_quick_hold_pages(&td->td_proc->p_vmspace->vm_map,
addr, len, prot, ma, io_hold_cnt + 2);
if (cnt == -1) {

Modified: head/sys/vm/vm_fault.c
==
--- head/sys/vm/vm_fault.c  Wed Nov 20 02:20:27 2013(r258364)
+++ head/sys/vm/vm_fault.c  Wed Nov 20 08:45:26 2013(r258365)
@@ -1074,12 +1074,12 @@ vm_fault_quick_hold_pages(vm_map_t map, 
 {
vm_offset_t end, va;
vm_page_t *mp;
-   vm_size_t count;
+   int count;
boolean_t pmap_failed;
 
if (len == 0)
return (0);
-   end = round_page(addr + len);   
+   end = round_page(addr + len);
addr = trunc_page(addr);
 
/*
@@ -1088,9 +1088,9 @@ vm_fault_quick_hold_pages(vm_map_t map, 
if (addr < vm_map_min(map) || addr > end || end > vm_map_max(map))
return (-1);
 
-   count = howmany(end - addr, PAGE_SIZE);
-   if (count > max_count)
+   if (atop(end - addr) > max_count)
panic("vm_fault_quick_hold_pages: count > max_count");
+   count = atop(end - addr);
 
/*
 * Most likely, the physical pages are resident in the pmap, so it is
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258366 - head/sys/vm

2013-11-20 Thread Konstantin Belousov 
Author: kib
Date: Wed Nov 20 08:47:54 2013
New Revision: 258366
URL: http://svnweb.freebsd.org/changeset/base/258366

Log:
  Add assertions to cover all places in the wiring and unwiring code
  where MAP_ENTRY_IN_TRANSITION is set or cleared.
  
  Tested by:pho
  Reviewed by:  alc
  Sponsored by: The FreeBSD Foundation
  MFC after:1 week

Modified:
  head/sys/vm/vm_map.c

Modified: head/sys/vm/vm_map.c
==
--- head/sys/vm/vm_map.cWed Nov 20 08:45:26 2013(r258365)
+++ head/sys/vm/vm_map.cWed Nov 20 08:47:54 2013(r258366)
@@ -2288,6 +2288,9 @@ vm_map_unwire(vm_map_t map, vm_offset_t 
 * Mark the entry in case the map lock is released.  (See
 * above.)
 */
+   KASSERT((entry->eflags & MAP_ENTRY_IN_TRANSITION) == 0 &&
+   entry->wiring_thread == NULL,
+   ("owned map entry %p", entry));
entry->eflags |= MAP_ENTRY_IN_TRANSITION;
entry->wiring_thread = curthread;
/*
@@ -2356,7 +2359,9 @@ done:
}
}
KASSERT((entry->eflags & MAP_ENTRY_IN_TRANSITION) != 0,
-   ("vm_map_unwire: in-transition flag missing"));
+   ("vm_map_unwire: in-transition flag missing %p", entry));
+   KASSERT(entry->wiring_thread == curthread,
+   ("vm_map_unwire: alien wire %p", entry));
entry->eflags &= ~MAP_ENTRY_IN_TRANSITION;
entry->wiring_thread = NULL;
if (entry->eflags & MAP_ENTRY_NEEDS_WAKEUP) {
@@ -2456,6 +2461,9 @@ vm_map_wire(vm_map_t map, vm_offset_t st
 * Mark the entry in case the map lock is released.  (See
 * above.)
 */
+   KASSERT((entry->eflags & MAP_ENTRY_IN_TRANSITION) == 0 &&
+   entry->wiring_thread == NULL,
+   ("owned map entry %p", entry));
entry->eflags |= MAP_ENTRY_IN_TRANSITION;
entry->wiring_thread = curthread;
if ((entry->protection & (VM_PROT_READ | VM_PROT_EXECUTE)) == 0
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258367 - head/sys/vm

2013-11-20 Thread Konstantin Belousov 
Author: kib
Date: Wed Nov 20 09:03:48 2013
New Revision: 258367
URL: http://svnweb.freebsd.org/changeset/base/258367

Log:
  Vm map code performs clipping when map entry covers region which is
  larger than the operational region.  If the op region size is zero,
  clipping would create a zero-sized map entry.  The result is that vm
  map splay starts behaving inconsistently, sometimes returning
  zero-sized entry, sometimes the next (or previous) entry.
  
  One step further, it could result in e.g. vm_map_wire() setting
  MAP_ENTRY_IN_TRANSITION on the zero-sized entry, but failing to clear
  it in the done part.  The vm_map_delete() than hangs forever waiting
  for the flag removal.
  
  Verify for zero-length requests and act as if it is always successfull
  without performing any action on the address space.
  
  Diagnosed by: pho
  Tested by:pho (previous version)
  Reviewed by:  alc (previous version)
  Sponsored by: The FreeBSD Foundation
  MFC after:1 week

Modified:
  head/sys/vm/vm_map.c

Modified: head/sys/vm/vm_map.c
==
--- head/sys/vm/vm_map.cWed Nov 20 08:47:54 2013(r258366)
+++ head/sys/vm/vm_map.cWed Nov 20 09:03:48 2013(r258367)
@@ -1876,6 +1876,9 @@ vm_map_protect(vm_map_t map, vm_offset_t
struct ucred *cred;
vm_prot_t old_prot;
 
+   if (start == end)
+   return (KERN_SUCCESS);
+
vm_map_lock(map);
 
VM_MAP_RANGE_CHECK(map, start, end);
@@ -2030,12 +2033,16 @@ vm_map_madvise(
case MADV_AUTOSYNC:
case MADV_NOCORE:
case MADV_CORE:
+   if (start == end)
+   return (KERN_SUCCESS);
modify_map = 1;
vm_map_lock(map);
break;
case MADV_WILLNEED:
case MADV_DONTNEED:
case MADV_FREE:
+   if (start == end)
+   return (KERN_SUCCESS);
vm_map_lock_read(map);
break;
default:
@@ -2190,6 +2197,8 @@ vm_map_inherit(vm_map_t map, vm_offset_t
default:
return (KERN_INVALID_ARGUMENT);
}
+   if (start == end)
+   return (KERN_SUCCESS);
vm_map_lock(map);
VM_MAP_RANGE_CHECK(map, start, end);
if (vm_map_lookup_entry(map, start, &temp_entry)) {
@@ -,6 +2231,8 @@ vm_map_unwire(vm_map_t map, vm_offset_t 
int rv;
boolean_t need_wakeup, result, user_unwire;
 
+   if (start == end)
+   return (KERN_SUCCESS);
user_unwire = (flags & VM_MAP_WIRE_USER) ? TRUE : FALSE;
vm_map_lock(map);
VM_MAP_RANGE_CHECK(map, start, end);
@@ -2392,6 +2403,8 @@ vm_map_wire(vm_map_t map, vm_offset_t st
boolean_t fictitious, need_wakeup, result, user_wire;
vm_prot_t prot;
 
+   if (start == end)
+   return (KERN_SUCCESS);
prot = 0;
if (flags & VM_MAP_WIRE_WRITE)
prot |= VM_PROT_WRITE;
@@ -2833,6 +2846,8 @@ vm_map_delete(vm_map_t map, vm_offset_t 
vm_map_entry_t first_entry;
 
VM_MAP_ASSERT_LOCKED(map);
+   if (start == end)
+   return (KERN_SUCCESS);
 
/*
 * Find the start of the region, and clip it
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258387 - head/sys/dev/nand

2013-11-20 Thread Grzegorz Bernacki 
Author: gber
Date: Wed Nov 20 11:10:23 2013
New Revision: 258387
URL: http://svnweb.freebsd.org/changeset/base/258387

Log:
  Split raw reading/programming into smaller chunks to avoid allocating too
  big chunk of kernel memory. Validate size of data. Add error handling to
  avoid calling copyout() when data has not been read correctly.
  
  Reviewed by:zbb
  Reported by:x90c 
  MFC after:  2 days

Modified:
  head/sys/dev/nand/nand_cdev.c
  head/sys/dev/nand/nand_geom.c

Modified: head/sys/dev/nand/nand_cdev.c
==
--- head/sys/dev/nand/nand_cdev.c   Wed Nov 20 11:09:12 2013
(r258386)
+++ head/sys/dev/nand/nand_cdev.c   Wed Nov 20 11:10:23 2013
(r258387)
@@ -294,19 +294,39 @@ nand_ioctl(struct cdev *dev, u_long cmd,
 struct thread *td)
 {
struct nand_chip *chip;
+   struct chip_geom  *cg;
struct nand_oob_rw *oob_rw = NULL;
struct nand_raw_rw *raw_rw = NULL;
device_t nandbus;
+   size_t bufsize, len, raw_size;
+   off_t off;
uint8_t *buf = NULL;
int ret = 0;
uint8_t status;
 
chip = (struct nand_chip *)dev->si_drv1;
+   cg = &chip->chip_geom;
nandbus = device_get_parent(chip->dev);
 
if ((cmd == NAND_IO_RAW_READ) || (cmd == NAND_IO_RAW_PROG)) {
raw_rw = (struct nand_raw_rw *)data;
-   buf = malloc(raw_rw->len, M_NAND, M_WAITOK);
+   raw_size =  cg->pgs_per_blk * (cg->page_size + cg->oob_size);
+
+   /* Check if len is not bigger than chip size */
+   if (raw_rw->len > raw_size)
+   return (EFBIG);
+
+   /*
+* Do not ask for too much memory, in case of large transfers
+* read/write in 16-pages chunks
+*/
+   bufsize = 16 * (cg->page_size + cg->oob_size);
+   if (raw_rw->len < bufsize)
+   bufsize = raw_rw->len;
+
+   buf = malloc(bufsize, M_NAND, M_WAITOK);
+   len = raw_rw->len;
+   off = 0;
}
switch(cmd) {
case NAND_IO_ERASE:
@@ -335,19 +355,37 @@ nand_ioctl(struct cdev *dev, u_long cmd,
break;
 
case NAND_IO_RAW_PROG:
-   ret = copyin(raw_rw->data, buf, raw_rw->len);
-   if (ret)
-   break;
-   ret = nand_prog_pages_raw(chip, raw_rw->off, buf,
-   raw_rw->len);
+   while (len > 0) {
+   if (len < bufsize)
+   bufsize = len;
+   ret = copyin(raw_rw->data + off, buf, bufsize);
+   if (ret)
+   break;
+   ret = nand_prog_pages_raw(chip, raw_rw->off + off, buf,
+   bufsize);
+   if (ret)
+   break;
+   len -= bufsize;
+   off += bufsize;
+   }
break;
 
case NAND_IO_RAW_READ:
-   ret = nand_read_pages_raw(chip, raw_rw->off, buf,
-   raw_rw->len);
-   if (ret)
-   break;
-   ret = copyout(buf, raw_rw->data, raw_rw->len);
+   while (len > 0) {
+   if (len < bufsize)
+   bufsize = len;
+
+   ret = nand_read_pages_raw(chip, raw_rw->off + off, buf,
+   bufsize);
+   if (ret)
+   break;
+
+   ret = copyout(buf, raw_rw->data + off, bufsize);
+   if (ret)
+   break;
+   len -= bufsize;
+   off += bufsize;
+   }
break;
 
case NAND_IO_PAGE_STAT:

Modified: head/sys/dev/nand/nand_geom.c
==
--- head/sys/dev/nand/nand_geom.c   Wed Nov 20 11:09:12 2013
(r258386)
+++ head/sys/dev/nand/nand_geom.c   Wed Nov 20 11:10:23 2013
(r258387)
@@ -193,20 +193,41 @@ nand_ioctl(struct disk *ndisk, u_long cm
 struct thread *td)
 {
struct nand_chip *chip;
+   struct chip_geom  *cg;
struct nand_oob_rw *oob_rw = NULL;
struct nand_raw_rw *raw_rw = NULL;
device_t nandbus;
+   size_t bufsize, len, raw_size;
+   off_t off;
uint8_t *buf = NULL;
int ret = 0;
uint8_t status;
 
chip = (struct nand_chip *)ndisk->d_drv1;
+   cg = &chip->chip_geom;
nandbus = device_get_parent(chip->dev);
 
if ((cmd == NAND_IO_RAW_READ) || (cmd == NAND_IO_RAW_PROG)) {
raw_rw = (struct nand_raw_rw *)data;
-   buf = malloc(raw_rw->len, M_NAND, M_WAITOK

svn commit: r258388 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs

2013-11-20 Thread Andriy Gapon 
Author: avg
Date: Wed Nov 20 11:47:50 2013
New Revision: 258388
URL: http://svnweb.freebsd.org/changeset/base/258388

Log:
  MFV r258377: 4088 use after free in arc_release()
  
  illumos/illumos-gate@ccc22e130479b5bd7c0002267fee1e0602d3f772
  
  MFC after:5 days

Modified:
  head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c
Directory Properties:
  head/sys/cddl/contrib/opensolaris/   (props changed)

Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c
==
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c   Wed Nov 20 
11:10:23 2013(r258387)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c   Wed Nov 20 
11:47:50 2013(r258388)
@@ -3540,6 +3540,7 @@ arc_release(arc_buf_t *buf, void *tag)
if (l2hdr) {
mutex_enter(&l2arc_buflist_mtx);
hdr->b_l2hdr = NULL;
+   list_remove(l2hdr->b_dev->l2ad_buflist, hdr);
}
buf_size = hdr->b_size;
 
@@ -3627,7 +3628,6 @@ arc_release(arc_buf_t *buf, void *tag)
ARCSTAT_INCR(arcstat_l2_asize, -l2hdr->b_asize);
trim_map_free(l2hdr->b_dev->l2ad_vdev, l2hdr->b_daddr,
hdr->b_size, 0);
-   list_remove(l2hdr->b_dev->l2ad_buflist, hdr);
kmem_free(l2hdr, sizeof (l2arc_buf_hdr_t));
ARCSTAT_INCR(arcstat_l2_size, -buf_size);
mutex_exit(&l2arc_buflist_mtx);
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258389 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs

2013-11-20 Thread Andriy Gapon 
Author: avg
Date: Wed Nov 20 11:52:32 2013
New Revision: 258389
URL: http://svnweb.freebsd.org/changeset/base/258389

Log:
  MFV r258378: 4089 NULL pointer dereference in arc_read()
  
  illumos/illumos-gate@57815f6b95a743697e148327725b7f568e75e6ea
  
  Tested by:adrian
  MFC after:4 days

Modified:
  head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c
Directory Properties:
  head/sys/cddl/contrib/opensolaris/   (props changed)

Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c
==
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c   Wed Nov 20 
11:47:50 2013(r258388)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c   Wed Nov 20 
11:52:32 2013(r258389)
@@ -20,9 +20,9 @@
  */
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright 2011 Nexenta Systems, Inc.  All rights reserved.
  * Copyright (c) 2013 by Delphix. All rights reserved.
  * Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
+ * Copyright 2013 Nexenta Systems, Inc.  All rights reserved.
  */
 
 /*
@@ -3173,6 +3173,8 @@ top:
vdev_t *vd = NULL;
uint64_t addr = 0;
boolean_t devw = B_FALSE;
+   enum zio_compress b_compress = ZIO_COMPRESS_OFF;
+   uint64_t b_asize = 0;
 
if (hdr == NULL) {
/* this block is not in the cache */
@@ -3242,10 +3244,12 @@ top:
hdr->b_acb = acb;
hdr->b_flags |= ARC_IO_IN_PROGRESS;
 
-   if (HDR_L2CACHE(hdr) && hdr->b_l2hdr != NULL &&
+   if (hdr->b_l2hdr != NULL &&
(vd = hdr->b_l2hdr->b_dev->l2ad_vdev) != NULL) {
devw = hdr->b_l2hdr->b_dev->l2ad_writing;
addr = hdr->b_l2hdr->b_daddr;
+   b_compress = hdr->b_l2hdr->b_compress;
+   b_asize = hdr->b_l2hdr->b_asize;
/*
 * Lock out device removal.
 */
@@ -3296,7 +3300,7 @@ top:
cb->l2rcb_bp = *bp;
cb->l2rcb_zb = *zb;
cb->l2rcb_flags = zio_flags;
-   cb->l2rcb_compress = hdr->b_l2hdr->b_compress;
+   cb->l2rcb_compress = b_compress;
 
ASSERT(addr >= VDEV_LABEL_START_SIZE &&
addr + size < vd->vdev_psize -
@@ -3308,8 +3312,7 @@ top:
 * Issue a null zio if the underlying buffer
 * was squashed to zero size by compression.
 */
-   if (hdr->b_l2hdr->b_compress ==
-   ZIO_COMPRESS_EMPTY) {
+   if (b_compress == ZIO_COMPRESS_EMPTY) {
rzio = zio_null(pio, spa, vd,
l2arc_read_done, cb,
zio_flags | ZIO_FLAG_DONT_CACHE |
@@ -3318,8 +3321,8 @@ top:
ZIO_FLAG_DONT_RETRY);
} else {
rzio = zio_read_phys(pio, vd, addr,
-   hdr->b_l2hdr->b_asize,
-   buf->b_data, ZIO_CHECKSUM_OFF,
+   b_asize, buf->b_data,
+   ZIO_CHECKSUM_OFF,
l2arc_read_done, cb, priority,
zio_flags | ZIO_FLAG_DONT_CACHE |
ZIO_FLAG_CANFAIL |
@@ -3328,8 +3331,7 @@ top:
}
DTRACE_PROBE2(l2arc__read, vdev_t *, vd,
zio_t *, rzio);
-   ARCSTAT_INCR(arcstat_l2_read_bytes,
-   hdr->b_l2hdr->b_asize);
+   ARCSTAT_INCR(arcstat_l2_read_bytes, b_asize);
 
if (*arc_flags & ARC_NOWAIT) {
zio_nowait(rzio);
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258390 - head/tools/tools/umastat

2013-11-20 Thread Alexander Motin 
Author: mav
Date: Wed Nov 20 12:32:34 2013
New Revision: 258390
URL: http://svnweb.freebsd.org/changeset/base/258390

Log:
  Print some more flags.

Modified:
  head/tools/tools/umastat/umastat.c

Modified: head/tools/tools/umastat/umastat.c
==
--- head/tools/tools/umastat/umastat.c  Wed Nov 20 11:52:32 2013
(r258389)
+++ head/tools/tools/umastat/umastat.c  Wed Nov 20 12:32:34 2013
(r258390)
@@ -117,6 +117,9 @@ static const struct flaginfo {
u_int32_tfi_flag;
const char  *fi_name;
 } flaginfo[] = {
+   { UMA_ZFLAG_MULTI, "multi" },
+   { UMA_ZFLAG_DRAINING, "draining" },
+   { UMA_ZFLAG_BUCKET, "bucket" },
{ UMA_ZFLAG_INTERNAL, "internal" },
{ UMA_ZFLAG_FULL, "full" },
{ UMA_ZFLAG_CACHEONLY, "cacheonly" },
@@ -132,6 +135,10 @@ static const struct flaginfo {
{ UMA_ZONE_SECONDARY, "secondary" },
{ UMA_ZONE_REFCNT, "refcnt" },
{ UMA_ZONE_MAXBUCKET, "maxbucket" },
+   { UMA_ZONE_CACHESPREAD, "cachespread" },
+   { UMA_ZONE_VTOSLAB, "vtoslab" },
+   { UMA_ZONE_NODUMP, "nodump" },
+   { UMA_ZONE_PCPU, "pcpu" },
 };
 static const int flaginfo_count = sizeof(flaginfo) / sizeof(struct flaginfo);
 
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258391 - head/sys/sys

2013-11-20 Thread Gleb Smirnoff 
Author: glebius
Date: Wed Nov 20 13:22:22 2013
New Revision: 258391
URL: http://svnweb.freebsd.org/changeset/base/258391

Log:
  Fix comment after r250551.

Modified:
  head/sys/sys/bufobj.h

Modified: head/sys/sys/bufobj.h
==
--- head/sys/sys/bufobj.h   Wed Nov 20 12:32:34 2013(r258390)
+++ head/sys/sys/bufobj.h   Wed Nov 20 13:22:22 2013(r258391)
@@ -63,7 +63,7 @@ extern struct buf_ops buf_ops_bio;
 
 TAILQ_HEAD(buflists, buf);
 
-/* A Buffer splay list */
+/* A Buffer list & trie */
 struct bufv {
struct buflists bv_hd;  /* Sorted blocklist */
struct pctrie   bv_root;/* Buf trie */
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r257831 - head/sys/cam/scsi

2013-11-20 Thread Slawa Olhovchenkov 
On Thu, Nov 07, 2013 at 11:21:52PM +, Steven Hartland wrote:

> Author: smh
> Date: Thu Nov  7 23:21:52 2013
> New Revision: 257831
> URL: http://svnweb.freebsd.org/changeset/base/257831
> 
> Log:
>   Corrected definition for old_rate to match d_rotation_rate
>   
>   MFC after:  2 Days
>   X-MFC-With: r256956

MFC?
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r257831 - head/sys/cam/scsi

2013-11-20 Thread Steven Hartland 
- Original Message - 
From: "Slawa Olhovchenkov" 

To: "Steven Hartland" 
Cc: ; ; 

Sent: Wednesday, November 20, 2013 2:05 PM
Subject: Re: svn commit: r257831 - head/sys/cam/scsi



On Thu, Nov 07, 2013 at 11:21:52PM +, Steven Hartland wrote:


Author: smh
Date: Thu Nov  7 23:21:52 2013
New Revision: 257831
URL: http://svnweb.freebsd.org/changeset/base/257831

Log:
  Corrected definition for old_rate to match d_rotation_rate
  
  MFC after: 2 Days

  X-MFC-With: r256956


MFC?


I have stable/9 done locally just pending me fixing my svn install
so it uses the FreeBSD fields again which it appears to have lost
for some unknown reason :(


This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 


In the event of misdirection, illegible or incomplete transmission please 
telephone +44 845 868 1337
or return the E.mail to postmas...@multiplay.co.uk.

___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r257831 - head/sys/cam/scsi

2013-11-20 Thread Slawa Olhovchenkov 
On Wed, Nov 20, 2013 at 02:17:32PM -, Steven Hartland wrote:

> - Original Message - 
> From: "Slawa Olhovchenkov" 
> To: "Steven Hartland" 
> Cc: ; ; 
> 
> Sent: Wednesday, November 20, 2013 2:05 PM
> Subject: Re: svn commit: r257831 - head/sys/cam/scsi
> 
> 
> > On Thu, Nov 07, 2013 at 11:21:52PM +, Steven Hartland wrote:
> > 
> >> Author: smh
> >> Date: Thu Nov  7 23:21:52 2013
> >> New Revision: 257831
> >> URL: http://svnweb.freebsd.org/changeset/base/257831
> >> 
> >> Log:
> >>   Corrected definition for old_rate to match d_rotation_rate
> >>   
> >>   MFC after: 2 Days
> >>   X-MFC-With: r256956
> > 
> > MFC?
> 
> I have stable/9 done locally just pending me fixing my svn install
> so it uses the FreeBSD fields again which it appears to have lost
> for some unknown reason :(

Can you fix this before next release cycle of 10.0?
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r257831 - head/sys/cam/scsi

2013-11-20 Thread Steven Hartland 
- Original Message - 
From: "Slawa Olhovchenkov" 

> On Thu, Nov 07, 2013 at 11:21:52PM +, Steven Hartland wrote:
> 
>> Author: smh

>> Date: Thu Nov  7 23:21:52 2013
>> New Revision: 257831
>> URL: http://svnweb.freebsd.org/changeset/base/257831
>> 
>> Log:

>>   Corrected definition for old_rate to match d_rotation_rate
>>   
>>   MFC after: 2 Days

>>   X-MFC-With: r256956
> 
> MFC?


I have stable/9 done locally just pending me fixing my svn install
so it uses the FreeBSD fields again which it appears to have lost
for some unknown reason :(


Can you fix this before next release cycle of 10.0?


As the original change didnt make the 10 cut off it wont be in 10
I'm afraid.

I did request consideration of re@ for it due the confusion about
the change request process but it was denied :(

If you feel strongly it should be in 10 you might want to ask re@
for reconsideration.

   Regards
   Steve


This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 


In the event of misdirection, illegible or incomplete transmission please 
telephone +44 845 868 1337
or return the E.mail to postmas...@multiplay.co.uk.

___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258392 - head/sys/arm/arm

2013-11-20 Thread Ian Lepore 
Author: ian
Date: Wed Nov 20 15:53:50 2013
New Revision: 258392
URL: http://svnweb.freebsd.org/changeset/base/258392

Log:
  Call cpu_setup() immediately after the page tables are installed.  This
  enables data cache and other chip-specific features.  It was previously
  done via an early SYSINIT, but it was being done after pmap and vm setup,
  and those setups need to use mutexes.  On some modern ARM platforms,
  the ldrex/strex instructions that implement mutexes require the data cache
  to be enabled.
  
  A nice side effect of enabling caching earlier is that it eliminates the
  multi-second pause that used to happen early in boot while physical memory
  and pmap and vm were being set up.  On boards with 1 GB or more of ram
  this pause was very noticible, sometimes 5-6 seconds.
  
  PR:   arm/183740

Modified:
  head/sys/arm/arm/machdep.c

Modified: head/sys/arm/arm/machdep.c
==
--- head/sys/arm/arm/machdep.c  Wed Nov 20 13:22:22 2013(r258391)
+++ head/sys/arm/arm/machdep.c  Wed Nov 20 15:53:50 2013(r258392)
@@ -361,7 +361,6 @@ cpu_startup(void *dummy)
 #endif
 #endif
 
-   cpu_setup("");
identify_arm_cpu();
 
printf("real memory  = %ju (%ju MB)\n", (uintmax_t)ptoa(physmem),
@@ -1431,6 +1430,12 @@ initarm(struct arm_boot_params *abp)
cpu_domains(DOMAIN_CLIENT << (PMAP_DOMAIN_KERNEL * 2));
 
/*
+* Now that proper page tables are installed, call cpu_setup() to enable
+* instruction and data caches and other chip-specific features.
+*/
+   cpu_setup("");
+
+   /*
 * Only after the SOC registers block is mapped we can perform device
 * tree fixups, as they may attempt to read parameters from hardware.
 */
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258393 - head/sys/arm/conf

2013-11-20 Thread Ian Lepore 
Author: ian
Date: Wed Nov 20 16:42:01 2013
New Revision: 258393
URL: http://svnweb.freebsd.org/changeset/base/258393

Log:
  Add USB_HOST_ALIGN=64; the cache line size on the am335x is 64 bytes.

Modified:
  head/sys/arm/conf/BEAGLEBONE

Modified: head/sys/arm/conf/BEAGLEBONE
==
--- head/sys/arm/conf/BEAGLEBONEWed Nov 20 15:53:50 2013
(r258392)
+++ head/sys/arm/conf/BEAGLEBONEWed Nov 20 16:42:01 2013
(r258393)
@@ -104,6 +104,7 @@ device  gpio
 
 # USB support
 device usb
+optionsUSB_HOST_ALIGN=64   # Cacheline size is 64 on AM335x.
 optionsUSB_DEBUG
 #options   USB_REQ_DEBUG
 #options   USB_VERBOSE
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258395 - head/usr.bin/uname

2013-11-20 Thread Peter Wemm 
Author: peter
Date: Wed Nov 20 17:48:38 2013
New Revision: 258395
URL: http://svnweb.freebsd.org/changeset/base/258395

Log:
  Change introduction history.

Modified:
  head/usr.bin/uname/uname.1

Modified: head/usr.bin/uname/uname.1
==
--- head/usr.bin/uname/uname.1  Wed Nov 20 17:46:23 2013(r258394)
+++ head/usr.bin/uname/uname.1  Wed Nov 20 17:48:38 2013(r258395)
@@ -28,7 +28,7 @@
 .\"@(#)uname.1 8.3 (Berkeley) 4/8/94
 .\" $FreeBSD$
 .\"
-.Dd January 26, 2010
+.Dd November 20, 2013
 .Dt UNAME 1
 .Os
 .Sh NAME
@@ -128,4 +128,4 @@ The
 and
 .Fl U
 extension flags appeared in
-.Fx 11.0 .
+.Fx 10.0 .
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258397 - head/sys/fs/pseudofs

2013-11-20 Thread Konstantin Belousov 
Author: kib
Date: Wed Nov 20 19:41:00 2013
New Revision: 258397
URL: http://svnweb.freebsd.org/changeset/base/258397

Log:
  Redo r258088 to avoid relying on signed arithmetic overflow, since
  compiler interprets this as an undefined behaviour.  Instead, ensure
  that the sum of uio_offset and uio_resid is below OFF_MAX using the
  operation which cannot overflow.
  
  Reported and tested by:   pho
  Discussed with:   bde
  Approved by:  des (pseudofs maintainer)
  Sponsored by: The FreeBSD Foundation
  MFC after:1 week

Modified:
  head/sys/fs/pseudofs/pseudofs_vnops.c

Modified: head/sys/fs/pseudofs/pseudofs_vnops.c
==
--- head/sys/fs/pseudofs/pseudofs_vnops.c   Wed Nov 20 18:58:07 2013
(r258396)
+++ head/sys/fs/pseudofs/pseudofs_vnops.c   Wed Nov 20 19:41:00 2013
(r258397)
@@ -616,8 +616,7 @@ pfs_read(struct vop_read_args *va)
struct proc *proc;
struct sbuf *sb = NULL;
int error, locked;
-   off_t offset;
-   ssize_t buflen, resid;
+   off_t buflen;
 
PFS_TRACE(("%s", pn->pn_name));
pfs_assert_not_owned(pn);
@@ -654,16 +653,12 @@ pfs_read(struct vop_read_args *va)
goto ret;
}
 
-   resid = uio->uio_resid;
-   offset = uio->uio_offset;
-   buflen = offset + resid;
-
-   /* beaucoup sanity checks so we don't ask for bogus allocation */
-   if (resid < 0 || buflen < offset || buflen < resid ||
-   buflen >= INT_MAX) {
+   if (uio->uio_resid < 0 || uio->uio_offset < 0 ||
+   uio->uio_resid > OFF_MAX - uio->uio_offset) {
error = EINVAL;
goto ret;
}
+   buflen = uio->uio_offset + uio->uio_resid;
if (buflen > MAXPHYS)
buflen = MAXPHYS;
 
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Julian Elischer 

On 11/19/13, 3:04 AM, Robert Watson wrote:

On Mon, 18 Nov 2013, George V. Neville-Neil wrote:

 Allow ethernet drivers to pass in packets connected via the 
nextpkt pointer.
 Handling packets in this way allows drivers to amortize work 
during packet reception.


 Submitted by:Vijay Singh
 Sponsored by:NetApp


Currently, it is quite easy to make mistakes regarding individual 
mbuf chains vs. lists of mbuf chains.  This leads me to wonder 
whether a new type, perhaps simply constructed on the stack before 
passing in, should be used for KPIs that accept lists of packets. E.g.,


/*
 * This structure is almost always allocated on a caller stack, so
 * cannot itself be queued without memory allocation in most cases.
 */
struct mbuf_queue {
struct mbuf*mq_head;
};


It's hard to believe that we don't have a structure around already 
that we can't use. With Luigi's comment, I wonder that there isn't an 
mbuf_list structure already we can just steal. it could almost be the 
current interface input queue structure.



___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Julian Elischer 

On 11/19/13, 3:04 AM, Robert Watson wrote:

On Mon, 18 Nov 2013, George V. Neville-Neil wrote:

 Allow ethernet drivers to pass in packets connected via the 
nextpkt pointer.
 Handling packets in this way allows drivers to amortize work 
during packet reception.


 Submitted by:Vijay Singh
 Sponsored by:NetApp


Currently, it is quite easy to make mistakes regarding individual 
mbuf chains vs. lists of mbuf chains.  This leads me to wonder 
whether a new type, perhaps simply constructed on the stack before 
passing in, should be used for KPIs that accept lists of packets. E.g.,


/*
 * This structure is almost always allocated on a caller stack, so
 * cannot itself be queued without memory allocation in most cases.
 */
struct mbuf_queue {
struct mbuf*mq_head;
};

int
ether_input(struct ifnet *ifp, struct mbuf_queue *m)
{

...
}

or separate entrypoints, old and and new


...
struct mbuf_queue mq = { m };

return (ether_input(ifp, &mq));
...

That way the compiler can help us figure out where we expect an 
individual packet but have accidentally leaked a queue.  Functions 
that accept only a single packet could also more agressively assert 
that m->m_nextpkt is NULL:


M_ASSERT_ONEPACKET(m);

Robert








___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Robert N. M. Watson 

On 20 Nov 2013, at 20:11, Julian Elischer  wrote:

>> Currently, it is quite easy to make mistakes regarding individual mbuf 
>> chains vs. lists of mbuf chains.  This leads me to wonder whether a new 
>> type, perhaps simply constructed on the stack before passing in, should be 
>> used for KPIs that accept lists of packets. E.g.,
>> 
>>/*
>> * This structure is almost always allocated on a caller stack, so
>> * cannot itself be queued without memory allocation in most cases.
>> */
>>struct mbuf_queue {
>>struct mbuf*mq_head;
>>};
>> 
>> 
> It's hard to believe that we don't have a structure around already that we 
> can't use. With Luigi's comment, I wonder that there isn't an mbuf_list 
> structure already we can just steal. it could almost be the current interface 
> input queue structure.

The exact details don't matter; what does matter is that we make it as easy as 
possible to detect mistakes using the compiler (e.g., queue passed where mbuf 
expected, or vice versa) and that those situations we can't check statically, 
we try to check dynamically (multi-entry queue passed where mbuf expected). In 
the past we've had bugs along similar lines, where code expects m->m_nextpkt to 
be NULL when it isn't leading to very unhappy times in socket buffers, etc. 
Ideally what we did would have no expense at runtime unless debugging features 
were turned on (INVARIANTS).

Robert
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258399 - head/sys/crypto/aesni

2013-11-20 Thread John-Mark Gurney 
Author: jmg
Date: Wed Nov 20 20:25:27 2013
New Revision: 258399
URL: http://svnweb.freebsd.org/changeset/base/258399

Log:
  flag that the aesni driver is sync...  This means we don't waste a
  context switch just to call the done callback...  On my machine, this
  improves geli/gzero decrypt performance by ~27% from 550MB/sec to
  ~700MB/sec...
  
  MFC after:3 days

Modified:
  head/sys/crypto/aesni/aesni.c

Modified: head/sys/crypto/aesni/aesni.c
==
--- head/sys/crypto/aesni/aesni.c   Wed Nov 20 20:24:59 2013
(r258398)
+++ head/sys/crypto/aesni/aesni.c   Wed Nov 20 20:25:27 2013
(r258399)
@@ -92,7 +92,8 @@ aesni_attach(device_t dev)
sc = device_get_softc(dev);
TAILQ_INIT(&sc->sessions);
sc->sid = 1;
-   sc->cid = crypto_get_driverid(dev, CRYPTOCAP_F_HARDWARE);
+   sc->cid = crypto_get_driverid(dev,
+   CRYPTOCAP_F_HARDWARE|CRYPTOCAP_F_SYNC);
if (sc->cid < 0) {
device_printf(dev, "Could not get crypto driver id.\n");
return (ENOMEM);
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r255736 - in head: share/man/man4 sys/amd64/conf sys/conf sys/dev/bxe sys/i386/conf sys/modules/bxe

2013-11-20 Thread hiren panchasara 
+ Eric

On Fri, Sep 20, 2013 at 1:18 PM, David Christensen  wrote:
> Author: davidch
> Date: Fri Sep 20 20:18:49 2013
> New Revision: 255736
> URL: http://svnweb.freebsd.org/changeset/base/255736
>
> Log:
>   Substantial rewrite of bxe(4) to add support for the BCM57712 and
>   BCM578XX controllers.
>
>   Approved by:  re
>   MFC after:4 weeks

David/Eric,

Are you guys planning to mfc this back to 9?

cheers,
Hiren
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258400 - in head/usr.sbin/bsdconfig: . includes includes/include share

2013-11-20 Thread Devin Teske 
Author: dteske
Date: Wed Nov 20 20:37:21 2013
New Revision: 258400
URL: http://svnweb.freebsd.org/changeset/base/258400

Log:
  Add new `includes' module for exploring the bsdconfig(8) API.

Added:
  head/usr.sbin/bsdconfig/includes/
  head/usr.sbin/bsdconfig/includes/INDEX   (contents, props changed)
  head/usr.sbin/bsdconfig/includes/Makefile   (contents, props changed)
  head/usr.sbin/bsdconfig/includes/USAGE   (contents, props changed)
  head/usr.sbin/bsdconfig/includes/include/
  head/usr.sbin/bsdconfig/includes/include/Makefile   (contents, props changed)
  head/usr.sbin/bsdconfig/includes/include/messages.subr   (contents, props 
changed)
  head/usr.sbin/bsdconfig/includes/includes   (contents, props changed)
Modified:
  head/usr.sbin/bsdconfig/Makefile
  head/usr.sbin/bsdconfig/share/variable.subr

Modified: head/usr.sbin/bsdconfig/Makefile
==
--- head/usr.sbin/bsdconfig/MakefileWed Nov 20 20:25:27 2013
(r258399)
+++ head/usr.sbin/bsdconfig/MakefileWed Nov 20 20:37:21 2013
(r258400)
@@ -6,6 +6,7 @@ SUBDIR= console \
dot \
examples \
include \
+   includes \
mouse \
networking \
packages \

Added: head/usr.sbin/bsdconfig/includes/INDEX
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/usr.sbin/bsdconfig/includes/INDEX  Wed Nov 20 20:37:21 2013
(r258400)
@@ -0,0 +1,56 @@
+# Copyright (c) 2013 Devin Teske
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+# 1. Redistributions of source code must retain the above copyright
+#notice, this list of conditions and the following disclaimer.
+# 2. Redistributions in binary form must reproduce the above copyright
+#notice, this list of conditions and the following disclaimer in the
+#documentation and/or other materials provided with the distribution.
+#
+# THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
+# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+# ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
+# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+# SUCH DAMAGE.
+#
+# $FreeBSD$
+
+#
+# Title that will be shown in the bsdconfig menu.
+#
+menu_title=""
+
+#
+# A short descriptive line shown at the bottom of the bsdconfig menu. keep it
+# short because any line longer than the terminal width will be truncated.
+#
+menu_help=""
+
+#
+# Two-part variable that defines an action to take when `keyword' is passed on
+# a bsdconfig command line. Variable takes the form "keyword|command" and
+# multiple occurrences of the variable (with different `keyword's, or different
+# `keyword's AND `command's) are allowed.  If `command' begins with a '/' then
+# the full path to the program is needed. If `command' begins with anything
+# else it is a path relative to the directory this INDEX file is in. `keyword'
+# can be i18n'ed but `command' is the name of a script.
+#
+menu_selection="includes|includes"
+
+#
+#  Items below this line do NOT need i18n translation 
+#
+# Name of the program to be run when this menu choice is selected. If it begins
+# with a '/' then the full path to the program is needed. If it begins with
+# anything else it is a path relative to the directory this INDEX file is in.
+#
+menu_program=""

Added: head/usr.sbin/bsdconfig/includes/Makefile
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/usr.sbin/bsdconfig/includes/Makefile   Wed Nov 20 20:37:21 2013
(r258400)
@@ -0,0 +1,16 @@
+# $FreeBSD$
+
+NO_OBJ=
+
+SUBDIR= include
+
+FILESDIR=  ${LIBEXECDIR}/bsdconfig/includes
+FILES= INDEX USAGE
+
+SCRIPTSDIR=${FILESDIR}
+SCRIPTS=   includes
+
+beforeinstall:
+   mkdir -p ${DESTDIR}${FILESDIR}
+
+.include 

Added: head/usr.sbin/bsdconfig/includes/USAGE
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/usr.sbin/bsdconfig/includes/USAGE  Wed Nov 20 20:37:21 2013
(r258400)
@@ -0,0 +1,66 @@
+# Copyright (c) 2013 Devin Teske
+# All ri

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Luigi Rizzo 
On Wed, Nov 20, 2013 at 12:11:20PM -0800, Julian Elischer wrote:
> On 11/19/13, 3:04 AM, Robert Watson wrote:
> > On Mon, 18 Nov 2013, George V. Neville-Neil wrote:
> >
> >>  Allow ethernet drivers to pass in packets connected via the 
> >> nextpkt pointer.
> >>  Handling packets in this way allows drivers to amortize work 
> >> during packet reception.
> >>
> >>  Submitted by:Vijay Singh
> >>  Sponsored by:NetApp
> >
> > Currently, it is quite easy to make mistakes regarding individual 
> > mbuf chains vs. lists of mbuf chains.  This leads me to wonder 
> > whether a new type, perhaps simply constructed on the stack before 
> > passing in, should be used for KPIs that accept lists of packets. E.g.,
> >
> > /*
> >  * This structure is almost always allocated on a caller stack, so
> >  * cannot itself be queued without memory allocation in most cases.
> >  */
> > struct mbuf_queue {
> > struct mbuf*mq_head;
> > };
> >
> >
> It's hard to believe that we don't have a structure around already 
> that we can't use. With Luigi's comment, I wonder that there isn't an 
> mbuf_list structure already we can just steal. it could almost be the 
> current interface input queue structure.

The fact is that the structure is so simple that every piece of code
seems to have its own implementation.

The closest thing is "struct ifqueue" in sys/net/ifq.h, but it
has some unnecessary fields (e.g. a struct mtx and a few counters).

sys/sys/sockbuf.h has a queue embedded in struct sockbuf
sys/netinet/tcp_lro.h has one in struct lro_entry

sctp and dummynet redefine their own queues, as well as a number
of device drivers (dev/xen/netfrnt/mbufq.h, cxgb,  ...)

I am not sure I see anything worth reusing.

cheers
luigi

___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r255736 - in head: share/man/man4 sys/amd64/conf sys/conf sys/dev/bxe sys/i386/conf sys/modules/bxe

2013-11-20 Thread Eric Davis 
Hi Hiren,

Yes I am planning on it.

- eric


On Wed, Nov 20, 2013 at 12:27 PM, hiren panchasara wrote:

> + Eric
>
> On Fri, Sep 20, 2013 at 1:18 PM, David Christensen 
> wrote:
> > Author: davidch
> > Date: Fri Sep 20 20:18:49 2013
> > New Revision: 255736
> > URL: http://svnweb.freebsd.org/changeset/base/255736
> >
> > Log:
> >   Substantial rewrite of bxe(4) to add support for the BCM57712 and
> >   BCM578XX controllers.
> >
> >   Approved by:  re
> >   MFC after:4 weeks
>
> David/Eric,
>
> Are you guys planning to mfc this back to 9?
>
> cheers,
> Hiren
>
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258401 - head/usr.sbin/bsdconfig/includes

2013-11-20 Thread Devin Teske 
Author: dteske
Date: Wed Nov 20 21:05:33 2013
New Revision: 258401
URL: http://svnweb.freebsd.org/changeset/base/258401

Log:
  Add a `-d' flag for printing the description of each function.

Modified:
  head/usr.sbin/bsdconfig/includes/USAGE
  head/usr.sbin/bsdconfig/includes/includes

Modified: head/usr.sbin/bsdconfig/includes/USAGE
==
--- head/usr.sbin/bsdconfig/includes/USAGE  Wed Nov 20 20:37:21 2013
(r258400)
+++ head/usr.sbin/bsdconfig/includes/USAGE  Wed Nov 20 21:05:33 2013
(r258401)
@@ -28,6 +28,7 @@ Usage: bsdconfig @PROGRAM_NAME@ [OPTIONS
 
 OPTIONS:
-aAlways use color even when output is not to a terminal.
+   -dPrint the description for each function selected.
-fShow functions for selected includes.
-F pattern
  If `-f', only print functions matching pattern. Without `-f'
@@ -64,3 +65,7 @@ EXAMPLES:
bsdconfig @PROGRAM_NAME@ -F show common
 
NB: The `.subr' suffix on the end of the include is optional.
+
+   Show descriptions of each of the `show' functions:
+
+   bsdconfig @PROGRAM_NAME@ -dfF show

Modified: head/usr.sbin/bsdconfig/includes/includes
==
--- head/usr.sbin/bsdconfig/includes/includes   Wed Nov 20 20:37:21 2013
(r258400)
+++ head/usr.sbin/bsdconfig/includes/includes   Wed Nov 20 21:05:33 2013
(r258401)
@@ -29,7 +29,7 @@
  INCLUDES
 
 # Prevent common.subr from auto initializing debugging (this is not an inter-
-# active utility that requires debugging).
+# active utility that requires debugging; also `-d' has been repurposed).
 #
 DEBUG_SELF_INITIALIZE=NO
 
@@ -50,6 +50,7 @@ ipgm=$( f_index_menusel_keyword $BSDCFG_
 # Options
 #
 USE_COLOR=1
+SHOW_DESC=
 SHOW_FUNCS=
 FUNC_PATTERN=
 
@@ -64,20 +65,33 @@ show_include()
local file="${1#./}"
 
local pattern="${FUNC_PATTERN:-.*}"
-   output=$( awk -v use_color=${USE_COLOR:-0} -v re="$pattern" '
+   output=$( awk \
+   -v use_color=${USE_COLOR:-0} \
+   -v re="$pattern" \
+   -v show_desc=${SHOW_DESC:-0} '
/^$/,/^#/ {
if ($0 ~ /^# f_/) {
if (!match($2, re)) next
if (use_color)
-   printf " %s%s%s\n",
+   printf "+%s%s%s\n",
   substr($0, 2, RSTART),
   substr($0, 2 + RSTART, RLENGTH),
   substr($0, 2 + RSTART + RLENGTH)
else
print substr($0, 2)
-   print_more = substr($0, length($0)) == "\\"
+   if (show_desc)
+   print_more = 1
+   else
+   print_more = substr($0, length($0)) == "\\"
}
-   while (print_more) {
+   if (show_desc && print_more) {
+   getline
+   while ($0 ~ /^#/) {
+   print substr($0, 2)
+   getline
+   }
+   print_more = 0
+   } else while (print_more) {
getline
print substr($0, 2)
print_more = substr($0, length($0)) == "\\"
@@ -89,10 +103,10 @@ show_include()
return $SUCCESS
fi
if [ "$FUNC_PATTERN" ]; then
-   printf "$msg_functions_in_matching\n" \
+   printf ">>> $msg_functions_in_matching\n" \
   "$file" "$FUNC_PATTERN"
else
-   printf "$msg_functions_in\n" "$file"
+   printf ">>> $msg_functions_in\n" "$file"
fi
echo "$output"
echo # blank line to simplify awk(1)-based reparse
@@ -110,9 +124,10 @@ show_include()
 #
 # Process command-line arguments
 #
-while getopts afF:hn flag; do
+while getopts adfF:hn flag; do
case "$flag" in
a) USE_COLOR=1 ;;
+   d) SHOW_DESC=1 ;;
f) SHOW_FUNCS=1 ;;
F) FUNC_PATTERN="$OPTARG" ;;
n) USE_COLOR= ;;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258402 - head/sys/ufs/ffs

2013-11-20 Thread John-Mark Gurney 
Author: jmg
Date: Wed Nov 20 21:16:53 2013
New Revision: 258402
URL: http://svnweb.freebsd.org/changeset/base/258402

Log:
  fix a use after free, jsegdep_merge will free wk, avoid the next check...
  
  CID:  1006098
  Sponsored by: Imaginary Forces
  Reviewed by:  mckusick
  MFC after:1 week

Modified:
  head/sys/ufs/ffs/ffs_softdep.c

Modified: head/sys/ufs/ffs/ffs_softdep.c
==
--- head/sys/ufs/ffs/ffs_softdep.c  Wed Nov 20 21:05:33 2013
(r258401)
+++ head/sys/ufs/ffs/ffs_softdep.c  Wed Nov 20 21:16:53 2013
(r258402)
@@ -1096,7 +1096,7 @@ jwork_move(dst, src)
LIST_FOREACH_SAFE(wk, dst, wk_list, wkn) {
if (wk->wk_type == D_JSEGDEP)
jsegdep = jsegdep_merge(WK_JSEGDEP(wk), jsegdep);
-   if (wk->wk_type == D_FREEDEP)
+   else if (wk->wk_type == D_FREEDEP)
freedep = freedep_merge(WK_FREEDEP(wk), freedep);
}
 
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258403 - head/sys/ufs/ffs

2013-11-20 Thread John-Mark Gurney 
Author: jmg
Date: Wed Nov 20 21:21:29 2013
New Revision: 258403
URL: http://svnweb.freebsd.org/changeset/base/258403

Log:
  fix white space...
  
  MFC after:1 week

Modified:
  head/sys/ufs/ffs/softdep.h

Modified: head/sys/ufs/ffs/softdep.h
==
--- head/sys/ufs/ffs/softdep.h  Wed Nov 20 21:16:53 2013(r258402)
+++ head/sys/ufs/ffs/softdep.h  Wed Nov 20 21:21:29 2013(r258403)
@@ -132,7 +132,7 @@
 #defineINPROGRESS  0x001000 /* dirrem, freeblks, freefrag, 
freefile only */
 #defineUFS1FMT 0x002000 /* indirdep only */
 #defineEXTDATA 0x004000 /* allocdirect only */
-#define ONWORKLIST 0x008000
+#defineONWORKLIST  0x008000
 #defineIOWAITING   0x01 /* Thread is waiting for IO to 
complete. */
 #defineONDEPLIST   0x02 /* Structure is on a dependency list. 
*/
 #defineUNLINKED0x04 /* inodedep has been unlinked. */
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258387 - head/sys/dev/nand

2013-11-20 Thread John-Mark Gurney 
Grzegorz Bernacki wrote this message on Wed, Nov 20, 2013 at 11:10 +:
> Author: gber
> Date: Wed Nov 20 11:10:23 2013
> New Revision: 258387
> URL: http://svnweb.freebsd.org/changeset/base/258387
> 
> Log:
>   Split raw reading/programming into smaller chunks to avoid allocating too
>   big chunk of kernel memory. Validate size of data. Add error handling to
>   avoid calling copyout() when data has not been read correctly.
>   
>   Reviewed by:zbb
>   Reported by:x90c 
>   MFC after:  2 days
> 
> Modified:
>   head/sys/dev/nand/nand_cdev.c
>   head/sys/dev/nand/nand_geom.c

Looks like this change errors w/ gcc:
http://tinderbox.freebsd.org/tinderbox-head-noclang-build-HEAD-armv6-arm.brief

-- 
  John-Mark Gurney  Voice: +1 415 225 5579

 "All that I will do, has been done, All that I have, has not."
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Adrian Chadd 
Can we revert this and just quickly put together something else that
won't potentially come back to bite us with weird side effects?

My suggestions (and I'm happy to do this work if needed):

* create a lightweight mbuf queue representation so an mbuf list isn't
just the mbuf header pointer;
* create a new ether input (say, ether_input_multi) that takes an mbuf
list, so existing driver code does the right thing.

After that it'd be nice to write a set of mbuf list macros for
abstract the whole queue, dequeue, concat, iterate, etc (like
sys/queue.h, but for mbufs.)

What do people think?

(I've been doing it for m->next chained things, but not m->m_nextpkt things..)



-adrian


On 18 November 2013 14:58, George V. Neville-Neil  wrote:
> Author: gnn
> Date: Mon Nov 18 22:58:14 2013
> New Revision: 258328
> URL: http://svnweb.freebsd.org/changeset/base/258328
>
> Log:
>   Allow ethernet drivers to pass in packets connected via the nextpkt pointer.
>   Handling packets in this way allows drivers to amortize work during packet 
> reception.
>
>   Submitted by: Vijay Singh
>   Sponsored by: NetApp
>
> Modified:
>   head/sys/net/if_ethersubr.c
>
> Modified: head/sys/net/if_ethersubr.c
> ==
> --- head/sys/net/if_ethersubr.c Mon Nov 18 22:55:50 2013(r258327)
> +++ head/sys/net/if_ethersubr.c Mon Nov 18 22:58:14 2013(r258328)
> @@ -708,13 +708,25 @@ static void
>  ether_input(struct ifnet *ifp, struct mbuf *m)
>  {
>
> +   struct mbuf *mn;
> +
> /*
> -* We will rely on rcvif being set properly in the deferred context,
> -* so assert it is correct here.
> +* The drivers are allowed to pass in a chain of packets linked with
> +* m_nextpkt. We split them up into separate packets here and pass
> +* them up. This allows the drivers to amortize the receive lock.
>  */
> -   KASSERT(m->m_pkthdr.rcvif == ifp, ("%s: ifnet mismatch", __func__));
> +   while (m) {
> +   mn = m->m_nextpkt;
> +   m->m_nextpkt = NULL;
>
> -   netisr_dispatch(NETISR_ETHER, m);
> +   /*
> +* We will rely on rcvif being set properly in the deferred 
> context,
> +* so assert it is correct here.
> +*/
> +   KASSERT(m->m_pkthdr.rcvif == ifp, ("%s: ifnet mismatch", 
> __func__));
> +   netisr_dispatch(NETISR_ETHER, m);
> +   m = mn;
> +   }
>  }
>
>  /*
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258406 - head/usr.sbin/bsdconfig/includes

2013-11-20 Thread Devin Teske 
Author: dteske
Date: Wed Nov 20 22:11:42 2013
New Revision: 258406
URL: http://svnweb.freebsd.org/changeset/base/258406

Log:
  Make `-d' implicitly enable `-f' for simplification.

Modified:
  head/usr.sbin/bsdconfig/includes/USAGE
  head/usr.sbin/bsdconfig/includes/includes

Modified: head/usr.sbin/bsdconfig/includes/USAGE
==
--- head/usr.sbin/bsdconfig/includes/USAGE  Wed Nov 20 21:59:24 2013
(r258405)
+++ head/usr.sbin/bsdconfig/includes/USAGE  Wed Nov 20 22:11:42 2013
(r258406)
@@ -28,7 +28,7 @@ Usage: bsdconfig @PROGRAM_NAME@ [OPTIONS
 
 OPTIONS:
-aAlways use color even when output is not to a terminal.
-   -dPrint the description for each function selected.
+   -dPrint description for each function selected. Implies `-f'.
-fShow functions for selected includes.
-F pattern
  If `-f', only print functions matching pattern. Without `-f'
@@ -68,4 +68,4 @@ EXAMPLES:
 
Show descriptions of each of the `show' functions:
 
-   bsdconfig @PROGRAM_NAME@ -dfF show
+   bsdconfig @PROGRAM_NAME@ -dF show

Modified: head/usr.sbin/bsdconfig/includes/includes
==
--- head/usr.sbin/bsdconfig/includes/includes   Wed Nov 20 21:59:24 2013
(r258405)
+++ head/usr.sbin/bsdconfig/includes/includes   Wed Nov 20 22:11:42 2013
(r258406)
@@ -127,7 +127,7 @@ show_include()
 while getopts adfF:hn flag; do
case "$flag" in
a) USE_COLOR=1 ;;
-   d) SHOW_DESC=1 ;;
+   d) SHOW_DESC=1 SHOW_FUNCS=1 ;;
f) SHOW_FUNCS=1 ;;
F) FUNC_PATTERN="$OPTARG" ;;
n) USE_COLOR= ;;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258407 - head/usr.sbin/bsdconfig/share/media

2013-11-20 Thread Devin Teske 
Author: dteske
Date: Wed Nov 20 22:12:21 2013
New Revision: 258407
URL: http://svnweb.freebsd.org/changeset/base/258407

Log:
  Update function description for API tool:
  bsdconfig includes -dF f_validate_hostname

Modified:
  head/usr.sbin/bsdconfig/share/media/tcpip.subr

Modified: head/usr.sbin/bsdconfig/share/media/tcpip.subr
==
--- head/usr.sbin/bsdconfig/share/media/tcpip.subr  Wed Nov 20 22:11:42 
2013(r258406)
+++ head/usr.sbin/bsdconfig/share/media/tcpip.subr  Wed Nov 20 22:12:21 
2013(r258407)
@@ -110,10 +110,11 @@ f_struct_define DHCP_LEASE \
 #  begin or end with a hyphen).
 #3 One or more individual labels within the hostname are null.
 #
-# f_dialog_validate_hostname $hostname
+# To call this function and display an appropriate error message to the user
+# based on the above error codes, use the following function defined in
+# dialog.subr:
 #
-# If the hostname is determined to be invalid, the appropriate error will be
-# displayed using the f_show_msg function.
+#  f_dialog_validate_hostname $hostname
 #
 f_validate_hostname()
 {
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread George Neville-Neil 

On Nov 20, 2013, at 17:02 , Adrian Chadd  wrote:

> Can we revert this and just quickly put together something else that
> won't potentially come back to bite us with weird side effects?
> 
> My suggestions (and I'm happy to do this work if needed):
> 
> * create a lightweight mbuf queue representation so an mbuf list isn't
> just the mbuf header pointer;
> * create a new ether input (say, ether_input_multi) that takes an mbuf
> list, so existing driver code does the right thing.
> 
> After that it'd be nice to write a set of mbuf list macros for
> abstract the whole queue, dequeue, concat, iterate, etc (like
> sys/queue.h, but for mbufs.)
> 
> What do people think?
> 
> (I've been doing it for m->next chained things, but not m->m_nextpkt things..)

I think the right way to do this is to move forwards and not backwards.
This change has the nice effect of not breaking anything else in the
tree while providing us with a feature that will be useful.  

If you want to add an ether_input_multi() that might make sense as an
adjunct to what’s there now.

Best,
Geoge
 


signature.asc
Description: Message signed with OpenPGP using GPGMail

svn commit: r258410 - head/sys/arm/arm

2013-11-20 Thread Olivier Houchard 
Author: cognet
Date: Wed Nov 20 23:06:54 2013
New Revision: 258410
URL: http://svnweb.freebsd.org/changeset/base/258410

Log:
  In pmap_unmapdev(), remember the size, and use that as an argument to
  kva_free(), or we'd end up always passing it a size of 0, and for some
  strange reason it doesn't seem to like it.

Modified:
  head/sys/arm/arm/devmap.c

Modified: head/sys/arm/arm/devmap.c
==
--- head/sys/arm/arm/devmap.c   Wed Nov 20 23:02:09 2013(r258409)
+++ head/sys/arm/arm/devmap.c   Wed Nov 20 23:06:54 2013(r258410)
@@ -240,6 +240,7 @@ void
 pmap_unmapdev(vm_offset_t va, vm_size_t size)
 {
vm_offset_t tmpva, offset;
+   vm_size_t origsize = size;

offset = va & PAGE_MASK;
va = trunc_page(va);
@@ -251,6 +252,6 @@ pmap_unmapdev(vm_offset_t va, vm_size_t 
tmpva += PAGE_SIZE;
}
 
-   kva_free(va, size);
+   kva_free(va, origsize);
 }
 
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Luigi Rizzo 
On Wed, Nov 20, 2013 at 05:29:37PM -0500, George Neville-Neil wrote:
> 
> On Nov 20, 2013, at 17:02 , Adrian Chadd  wrote:
> 
> > Can we revert this and just quickly put together something else that
> > won't potentially come back to bite us with weird side effects?
> > 
> > My suggestions (and I'm happy to do this work if needed):
> > 
> > * create a lightweight mbuf queue representation so an mbuf list isn't
> > just the mbuf header pointer;
> > * create a new ether input (say, ether_input_multi) that takes an mbuf
> > list, so existing driver code does the right thing.
> > 
> > After that it'd be nice to write a set of mbuf list macros for
> > abstract the whole queue, dequeue, concat, iterate, etc (like
> > sys/queue.h, but for mbufs.)
> > 
> > What do people think?
> > 
> > (I've been doing it for m->next chained things, but not m->m_nextpkt 
> > things..)
> 
> I think the right way to do this is to move forwards and not backwards.
> This change has the nice effect of not breaking anything else in the
> tree while providing us with a feature that will be useful.  

I am a bit torn on this.

George is right - we should move forward, and we have been
discussing this trivial change for two years.
It is impossible to tell when someone will find the time to
implement the mbuf queue and the new method (more on this later).
So we should not back it out.

At the same time: in principle this change does not break
anything, but Robert is right that some of the 50-100 drivers
we have might forget to clear m_nextpkt and cause trouble.
So it would be nice to implement some protection, maybe
something as simple as setting a flag in the first mbuf to
tell that this is a chain and not a single mbuf.
We have 12 M_PROTO* flags to use.

For the next step:

I am still wondering what is the best way to implement
an alternate ifp->if_input_multi method (which takes an mbuf queue
as an argument, and can be checked by the compiler).
I am not too fond of having two different input methods,
as it complicates life when one has to intercept packets.
On the other hand, globally changing all drivers to use the
if_input_multi() method is too much trouble, so i suspect
we will have to come up with some nice trick, e.g.
use a pseudo mbuf with m_type = MT_MBQUEUE
and overload the m_hdr fields as queue pointers.

cheers
luigi

___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Adrian Chadd 
Hi,

We should migrate drivers to use a multi-input method where it's
appropriate. It's the same pain as if_transmit() is/was.

I'd really like to avoid having hacky solutions like mbufs with magic
types. If we're going down that path, we should create a correct
inline messaging mechanism that includes arbitrary messages in the
stream, where some may or may not be mbufs. Magic mbufs just makes me
want to tear out my eyes a little.

So, the reason I'd like to back it out is because we should be doing
it via a multi method with some type that represents an mbuf list. If
George doesn't mind, I'll add a multi input method, move this stuff
into it, and make ether_input just be single frames.



-adrian
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Luigi Rizzo 
On Wed, Nov 20, 2013 at 04:07:07PM -0800, Adrian Chadd wrote:
> Hi,
> 
> We should migrate drivers to use a multi-input method where it's
> appropriate. It's the same pain as if_transmit() is/was.

right, and i think that is very confusing and i'd rather
not replicate the experience.

But what is your plan, have both if_input and if_input_multi ?
And then make the vlan and all similar filters intercept both ?
Because all of the existing options have pros and cons:

1. having both if_input and if_input_multi is visually cleaner
   but requires extending ifnet and some convoluted code in the
   initialization (same as if_transmit)
   
2. just if_input with typed mbufs is less clean but has the big
   advantage thay you only need to change ether_input() (and equivalent
   for other L2 protocols), and it is not error prone

3. having only if_input_multi (even without renaming if_input)
   requires you to change all the 100+ drivers.

It seems to me that #2 at least preserves binary compatibility
with driver modules and is easier to backport to other versions
of FreeBSD, this is why i prefer it.

my two cents...

cheers
luigi

> I'd really like to avoid having hacky solutions like mbufs with magic
> types. If we're going down that path, we should create a correct
> inline messaging mechanism that includes arbitrary messages in the
> stream, where some may or may not be mbufs. Magic mbufs just makes me
> want to tear out my eyes a little.
> 
> So, the reason I'd like to back it out is because we should be doing
> it via a multi method with some type that represents an mbuf list. If
> George doesn't mind, I'll add a multi input method, move this stuff
> into it, and make ether_input just be single frames.
> 
> 
> 
> -adrian
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258411 - head/usr.sbin/sysrc

2013-11-20 Thread Devin Teske 
Author: dteske
Date: Thu Nov 21 00:54:26 2013
New Revision: 258411
URL: http://svnweb.freebsd.org/changeset/base/258411

Log:
  Revert SVN r257830 -- that feature was annoying and was removed long ago.
  Was a momentary memory lapse induced by old code that was lying in my ~/bin

Modified:
  head/usr.sbin/sysrc/sysrc.8

Modified: head/usr.sbin/sysrc/sysrc.8
==
--- head/usr.sbin/sysrc/sysrc.8 Wed Nov 20 23:06:54 2013(r258410)
+++ head/usr.sbin/sysrc/sysrc.8 Thu Nov 21 00:54:26 2013(r258411)
@@ -24,7 +24,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd Nov 7, 2013
+.Dd Nov 20, 2013
 .Dt SYSRC 8
 .Os
 .Sh NAME
@@ -185,7 +185,7 @@ be called repeatedly).
 .Sh ENVIRONMENT
 The following environment variables are referenced by
 .Nm :
-.Bl -tag -width ".Ev SYSRC_VERBOSE"
+.Bl -tag -width ".Ev RC_DEFAULTS"
 .It Ev RC_CONFS
 Override default
 .Ql rc_conf_files
@@ -194,9 +194,6 @@ Override default
 Location of
 .Ql /etc/defaults/rc.conf
 file.
-.It Ev SYSRC_VERBOSE
-Default verbosity.
-Set to non-NULL to enable.
 .El
 .Sh DEPENDENCIES
 The following standard commands are required by
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258412 - in head/sys/arm: at91 econa s3c2xx0 sa11x0 xscale/i80321 xscale/i8134x xscale/ixp425 xscale/pxa

2013-11-20 Thread Ian Lepore 
Author: ian
Date: Thu Nov 21 01:08:10 2013
New Revision: 258412
URL: http://svnweb.freebsd.org/changeset/base/258412

Log:
  Call cpu_setup() from the initarm() routine on platforms that don't use
  the common FDT-aware initarm() in arm/machdep.c.
  
  Pointed out by:cognet
  Pointy hat to: ian

Modified:
  head/sys/arm/at91/at91_machdep.c
  head/sys/arm/econa/econa_machdep.c
  head/sys/arm/s3c2xx0/s3c24x0_machdep.c
  head/sys/arm/sa11x0/assabet_machdep.c
  head/sys/arm/xscale/i80321/ep80219_machdep.c
  head/sys/arm/xscale/i80321/iq31244_machdep.c
  head/sys/arm/xscale/i8134x/crb_machdep.c
  head/sys/arm/xscale/ixp425/avila_machdep.c
  head/sys/arm/xscale/pxa/pxa_machdep.c

Modified: head/sys/arm/at91/at91_machdep.c
==
--- head/sys/arm/at91/at91_machdep.cThu Nov 21 00:54:26 2013
(r258411)
+++ head/sys/arm/at91/at91_machdep.cThu Nov 21 01:08:10 2013
(r258412)
@@ -604,6 +604,7 @@ initarm(struct arm_boot_params *abp)
 * of the stack memory.
 */
cpu_control(CPU_CONTROL_MMU_ENABLE, CPU_CONTROL_MMU_ENABLE);
+   cpu_setup("");
 
set_stackptrs(0);
 

Modified: head/sys/arm/econa/econa_machdep.c
==
--- head/sys/arm/econa/econa_machdep.c  Thu Nov 21 00:54:26 2013
(r258411)
+++ head/sys/arm/econa/econa_machdep.c  Thu Nov 21 01:08:10 2013
(r258412)
@@ -309,6 +309,7 @@ initarm(struct arm_boot_params *abp)
 * this problem will not occur after initarm().
 */
cpu_idcache_wbinv_all();
+   cpu_setup("");
 
/* Set stack for exception handlers */
data_abort_handler_address = (u_int)data_abort_handler;

Modified: head/sys/arm/s3c2xx0/s3c24x0_machdep.c
==
--- head/sys/arm/s3c2xx0/s3c24x0_machdep.c  Thu Nov 21 00:54:26 2013
(r258411)
+++ head/sys/arm/s3c2xx0/s3c24x0_machdep.c  Thu Nov 21 01:08:10 2013
(r258412)
@@ -355,6 +355,7 @@ initarm(struct arm_boot_params *abp)
 * this problem will not occur after initarm().
 */
cpu_idcache_wbinv_all();
+   cpu_setup("");
 
/* Disable all peripheral interrupts */
ioreg_write32(S3C24X0_INTCTL_BASE + INTCTL_INTMSK, ~0);

Modified: head/sys/arm/sa11x0/assabet_machdep.c
==
--- head/sys/arm/sa11x0/assabet_machdep.c   Thu Nov 21 00:54:26 2013
(r258411)
+++ head/sys/arm/sa11x0/assabet_machdep.c   Thu Nov 21 01:08:10 2013
(r258412)
@@ -371,6 +371,7 @@ initarm(struct arm_boot_params *abp)
 
cpufunc_control(0x337f, 0x107d);
arm_vector_init(ARM_VECTORS_LOW, ARM_VEC_ALL);
+   cpu_setup("");
 
pmap_curmaxkvaddr = freemempos + KERNEL_PT_VMDATA_NUM * 0x40;
 

Modified: head/sys/arm/xscale/i80321/ep80219_machdep.c
==
--- head/sys/arm/xscale/i80321/ep80219_machdep.cThu Nov 21 00:54:26 
2013(r258411)
+++ head/sys/arm/xscale/i80321/ep80219_machdep.cThu Nov 21 01:08:10 
2013(r258412)
@@ -334,6 +334,8 @@ initarm(struct arm_boot_params *abp)
 * this problem will not occur after initarm().
 */
cpu_idcache_wbinv_all();
+   cpu_setup("");
+
/*
 * Fetch the SDRAM start/size from the i80321 SDRAM configration
 * registers.

Modified: head/sys/arm/xscale/i80321/iq31244_machdep.c
==
--- head/sys/arm/xscale/i80321/iq31244_machdep.cThu Nov 21 00:54:26 
2013(r258411)
+++ head/sys/arm/xscale/i80321/iq31244_machdep.cThu Nov 21 01:08:10 
2013(r258412)
@@ -335,6 +335,8 @@ initarm(struct arm_boot_params *abp)
 * this problem will not occur after initarm().
 */
cpu_idcache_wbinv_all();
+   cpu_setup("");
+
/*
 * Fetch the SDRAM start/size from the i80321 SDRAM configration
 * registers.

Modified: head/sys/arm/xscale/i8134x/crb_machdep.c
==
--- head/sys/arm/xscale/i8134x/crb_machdep.cThu Nov 21 00:54:26 2013
(r258411)
+++ head/sys/arm/xscale/i8134x/crb_machdep.cThu Nov 21 01:08:10 2013
(r258412)
@@ -320,6 +320,8 @@ initarm(struct arm_boot_params *abp)
 * this problem will not occur after initarm().
 */
cpu_idcache_wbinv_all();
+   cpu_setup("");
+
i80321_calibrate_delay();
i81342_sdram_bounds(&obio_bs_tag, IOP34X_VADDR, &memstart, &memsize);
physmem = memsize / PAGE_SIZE;

Modified: head/sys/arm/xscale/ixp425/avila_machdep.c
==

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Julian Elischer 

On 11/20/13, 2:02 PM, Adrian Chadd wrote:

Can we revert this and just quickly put together something else that
won't potentially come back to bite us with weird side effects?

My suggestions (and I'm happy to do this work if needed):

* create a lightweight mbuf queue representation so an mbuf list isn't
just the mbuf header pointer;
* create a new ether input (say, ether_input_multi) that takes an mbuf
list, so existing driver code does the right thing.

After that it'd be nice to write a set of mbuf list macros for
abstract the whole queue, dequeue, concat, iterate, etc (like
sys/queue.h, but for mbufs.)

What do people think?

(I've been doing it for m->next chained things, but not m->m_nextpkt things..)


I was thinking:
new interfaces.. (your -multi names sound good).
macros for handling said lists so that people don't screw them up
Old drivers run with no change.





-adrian


On 18 November 2013 14:58, George V. Neville-Neil  wrote:

Author: gnn
Date: Mon Nov 18 22:58:14 2013
New Revision: 258328
URL: http://svnweb.freebsd.org/changeset/base/258328

Log:
   Allow ethernet drivers to pass in packets connected via the nextpkt pointer.
   Handling packets in this way allows drivers to amortize work during packet 
reception.

   Submitted by: Vijay Singh
   Sponsored by: NetApp

Modified:
   head/sys/net/if_ethersubr.c

Modified: head/sys/net/if_ethersubr.c
==
--- head/sys/net/if_ethersubr.c Mon Nov 18 22:55:50 2013(r258327)
+++ head/sys/net/if_ethersubr.c Mon Nov 18 22:58:14 2013(r258328)
@@ -708,13 +708,25 @@ static void
  ether_input(struct ifnet *ifp, struct mbuf *m)
  {

+   struct mbuf *mn;
+
 /*
-* We will rely on rcvif being set properly in the deferred context,
-* so assert it is correct here.
+* The drivers are allowed to pass in a chain of packets linked with
+* m_nextpkt. We split them up into separate packets here and pass
+* them up. This allows the drivers to amortize the receive lock.
  */
-   KASSERT(m->m_pkthdr.rcvif == ifp, ("%s: ifnet mismatch", __func__));
+   while (m) {
+   mn = m->m_nextpkt;
+   m->m_nextpkt = NULL;

-   netisr_dispatch(NETISR_ETHER, m);
+   /*
+* We will rely on rcvif being set properly in the deferred 
context,
+* so assert it is correct here.
+*/
+   KASSERT(m->m_pkthdr.rcvif == ifp, ("%s: ifnet mismatch", 
__func__));
+   netisr_dispatch(NETISR_ETHER, m);
+   m = mn;
+   }
  }

  /*


___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Re: svn commit: r258328 - head/sys/net

2013-11-20 Thread Julian Elischer 

On 11/20/13, 4:35 PM, Luigi Rizzo wrote:

On Wed, Nov 20, 2013 at 04:07:07PM -0800, Adrian Chadd wrote:

Hi,

We should migrate drivers to use a multi-input method where it's
appropriate. It's the same pain as if_transmit() is/was.

right, and i think that is very confusing and i'd rather
not replicate the experience.

But what is your plan, have both if_input and if_input_multi ?
And then make the vlan and all similar filters intercept both ?
Because all of the existing options have pros and cons:

1. having both if_input and if_input_multi is visually cleaner
but requires extending ifnet and some convoluted code in the
initialization (same as if_transmit)

2. just if_input with typed mbufs is less clean but has the big

advantage thay you only need to change ether_input() (and equivalent
for other L2 protocols), and it is not error prone


if_input_multi() or whatever it is called is just this patch split 
into two bits.
in the simple case it just calls the single one multiple times, just 
like this patch does.

it is clean from teh compiler's point of view,


3. having only if_input_multi (even without renaming if_input)
requires you to change all the 100+ drivers.

It seems to me that #2 at least preserves binary compatibility
with driver modules and is easier to backport to other versions
of FreeBSD, this is why i prefer it.

my two cents...

cheers
luigi


I'd really like to avoid having hacky solutions like mbufs with magic
types. If we're going down that path, we should create a correct
inline messaging mechanism that includes arbitrary messages in the
stream, where some may or may not be mbufs. Magic mbufs just makes me
want to tear out my eyes a little.

So, the reason I'd like to back it out is because we should be doing
it via a multi method with some type that represents an mbuf list. If
George doesn't mind, I'll add a multi input method, move this stuff
into it, and make ether_input just be single frames.



-adrian

___
freebsd-a...@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-arch
To unsubscribe, send any mail to "freebsd-arch-unsubscr...@freebsd.org"



___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258418 - head/usr.sbin/bsdconfig/networking/share

2013-11-20 Thread Devin Teske 
Author: dteske
Date: Thu Nov 21 03:22:13 2013
New Revision: 258418
URL: http://svnweb.freebsd.org/changeset/base/258418

Log:
  Quote the interface name for good measure.

Modified:
  head/usr.sbin/bsdconfig/networking/share/device.subr

Modified: head/usr.sbin/bsdconfig/networking/share/device.subr
==
--- head/usr.sbin/bsdconfig/networking/share/device.subrThu Nov 21 
03:01:28 2013(r258417)
+++ head/usr.sbin/bsdconfig/networking/share/device.subrThu Nov 21 
03:22:13 2013(r258418)
@@ -278,22 +278,23 @@ f_dialog_menu_netdev_edit()
  msg=$( printf "$msg_scanning_for_dhcp" "$interface" )
  if [ "$USE_XDIALOG" ]; then
(
- f_quietly ifconfig $interface delete
- f_quietly dhclient $interface
+ f_quietly ifconfig "$interface" delete
+ f_quietly dhclient "$interface"
) |
  f_xdialog_info "$msg"
  else
f_dialog_info "$msg"
-   f_quietly ifconfig $interface delete
-   f_quietly dhclient $interface
+   f_quietly ifconfig "$interface" delete
+   f_quietly dhclient "$interface"
  fi
)
retval=$?
trap 'interrupt' SIGINT
if [ $retval -eq $DIALOG_OK ]; then
dhcp=1
-   ipaddr=$( f_ifconfig_inet $interface )
-   netmask=$( f_ifconfig_netmask $interface )
+   f_ifconfig_inet "$interface" ipaddr
+   f_ifconfig_inet6 "$interface" ipaddr6
+   f_ifconfig_netmask "$interface" netmask
options=
 
# Fixup search/domain in resolv.conf(5)
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

svn commit: r258424 - head/sys/powerpc/aim

2013-11-20 Thread Justin Hibbits 
Author: jhibbits
Date: Thu Nov 21 06:54:28 2013
New Revision: 258424
URL: http://svnweb.freebsd.org/changeset/base/258424

Log:
  Remove stale comment.  The PID provider is handled elsewhere already.

Modified:
  head/sys/powerpc/aim/trap.c

Modified: head/sys/powerpc/aim/trap.c
==
--- head/sys/powerpc/aim/trap.c Thu Nov 21 05:17:36 2013(r258423)
+++ head/sys/powerpc/aim/trap.c Thu Nov 21 06:54:28 2013(r258424)
@@ -205,9 +205,6 @@ trap(struct trapframe *frame)
 * handled the trap and modified the trap frame so that this
 * function can return normally.
 */
-   /*
-* XXXDTRACE: add pid probe handler here (if ever)
-*/
if (dtrace_trap_func != NULL && (*dtrace_trap_func)(frame, type))
return;
 #endif
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"