d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
ae0669e390
freebsd-nq/sys/net/bpf.h

474 lines
15 KiB
C
Raw Normal View History

/* -> /*- for license, minor formatting changes
2005-01-07 01:45:51 +00:00
/*-
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* Copyright (c) 1990, 1991, 1993
Revert r228986 until it can be reworked to avoid panicing the kernel when the same interface is attached multiple times with different DLTs, as is done in net80211 for example. Reported by: adrian
2011-12-31 07:21:28 +00:00
* The Regents of the University of California. All rights reserved.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*
* This code is derived from the Stanford/CMU enet packet filter,
* (net/enet.c) distributed as part of 4.3BSD, and code contributed
* to Berkeley by Steven McCanne and Van Jacobson both of Lawrence
* Berkeley Laboratory.
*
* 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.
Renumber copyright clause 4 Renumber cluase 4 to 3, per what everybody else did when BSD granted them permission to remove clause 3. My insistance on keeping the same numbering for legal reasons is too pedantic, so give up on that point. Submitted by: Jan Schaumann <jschauma@stevens.edu> Pull Request: https://github.com/freebsd/freebsd/pull/96
2017-02-28 23:42:47 +00:00
* 3. Neither the name of the University nor the names of its contributors
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
*
* @(#)bpf.h 8.1 (Berkeley) 6/10/93
Update to match definitions in LBL June 96 release
1996-08-19 20:28:25 +00:00
* @(#)bpf.h 1.34 (LBL) 6/16/96
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*
$Id$ -> $FreeBSD$
1999-08-28 01:08:13 +00:00
* $FreeBSD$
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Make idempotent. Submitted by: Paul
1994-08-21 05:11:48 +00:00
#ifndef _NET_BPF_H_
#define _NET_BPF_H_
Update to match definitions in LBL June 96 release
1996-08-19 20:28:25 +00:00
/* BSD style release date */
#define BPF_RELEASE 199606
typedef int32_t bpf_int32;
typedef u_int32_t bpf_u_int32;
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
typedef int64_t bpf_int64;
typedef u_int64_t bpf_u_int64;
Update to match definitions in LBL June 96 release
1996-08-19 20:28:25 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Remove trailing whitespace.
1995-05-30 08:16:23 +00:00
* Alignment macros. BPF_WORDALIGN rounds up to the next
* even multiple of BPF_ALIGNMENT.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Change BPF_ALIGNMENT to long, necessary for correct alignment on Alpha.
1998-10-04 21:53:59 +00:00
#define BPF_ALIGNMENT sizeof(long)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_WORDALIGN(x) (((x)+(BPF_ALIGNMENT-1))&~(BPF_ALIGNMENT-1))
#define BPF_MAXINSNS 512
|The hard limit for the BPF buffer size is 32KB, which appears too low |for high speed networks (even at 100Mbit/s this corresponds to 1/300th |of a second). The default buffer size is 4KB, but libpcap and ipfilter |both override this (using the BIOCSBLEN ioctl) and allocate 32KB. | |The following patch adds an sysctl for bpf_maxbufsize, similar to the |one for bpf_bufsize that you added back in December 1995. I choose to |make the default for this limit 512KB (the value suggested by NFR). Submitted by: se Reviewed by: phk
2000-01-15 19:46:12 +00:00
#define BPF_MAXBUFSIZE 0x80000
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_MINBUFSIZE 32
/*
* Structure for BIOCSETF.
*/
struct bpf_program {
u_int bf_len;
struct bpf_insn *bf_insns;
};
Remove trailing whitespace.
1995-05-30 08:16:23 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Struct returned by BIOCGSTATS.
*/
struct bpf_stat {
u_int bs_recv; /* number of packets received */
u_int bs_drop; /* number of packets dropped */
};
/*
Remove trailing whitespace.
1995-05-30 08:16:23 +00:00
* Struct return by BIOCVERSION. This represents the version number of
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* the filter language described by the instruction encodings below.
* bpf understands a program iff kernel_major == filter_major &&
* kernel_minor >= filter_minor, that is, if the value returned by the
* running kernel has the same major number and a minor number equal
* equal to or less than the filter being downloaded. Otherwise, the
* results are undefined, meaning an error may be returned or packets
* may be accepted haphazardly.
* It has nothing to do with the source code version.
*/
struct bpf_version {
u_short bv_major;
u_short bv_minor;
};
Update to match definitions in LBL June 96 release
1996-08-19 20:28:25 +00:00
/* Current version number of filter architecture. */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_MAJOR_VERSION 1
#define BPF_MINOR_VERSION 1
Introduce support for zero-copy BPF buffering, which reduces the overhead of packet capture by allowing a user process to directly "loan" buffer memory to the kernel rather than using read(2) to explicitly copy data from kernel address space. The user process will issue new BPF ioctls to set the shared memory buffer mode and provide pointers to buffers and their size. The kernel then wires and maps the pages into kernel address space using sf_buf(9), which on supporting architectures will use the direct map region. The current "buffered" access mode remains the default, and support for zero-copy buffers must, for the time being, be explicitly enabled using a sysctl for the kernel to accept requests to use it. The kernel and user process synchronize use of the buffers with atomic operations, avoiding the need for system calls under load; the user process may use select()/poll()/kqueue() to manage blocking while waiting for network data if the user process is able to consume data faster than the kernel generates it. Patchs to libpcap are available to allow libpcap applications to transparently take advantage of this support. Detailed information on the new API may be found in bpf(4), including specific atomic operations and memory barriers required to synchronize buffer use safely. These changes modify the base BPF implementation to (roughly) abstrac the current buffer model, allowing the new shared memory model to be added, and add new monitoring statistics for netstat to print. The implementation, with the exception of some monitoring hanges that break the netstat monitoring ABI for BPF, will be MFC'd. Zerocopy bpf buffers are still considered experimental are disabled by default. To experiment with this new facility, adjust the net.bpf.zerocopy_enable sysctl variable to 1. Changes to libpcap will be made available as a patch for the time being, and further refinements to the implementation are expected. Sponsored by: Seccuris Inc. In collaboration with: rwatson Tested by: pwood, gallatin MFC after: 4 months [1] [1] Certain portions will probably not be MFCed, specifically things that can break the monitoring ABI.
2008-03-24 13:49:17 +00:00
/*
* Historically, BPF has supported a single buffering model, first using mbuf
* clusters in kernel, and later using malloc(9) buffers in kernel. We now
* support multiple buffering modes, which may be queried and set using
* BIOCGETBUFMODE and BIOCSETBUFMODE. So as to avoid handling the complexity
* of changing modes while sniffing packets, the mode becomes fixed once an
* interface has been attached to the BPF descriptor.
*/
#define BPF_BUFMODE_BUFFER 1 /* Kernel buffers with read(). */
#define BPF_BUFMODE_ZBUF 2 /* Zero-copy buffers. */
/*-
* Struct used by BIOCSETZBUF, BIOCROTZBUF: describes up to two zero-copy
* buffer as used by BPF.
*/
struct bpf_zbuf {
void *bz_bufa; /* Location of 'a' zero-copy buffer. */
void *bz_bufb; /* Location of 'b' zero-copy buffer. */
size_t bz_buflen; /* Size of zero-copy buffers. */
};
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
#define BIOCGBLEN _IOR('B', 102, u_int)
#define BIOCSBLEN _IOWR('B', 102, u_int)
#define BIOCSETF _IOW('B', 103, struct bpf_program)
#define BIOCFLUSH _IO('B', 104)
#define BIOCPROMISC _IO('B', 105)
#define BIOCGDLT _IOR('B', 106, u_int)
#define BIOCGETIF _IOR('B', 107, struct ifreq)
#define BIOCSETIF _IOW('B', 108, struct ifreq)
#define BIOCSRTIMEOUT _IOW('B', 109, struct timeval)
#define BIOCGRTIMEOUT _IOR('B', 110, struct timeval)
#define BIOCGSTATS _IOR('B', 111, struct bpf_stat)
#define BIOCIMMEDIATE _IOW('B', 112, u_int)
#define BIOCVERSION _IOR('B', 113, struct bpf_version)
#define BIOCGRSIG _IOR('B', 114, u_int)
#define BIOCSRSIG _IOW('B', 115, u_int)
#define BIOCGHDRCMPLT _IOR('B', 116, u_int)
#define BIOCSHDRCMPLT _IOW('B', 117, u_int)
#define BIOCGDIRECTION _IOR('B', 118, u_int)
#define BIOCSDIRECTION _IOW('B', 119, u_int)
#define BIOCSDLT _IOW('B', 120, u_int)
#define BIOCGDLTLIST _IOWR('B', 121, struct bpf_dltlist)
Introduce two new ioctl(2) commands, BIOCLOCK and BIOCSETWF. These commands enhance the security of bpf(4) by further relinquishing the privilege of the bpf(4) consumer (assuming the ioctl commands are being implemented). Once BIOCLOCK is executed, the device becomes locked which prevents the execution of ioctl(2) commands which can change the underly parameters of the bpf(4) device. An example might be the setting of bpf(4) filter programs or attaching to different network interfaces. BIOCSETWF can be used to set write filters for outgoing packets. Currently if a bpf(4) consumer is compromised, the bpf(4) descriptor can essentially be used as a raw socket, regardless of consumer's UID. Write filters give users the ability to constrain which packets can be sent through the bpf(4) descriptor. These features are currently implemented by a couple programs which came from OpenBSD, such as the new dhclient and pflogd. -Modify bpf_setf(9) to accept a "cmd" parameter. This will be used to specify whether a read or write filter is to be set. -Add a bpf(4) filter program as a parameter to bpf_movein(9) as we will run the filter program on the mbuf data once we move the packet in from user-space. -Rather than execute two uiomove operations, (one for the link header and the other for the packet data), execute one and manually copy the linker header into the sockaddr structure via bcopy. -Restructure bpf_setf to compensate for write filters, as well as read. -Adjust bpf(4) stats structures to include a bd_locked member. It should be noted that the FreeBSD and OpenBSD implementations differ a bit in the sense that we unconditionally enforce the lock, where OpenBSD enforces it only if the calling credential is not root. Idea from: OpenBSD Reviewed by: mlaier
2005-08-22 19:35:48 +00:00
#define BIOCLOCK _IO('B', 122)
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
#define BIOCSETWF _IOW('B', 123, struct bpf_program)
#define BIOCFEEDBACK _IOW('B', 124, u_int)
#define BIOCGETBUFMODE _IOR('B', 125, u_int)
#define BIOCSETBUFMODE _IOW('B', 126, u_int)
#define BIOCGETZMAX _IOR('B', 127, size_t)
#define BIOCROTZBUF _IOR('B', 128, struct bpf_zbuf)
#define BIOCSETZBUF _IOW('B', 129, struct bpf_zbuf)
#define BIOCSETFNR _IOW('B', 130, struct bpf_program)
#define BIOCGTSTAMP _IOR('B', 131, u_int)
#define BIOCSTSTAMP _IOW('B', 132, u_int)
Add three new ioctl(2) commands for bpf(4). - BIOCGDIRECTION and BIOCSDIRECTION get or set the setting determining whether incoming, outgoing, or all packets on the interface should be returned by BPF. Set to BPF_D_IN to see only incoming packets on the interface. Set to BPF_D_INOUT to see packets originating locally and remotely on the interface. Set to BPF_D_OUT to see only outgoing packets on the interface. This setting is initialized to BPF_D_INOUT by default. BIOCGSEESENT and BIOCSSEESENT are obsoleted by these but kept for backward compatibility. - BIOCFEEDBACK sets packet feedback mode. This allows injected packets to be fed back as input to the interface when output via the interface is successful. When BPF_D_INOUT direction is set, injected outgoing packet is not returned by BPF to avoid duplication. This flag is initialized to zero by default. Note that libpcap has been modified to support BPF_D_OUT direction for pcap_setdirection(3) and PCAP_D_OUT direction is functional now. Reviewed by: rwatson
2007-02-26 22:24:14 +00:00
/* Obsolete */
#define BIOCGSEESENT BIOCGDIRECTION
#define BIOCSSEESENT BIOCSDIRECTION
/* Packet directions */
enum bpf_direction {
BPF_D_IN, /* See incoming packets */
BPF_D_INOUT, /* See incoming and outgoing packets */
BPF_D_OUT /* See outgoing packets */
};
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
/* Time stamping functions */
#define BPF_T_MICROTIME 0x0000
#define BPF_T_NANOTIME 0x0001
#define BPF_T_BINTIME 0x0002
#define BPF_T_NONE 0x0003
#define BPF_T_FORMAT_MASK 0x0003
#define BPF_T_NORMAL 0x0000
Revert r228986 until it can be reworked to avoid panicing the kernel when the same interface is attached multiple times with different DLTs, as is done in net80211 for example. Reported by: adrian
2011-12-31 07:21:28 +00:00
#define BPF_T_FAST 0x0100
#define BPF_T_MONOTONIC 0x0200
#define BPF_T_MONOTONIC_FAST (BPF_T_FAST | BPF_T_MONOTONIC)
#define BPF_T_FLAG_MASK 0x0300
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
#define BPF_T_FORMAT(t) ((t) & BPF_T_FORMAT_MASK)
#define BPF_T_FLAG(t) ((t) & BPF_T_FLAG_MASK)
#define BPF_T_VALID(t) \
((t) == BPF_T_NONE || (BPF_T_FORMAT(t) != BPF_T_NONE && \
((t) & ~(BPF_T_FORMAT_MASK | BPF_T_FLAG_MASK)) == 0))
Revert r228986 until it can be reworked to avoid panicing the kernel when the same interface is attached multiple times with different DLTs, as is done in net80211 for example. Reported by: adrian
2011-12-31 07:21:28 +00:00
#define BPF_T_MICROTIME_FAST (BPF_T_MICROTIME | BPF_T_FAST)
#define BPF_T_NANOTIME_FAST (BPF_T_NANOTIME | BPF_T_FAST)
#define BPF_T_BINTIME_FAST (BPF_T_BINTIME | BPF_T_FAST)
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
#define BPF_T_MICROTIME_MONOTONIC (BPF_T_MICROTIME | BPF_T_MONOTONIC)
#define BPF_T_NANOTIME_MONOTONIC (BPF_T_NANOTIME | BPF_T_MONOTONIC)
#define BPF_T_BINTIME_MONOTONIC (BPF_T_BINTIME | BPF_T_MONOTONIC)
Revert r228986 until it can be reworked to avoid panicing the kernel when the same interface is attached multiple times with different DLTs, as is done in net80211 for example. Reported by: adrian
2011-12-31 07:21:28 +00:00
#define BPF_T_MICROTIME_MONOTONIC_FAST (BPF_T_MICROTIME | BPF_T_MONOTONIC_FAST)
#define BPF_T_NANOTIME_MONOTONIC_FAST (BPF_T_NANOTIME | BPF_T_MONOTONIC_FAST)
#define BPF_T_BINTIME_MONOTONIC_FAST (BPF_T_BINTIME | BPF_T_MONOTONIC_FAST)
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Structure prepended to each packet.
*/
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
struct bpf_ts {
bpf_int64 bt_sec; /* seconds */
bpf_u_int64 bt_frac; /* fraction */
};
struct bpf_xhdr {
struct bpf_ts bh_tstamp; /* time stamp */
bpf_u_int32 bh_caplen; /* length of captured portion */
bpf_u_int32 bh_datalen; /* original length of packet */
u_short bh_hdrlen; /* length of bpf header (this struct
plus alignment padding) */
};
/* Obsolete */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct bpf_hdr {
struct timeval bh_tstamp; /* time stamp */
Update to match definitions in LBL June 96 release
1996-08-19 20:28:25 +00:00
bpf_u_int32 bh_caplen; /* length of captured portion */
bpf_u_int32 bh_datalen; /* original length of packet */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
u_short bh_hdrlen; /* length of bpf header (this struct
plus alignment padding) */
};
Change #ifdef KERNEL to #ifdef _KERNEL in the public headers. "KERNEL" is an application space macro and the applications are supposed to be free to use it as they please (but cannot). This is consistant with the other BSD's who made this change quite some time ago. More commits to come.
1999-12-29 04:46:21 +00:00
#ifdef _KERNEL
Implement flexible BPF timestamping framework. - Allow setting format, resolution and accuracy of BPF time stamps per listener. Previously, we were only able to use microtime(9). Now we can set various resolutions and accuracies with ioctl(2) BIOCSTSTAMP command. Similarly, we can get the current resolution and accuracy with BIOCGTSTAMP command. Document all supported options in bpf(4) and their uses. - Introduce new time stamp 'struct bpf_ts' and header 'struct bpf_xhdr'. The new time stamp has both 64-bit second and fractional parts. bpf_xhdr has this time stamp instead of 'struct timeval' for bh_tstamp. The new structures let us use bh_tstamp of same size on both 32-bit and 64-bit platforms without adding additional shims for 32-bit binaries. On 64-bit platforms, size of BPF header does not change compared to bpf_hdr as its members are already all 64-bit long. On 32-bit platforms, the size may increase by 8 bytes. For backward compatibility, struct bpf_hdr with struct timeval is still the default header unless new time stamp format is explicitly requested. However, the behaviour may change in the future and all relevant code is wrapped around "#ifdef BURN_BRIDGES" for now. - Add experimental support for tagging mbufs with time stamps from a lower layer, e.g., device driver. Currently, mbuf_tags(9) is used to tag mbufs. The time stamps must be uptime in 'struct bintime' format as binuptime(9) and getbinuptime(9) do. Reviewed by: net@
2010-06-15 19:28:44 +00:00
#define MTAG_BPF 0x627066
#define MTAG_BPF_TIMESTAMP 0
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#endif
Introduce support for zero-copy BPF buffering, which reduces the overhead of packet capture by allowing a user process to directly "loan" buffer memory to the kernel rather than using read(2) to explicitly copy data from kernel address space. The user process will issue new BPF ioctls to set the shared memory buffer mode and provide pointers to buffers and their size. The kernel then wires and maps the pages into kernel address space using sf_buf(9), which on supporting architectures will use the direct map region. The current "buffered" access mode remains the default, and support for zero-copy buffers must, for the time being, be explicitly enabled using a sysctl for the kernel to accept requests to use it. The kernel and user process synchronize use of the buffers with atomic operations, avoiding the need for system calls under load; the user process may use select()/poll()/kqueue() to manage blocking while waiting for network data if the user process is able to consume data faster than the kernel generates it. Patchs to libpcap are available to allow libpcap applications to transparently take advantage of this support. Detailed information on the new API may be found in bpf(4), including specific atomic operations and memory barriers required to synchronize buffer use safely. These changes modify the base BPF implementation to (roughly) abstrac the current buffer model, allowing the new shared memory model to be added, and add new monitoring statistics for netstat to print. The implementation, with the exception of some monitoring hanges that break the netstat monitoring ABI for BPF, will be MFC'd. Zerocopy bpf buffers are still considered experimental are disabled by default. To experiment with this new facility, adjust the net.bpf.zerocopy_enable sysctl variable to 1. Changes to libpcap will be made available as a patch for the time being, and further refinements to the implementation are expected. Sponsored by: Seccuris Inc. In collaboration with: rwatson Tested by: pwood, gallatin MFC after: 4 months [1] [1] Certain portions will probably not be MFCed, specifically things that can break the monitoring ABI.
2008-03-24 13:49:17 +00:00
/*
* When using zero-copy BPF buffers, a shared memory header is present
* allowing the kernel BPF implementation and user process to synchronize
* without using system calls. This structure defines that header. When
* accessing these fields, appropriate atomic operation and memory barriers
* are required in order not to see stale or out-of-order data; see bpf(4)
* for reference code to access these fields from userspace.
*
* The layout of this structure is critical, and must not be changed; if must
* fit in a single page on all architectures.
*/
struct bpf_zbuf_header {
volatile u_int bzh_kernel_gen; /* Kernel generation number. */
volatile u_int bzh_kernel_len; /* Length of data in the buffer. */
volatile u_int bzh_user_gen; /* User generation number. */
u_int _bzh_pad[5];
};
MFV r313676: libpcap 1.8.1 MFC after: 1 month
2017-02-13 08:23:39 +00:00
/* Pull in data-link level type codes. */
#include <net/dlt.h>
Sync DLTs with latest libpcap version.
2009-04-02 13:02:12 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Fix a bunch of spelling errors in the comment fields of a bunch of system include files.
1996-01-30 23:02:38 +00:00
* The instruction encodings.
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
*
* Please inform tcpdump-workers@lists.tcpdump.org if you use any
* of the reserved values, so that we can note that they're used
* (and perhaps implement it in the reference BPF implementation
* and encourage its implementation elsewhere).
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/*
* The upper 8 bits of the opcode aren't used. BSD/OS used 0x8000.
Import of libpcap 1.0.0.
2009-03-21 20:43:56 +00:00
*/
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/* instruction classes */
#define BPF_CLASS(code) ((code) & 0x07)
#define BPF_LD 0x00
#define BPF_LDX 0x01
#define BPF_ST 0x02
#define BPF_STX 0x03
#define BPF_ALU 0x04
#define BPF_JMP 0x05
#define BPF_RET 0x06
#define BPF_MISC 0x07
/* ld/ldx fields */
#define BPF_SIZE(code) ((code) & 0x18)
#define BPF_W 0x00
#define BPF_H 0x08
#define BPF_B 0x10
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/* 0x18 reserved; used by BSD/OS */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_MODE(code) ((code) & 0xe0)
#define BPF_IMM 0x00
#define BPF_ABS 0x20
#define BPF_IND 0x40
#define BPF_MEM 0x60
#define BPF_LEN 0x80
#define BPF_MSH 0xa0
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/* 0xc0 reserved; used by BSD/OS */
/* 0xe0 reserved; used by BSD/OS */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/* alu/jmp fields */
#define BPF_OP(code) ((code) & 0xf0)
#define BPF_ADD 0x00
#define BPF_SUB 0x10
#define BPF_MUL 0x20
#define BPF_DIV 0x30
#define BPF_OR 0x40
#define BPF_AND 0x50
#define BPF_LSH 0x60
#define BPF_RSH 0x70
#define BPF_NEG 0x80
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
#define BPF_MOD 0x90
#define BPF_XOR 0xa0
/* 0xb0 reserved */
/* 0xc0 reserved */
/* 0xd0 reserved */
/* 0xe0 reserved */
/* 0xf0 reserved */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_JA 0x00
#define BPF_JEQ 0x10
#define BPF_JGT 0x20
#define BPF_JGE 0x30
#define BPF_JSET 0x40
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/* 0x50 reserved; used on BSD/OS */
/* 0x60 reserved */
/* 0x70 reserved */
/* 0x80 reserved */
/* 0x90 reserved */
/* 0xa0 reserved */
/* 0xb0 reserved */
/* 0xc0 reserved */
/* 0xd0 reserved */
/* 0xe0 reserved */
/* 0xf0 reserved */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_SRC(code) ((code) & 0x08)
#define BPF_K 0x00
#define BPF_X 0x08
/* ret - BPF_K and BPF_X also apply */
#define BPF_RVAL(code) ((code) & 0x18)
#define BPF_A 0x10
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/* 0x18 reserved */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/* misc */
#define BPF_MISCOP(code) ((code) & 0xf8)
#define BPF_TAX 0x00
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/* 0x08 reserved */
/* 0x10 reserved */
/* 0x18 reserved */
/* #define BPF_COP 0x20 NetBSD "coprocessor" extensions */
/* 0x28 reserved */
/* 0x30 reserved */
/* 0x38 reserved */
/* #define BPF_COPX 0x40 NetBSD "coprocessor" extensions */
/* also used on BSD/OS */
/* 0x48 reserved */
/* 0x50 reserved */
/* 0x58 reserved */
/* 0x60 reserved */
/* 0x68 reserved */
/* 0x70 reserved */
/* 0x78 reserved */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#define BPF_TXA 0x80
Vendor import of libpcap 1.6.2.
2015-01-06 18:58:31 +00:00
/* 0x88 reserved */
/* 0x90 reserved */
/* 0x98 reserved */
/* 0xa0 reserved */
/* 0xa8 reserved */
/* 0xb0 reserved */
/* 0xb8 reserved */
/* 0xc0 reserved; used on BSD/OS */
/* 0xc8 reserved */
/* 0xd0 reserved */
/* 0xd8 reserved */
/* 0xe0 reserved */
/* 0xe8 reserved */
/* 0xf0 reserved */
/* 0xf8 reserved */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* The instruction data structure.
*/
struct bpf_insn {
Update to match definitions in LBL June 96 release
1996-08-19 20:28:25 +00:00
u_short code;
u_char jt;
u_char jf;
bpf_u_int32 k;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
/*
* Macros for insn array initializers.
*/
#define BPF_STMT(code, k) { (u_short)(code), 0, 0, k }
#define BPF_JUMP(code, k, jt, jf) { (u_short)(code), jt, jf, k }
o add BIOCGDLTLIST and BIOCSDLT ioctls to get the data link type list and set the link type for use by libpcap and tcpdump o move mtx unlock in bpfdetach up; it doesn't need to be held so long o change printf in bpf_detach to distinguish it from the same one in bpfsetdlt Note there are locking issues here related to ioctl processing; they have not been addressed here. Submitted by: Guy Harris <guy@alum.mit.edu> Obtained from: NetBSD (w/ locking modifications)
2003-01-20 19:08:46 +00:00
/*
* Structure to retrieve available DLTs for the interface.
*/
struct bpf_dltlist {
u_int bfl_len; /* number of bfd_list array */
u_int *bfl_list; /* array of DLTs */
};
Change #ifdef KERNEL to #ifdef _KERNEL in the public headers. "KERNEL" is an application space macro and the applications are supposed to be free to use it as they please (but cannot). This is consistant with the other BSD's who made this change quite some time ago. More commits to come.
1999-12-29 04:46:21 +00:00
#ifdef _KERNEL
Introduce support for zero-copy BPF buffering, which reduces the overhead of packet capture by allowing a user process to directly "loan" buffer memory to the kernel rather than using read(2) to explicitly copy data from kernel address space. The user process will issue new BPF ioctls to set the shared memory buffer mode and provide pointers to buffers and their size. The kernel then wires and maps the pages into kernel address space using sf_buf(9), which on supporting architectures will use the direct map region. The current "buffered" access mode remains the default, and support for zero-copy buffers must, for the time being, be explicitly enabled using a sysctl for the kernel to accept requests to use it. The kernel and user process synchronize use of the buffers with atomic operations, avoiding the need for system calls under load; the user process may use select()/poll()/kqueue() to manage blocking while waiting for network data if the user process is able to consume data faster than the kernel generates it. Patchs to libpcap are available to allow libpcap applications to transparently take advantage of this support. Detailed information on the new API may be found in bpf(4), including specific atomic operations and memory barriers required to synchronize buffer use safely. These changes modify the base BPF implementation to (roughly) abstrac the current buffer model, allowing the new shared memory model to be added, and add new monitoring statistics for netstat to print. The implementation, with the exception of some monitoring hanges that break the netstat monitoring ABI for BPF, will be MFC'd. Zerocopy bpf buffers are still considered experimental are disabled by default. To experiment with this new facility, adjust the net.bpf.zerocopy_enable sysctl variable to 1. Changes to libpcap will be made available as a patch for the time being, and further refinements to the implementation are expected. Sponsored by: Seccuris Inc. In collaboration with: rwatson Tested by: pwood, gallatin MFC after: 4 months [1] [1] Certain portions will probably not be MFCed, specifically things that can break the monitoring ABI.
2008-03-24 13:49:17 +00:00
#ifdef MALLOC_DECLARE
MALLOC_DECLARE(M_BPF);
#endif
#ifdef SYSCTL_DECL
SYSCTL_DECL(_net_bpf);
#endif
/*
* Rotate the packet buffers in descriptor d. Move the store buffer into the
Remove two unnecessary sleeps from the hot path in bpf(4). The first one never triggers because bpf_canfreebuf() can only be true for zero-copy buffers and zero-copy buffers are not read with read(2). The second also never triggers, because we check the free buffer before calling ROTATE_BUFFERS(). If the hold buffer is in use the free buffer will be NULL and there is nothing else to do besides drop the packet. If the free buffer isn't NULL the hold buffer _is_ free and it is safe to rotate the buffers. Update the comment in ROTATE_BUFFERS macro to match the logic described here. While here fix a few typos in comments. MFC after: 2 weeks Sponsored by: Rubicon Communications (Netgate)
2015-07-31 21:43:27 +00:00
* hold slot, and the free buffer into the store slot. Zero the length of the
* new store buffer. Descriptor lock should be held. One must be careful to
* not rotate the buffers twice, i.e. if fbuf != NULL.
Introduce support for zero-copy BPF buffering, which reduces the overhead of packet capture by allowing a user process to directly "loan" buffer memory to the kernel rather than using read(2) to explicitly copy data from kernel address space. The user process will issue new BPF ioctls to set the shared memory buffer mode and provide pointers to buffers and their size. The kernel then wires and maps the pages into kernel address space using sf_buf(9), which on supporting architectures will use the direct map region. The current "buffered" access mode remains the default, and support for zero-copy buffers must, for the time being, be explicitly enabled using a sysctl for the kernel to accept requests to use it. The kernel and user process synchronize use of the buffers with atomic operations, avoiding the need for system calls under load; the user process may use select()/poll()/kqueue() to manage blocking while waiting for network data if the user process is able to consume data faster than the kernel generates it. Patchs to libpcap are available to allow libpcap applications to transparently take advantage of this support. Detailed information on the new API may be found in bpf(4), including specific atomic operations and memory barriers required to synchronize buffer use safely. These changes modify the base BPF implementation to (roughly) abstrac the current buffer model, allowing the new shared memory model to be added, and add new monitoring statistics for netstat to print. The implementation, with the exception of some monitoring hanges that break the netstat monitoring ABI for BPF, will be MFC'd. Zerocopy bpf buffers are still considered experimental are disabled by default. To experiment with this new facility, adjust the net.bpf.zerocopy_enable sysctl variable to 1. Changes to libpcap will be made available as a patch for the time being, and further refinements to the implementation are expected. Sponsored by: Seccuris Inc. In collaboration with: rwatson Tested by: pwood, gallatin MFC after: 4 months [1] [1] Certain portions will probably not be MFCed, specifically things that can break the monitoring ABI.
2008-03-24 13:49:17 +00:00
*/
#define ROTATE_BUFFERS(d) do { \
(d)->bd_hbuf = (d)->bd_sbuf; \
(d)->bd_hlen = (d)->bd_slen; \
(d)->bd_sbuf = (d)->bd_fbuf; \
(d)->bd_slen = 0; \
(d)->bd_fbuf = NULL; \
bpf_bufheld(d); \
} while (0)
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
/*
* Descriptor associated with each attached hardware interface.
Move the definition of struct bpf_if to bpf.c. A couple of fields are still exposed via struct bpf_if_ext so that bpf_peers_present() can be inlined into its callers. However, this change eliminates some type duplication in the resulting CTF container, since otherwise ctfmerge(1) propagates the duplication through all types that contain a struct bpf_if. Differential Revision: https://reviews.freebsd.org/D2319 Reviewed by: melifaro, rpaulo
2015-04-20 22:08:11 +00:00
* Part of this structure is exposed to external callers to speed up
* bpf_peers_present() calls.
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
*/
Move the definition of struct bpf_if to bpf.c. A couple of fields are still exposed via struct bpf_if_ext so that bpf_peers_present() can be inlined into its callers. However, this change eliminates some type duplication in the resulting CTF container, since otherwise ctfmerge(1) propagates the duplication through all types that contain a struct bpf_if. Differential Revision: https://reviews.freebsd.org/D2319 Reviewed by: melifaro, rpaulo
2015-04-20 22:08:11 +00:00
struct bpf_if;
struct bpf_if_ext {
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
LIST_ENTRY(bpf_if) bif_next; /* list of all interfaces */
LIST_HEAD(, bpf_d) bif_dlist; /* descriptor list */
};
Introduce support for zero-copy BPF buffering, which reduces the overhead of packet capture by allowing a user process to directly "loan" buffer memory to the kernel rather than using read(2) to explicitly copy data from kernel address space. The user process will issue new BPF ioctls to set the shared memory buffer mode and provide pointers to buffers and their size. The kernel then wires and maps the pages into kernel address space using sf_buf(9), which on supporting architectures will use the direct map region. The current "buffered" access mode remains the default, and support for zero-copy buffers must, for the time being, be explicitly enabled using a sysctl for the kernel to accept requests to use it. The kernel and user process synchronize use of the buffers with atomic operations, avoiding the need for system calls under load; the user process may use select()/poll()/kqueue() to manage blocking while waiting for network data if the user process is able to consume data faster than the kernel generates it. Patchs to libpcap are available to allow libpcap applications to transparently take advantage of this support. Detailed information on the new API may be found in bpf(4), including specific atomic operations and memory barriers required to synchronize buffer use safely. These changes modify the base BPF implementation to (roughly) abstrac the current buffer model, allowing the new shared memory model to be added, and add new monitoring statistics for netstat to print. The implementation, with the exception of some monitoring hanges that break the netstat monitoring ABI for BPF, will be MFC'd. Zerocopy bpf buffers are still considered experimental are disabled by default. To experiment with this new facility, adjust the net.bpf.zerocopy_enable sysctl variable to 1. Changes to libpcap will be made available as a patch for the time being, and further refinements to the implementation are expected. Sponsored by: Seccuris Inc. In collaboration with: rwatson Tested by: pwood, gallatin MFC after: 4 months [1] [1] Certain portions will probably not be MFCed, specifically things that can break the monitoring ABI.
2008-03-24 13:49:17 +00:00
void bpf_bufheld(struct bpf_d *d);
Remove __P.
2002-03-19 21:54:18 +00:00
int bpf_validate(const struct bpf_insn *, int);
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
void bpf_tap(struct bpf_if *, u_char *, u_int);
void bpf_mtap(struct bpf_if *, struct mbuf *);
o eliminate widespread on-stack mbuf use for bpf by introducing a new bpf_mtap2 routine that does the right thing for an mbuf and a variable-length chunk of data that should be prepended. o while we're sweeping the drivers, use u_int32_t uniformly when when prepending the address family (several places were assuming sizeof(int) was 4) o return M_ASSERTVALID to BPF_MTAP* now that all stack-allocated mbufs have been eliminated; this may better be moved to the bpf routines Reviewed by: arch@ and several others
2003-12-28 03:56:00 +00:00
void bpf_mtap2(struct bpf_if *, void *, u_int, struct mbuf *);
Remove __P.
2002-03-19 21:54:18 +00:00
void bpfattach(struct ifnet *, u_int, u_int);
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
void bpfattach2(struct ifnet *, u_int, u_int, struct bpf_if **);
Remove __P.
2002-03-19 21:54:18 +00:00
void bpfdetach(struct ifnet *);
During if_vmove() we call if_detach_internal() which in turn calls the event handler notifying about interface departure and one of the consumers will detach if_bpf. There is no way for us to re-attach this easily as the DLT and hdrlen are only given on interface creation. Add a function to allow us to query the DLT and hdrlen from a current BPF attachment and after if_attach_internal() manually re-add the if_bpf attachment using these values. Found by panics triggered by nd6 packets running past BPF_MTAP() with no proper if_bpf pointer on the interface. Also add a basic DDB show function to investigate the if_bpf attachment of an interface. Reviewed by: gnn MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D5896
2016-04-11 10:00:38 +00:00
#ifdef VIMAGE
int bpf_get_bp_params(struct bpf_if *, u_int *, u_int *);
#endif
The advent of if_detach, allowing interface removal at runtime, makes it possible for a panic to occur if BPF is in use on the interface at the time of the call to if_detach. This happens because BPF maintains pointers to the struct ifnet describing the interface, which is freed by if_detach. To correct this problem, a new call, bpfdetach, is introduced. bpfdetach locates BPF descriptor references to the interface, and NULLs them. Other BPF code is modified so that discovery of a NULL interface results in ENXIO (already implemented for some calls). Processes blocked on a BPF call will also be woken up so that they can receive ENXIO. Interface drivers that invoke bpfattach and if_detach must be modified to also call bpfattach(ifp) before calling if_detach(ifp). This is relevant for buses that support hot removal, such as pccard and usb. Patches to all effected devices will not be committed, only to if_wi.c, due to testing limitations. To reproduce the crash, load up tcpdump on you favorite pccard ethernet card, and then eject the card. As some pccard drivers do not invoke if_detach(ifp), this bug will not manifest itself for those drivers. Reviewed by: wes
2000-03-19 05:42:34 +00:00
Remove __P.
2002-03-19 21:54:18 +00:00
void bpfilterattach(int);
Revert the previous commit to fix buildworld for now. We have constified 'struct bpf_insn *' for bpf_filter(9) and bpf_validate(9) since r1.19 but they conflict with pcap.h from libpcap.
2008-08-26 16:12:49 +00:00
u_int bpf_filter(const struct bpf_insn *, u_char *, u_int, u_int);
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
static __inline int
bpf_peers_present(struct bpf_if *bpf)
{
Move the definition of struct bpf_if to bpf.c. A couple of fields are still exposed via struct bpf_if_ext so that bpf_peers_present() can be inlined into its callers. However, this change eliminates some type duplication in the resulting CTF container, since otherwise ctfmerge(1) propagates the duplication through all types that contain a struct bpf_if. Differential Revision: https://reviews.freebsd.org/D2319 Reviewed by: melifaro, rpaulo
2015-04-20 22:08:11 +00:00
struct bpf_if_ext *ext;
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
Move the definition of struct bpf_if to bpf.c. A couple of fields are still exposed via struct bpf_if_ext so that bpf_peers_present() can be inlined into its callers. However, this change eliminates some type duplication in the resulting CTF container, since otherwise ctfmerge(1) propagates the duplication through all types that contain a struct bpf_if. Differential Revision: https://reviews.freebsd.org/D2319 Reviewed by: melifaro, rpaulo
2015-04-20 22:08:11 +00:00
ext = (struct bpf_if_ext *)bpf;
if (!LIST_EMPTY(&ext->bif_dlist))
Back out previous two commits, this caused some problems in the namespace resulting in some build failures. Instead, to fix the problem of bpf not being present, check the pointer before dereferencing it. This is a temporary bandaid until we can decide on how we want to handle the bpf code not being present. This will be fixed shortly.
2006-06-03 18:48:14 +00:00
return (1);
Make sure we don't try to dereference the the if_bpf pointer when bpf has not been compiled into the the kernel. Submitted by: benno
2006-06-03 06:37:00 +00:00
return (0);
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
}
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
#define BPF_TAP(_ifp,_pkt,_pktlen) do { \
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
if (bpf_peers_present((_ifp)->if_bpf)) \
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
bpf_tap((_ifp)->if_bpf, (_pkt), (_pktlen)); \
} while (0)
#define BPF_MTAP(_ifp,_m) do { \
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
if (bpf_peers_present((_ifp)->if_bpf)) { \
o eliminate widespread on-stack mbuf use for bpf by introducing a new bpf_mtap2 routine that does the right thing for an mbuf and a variable-length chunk of data that should be prepended. o while we're sweeping the drivers, use u_int32_t uniformly when when prepending the address family (several places were assuming sizeof(int) was 4) o return M_ASSERTVALID to BPF_MTAP* now that all stack-allocated mbufs have been eliminated; this may better be moved to the bpf routines Reviewed by: arch@ and several others
2003-12-28 03:56:00 +00:00
M_ASSERTVALID(_m); \
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
bpf_mtap((_ifp)->if_bpf, (_m)); \
o eliminate widespread on-stack mbuf use for bpf by introducing a new bpf_mtap2 routine that does the right thing for an mbuf and a variable-length chunk of data that should be prepended. o while we're sweeping the drivers, use u_int32_t uniformly when when prepending the address family (several places were assuming sizeof(int) was 4) o return M_ASSERTVALID to BPF_MTAP* now that all stack-allocated mbufs have been eliminated; this may better be moved to the bpf routines Reviewed by: arch@ and several others
2003-12-28 03:56:00 +00:00
} \
} while (0)
#define BPF_MTAP2(_ifp,_data,_dlen,_m) do { \
Fix the following bpf(4) race condition which can result in a panic: (1) bpf peer attaches to interface netif0 (2) Packet is received by netif0 (3) ifp->if_bpf pointer is checked and handed off to bpf (4) bpf peer detaches from netif0 resulting in ifp->if_bpf being initialized to NULL. (5) ifp->if_bpf is dereferenced by bpf machinery (6) Kaboom This race condition likely explains the various different kernel panics reported around sending SIGINT to tcpdump or dhclient processes. But really this race can result in kernel panics anywhere you have frequent bpf attach and detach operations with high packet per second load. Summary of changes: - Remove the bpf interface's "driverp" member - When we attach bpf interfaces, we now set the ifp->if_bpf member to the bpf interface structure. Once this is done, ifp->if_bpf should never be NULL. [1] - Introduce bpf_peers_present function, an inline operation which will do a lockless read bpf peer list associated with the interface. It should be noted that the bpf code will pickup the bpf_interface lock before adding or removing bpf peers. This should serialize the access to the bpf descriptor list, removing the race. - Expose the bpf_if structure in bpf.h so that the bpf_peers_present function can use it. This also removes the struct bpf_if; hack that was there. - Adjust all consumers of the raw if_bpf structure to use bpf_peers_present Now what happens is: (1) Packet is received by netif0 (2) Check to see if bpf descriptor list is empty (3) Pickup the bpf interface lock (4) Hand packet off to process From the attach/detach side: (1) Pickup the bpf interface lock (2) Add/remove from bpf descriptor list Now that we are storing the bpf interface structure with the ifnet, there is is no need to walk the bpf interface list to locate the correct bpf interface. We now simply look up the interface, and initialize the pointer. This has a nice side effect of changing a bpf interface attach operation from O(N) (where N is the number of bpf interfaces), to O(1). [1] From now on, we can no longer check ifp->if_bpf to tell us whether or not we have any bpf peers that might be interested in receiving packets. In collaboration with: sam@ MFC after: 1 month
2006-06-02 19:59:33 +00:00
if (bpf_peers_present((_ifp)->if_bpf)) { \
o eliminate widespread on-stack mbuf use for bpf by introducing a new bpf_mtap2 routine that does the right thing for an mbuf and a variable-length chunk of data that should be prepended. o while we're sweeping the drivers, use u_int32_t uniformly when when prepending the address family (several places were assuming sizeof(int) was 4) o return M_ASSERTVALID to BPF_MTAP* now that all stack-allocated mbufs have been eliminated; this may better be moved to the bpf routines Reviewed by: arch@ and several others
2003-12-28 03:56:00 +00:00
M_ASSERTVALID(_m); \
bpf_mtap2((_ifp)->if_bpf,(_data),(_dlen),(_m)); \
} \
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
} while (0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#endif
/*
* Number of scratch memory words (for BPF_LD|BPF_MEM and BPF_ST).
*/
#define BPF_MEMWORDS 16
Style: s/SYS_EVENTHANDLER_H/_SYS_EVENTHANDLER_H_/g Submitted by: bde
2013-10-28 20:32:05 +00:00
#ifdef _SYS_EVENTHANDLER_H_
Instead of putting ifnet declaration into eventhandler.h, move bpf(4) and vlan(4) related event declarations to bpf.h and if_vlan_var.h. To avoid dependency on eventhandler.h, protect these declarations with ifdef SYS_EVENTHANDLER_H. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-28 07:45:03 +00:00
/* BPF attach/detach events */
struct ifnet;
typedef void (*bpf_track_fn)(void *, struct ifnet *, int /* dlt */,
int /* 1 =>'s attach */);
EVENTHANDLER_DECLARE(bpf_track, bpf_track_fn);
Style: s/SYS_EVENTHANDLER_H/_SYS_EVENTHANDLER_H_/g Submitted by: bde
2013-10-28 20:32:05 +00:00
#endif /* _SYS_EVENTHANDLER_H_ */
Instead of putting ifnet declaration into eventhandler.h, move bpf(4) and vlan(4) related event declarations to bpf.h and if_vlan_var.h. To avoid dependency on eventhandler.h, protect these declarations with ifdef SYS_EVENTHANDLER_H. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-28 07:45:03 +00:00
o add support for multiple link types per interface (e.g. 802.11 and Ethernet) o introduce BPF_TAP and BPF_MTAP macros to hide implementation details and ease code portability o use m_getcl where appropriate Reviewed by: many Approved by: re Obtained from: NetBSD (multiple link type support)
2002-11-14 23:24:13 +00:00
#endif /* _NET_BPF_H_ */
Reference in New Issue Copy Permalink