-sphinx.addnodesdocument)}( rawsourcechildren]( translations LanguagesNode)}(hhh](h pending_xref)}(hhh]docutils.nodesTextChinese (Simplified)}parenthsba attributes}(ids]classes]names]dupnames]backrefs] refdomainstdreftypedoc reftarget*/translations/zh_CN/networking/packet_mmapmodnameN classnameN refexplicitutagnamehhh ubh)}(hhh]hChinese (Traditional)}hh2sbah}(h]h ]h"]h$]h&] refdomainh)reftypeh+ reftarget*/translations/zh_TW/networking/packet_mmapmodnameN classnameN refexplicituh1hhh ubh)}(hhh]hItalian}hhFsbah}(h]h ]h"]h$]h&] refdomainh)reftypeh+ reftarget*/translations/it_IT/networking/packet_mmapmodnameN classnameN refexplicituh1hhh ubh)}(hhh]hJapanese}hhZsbah}(h]h ]h"]h$]h&] refdomainh)reftypeh+ reftarget*/translations/ja_JP/networking/packet_mmapmodnameN classnameN refexplicituh1hhh ubh)}(hhh]hKorean}hhnsbah}(h]h ]h"]h$]h&] refdomainh)reftypeh+ reftarget*/translations/ko_KR/networking/packet_mmapmodnameN classnameN refexplicituh1hhh ubh)}(hhh]hSpanish}hhsbah}(h]h ]h"]h$]h&] refdomainh)reftypeh+ reftarget*/translations/sp_SP/networking/packet_mmapmodnameN classnameN refexplicituh1hhh ubeh}(h]h ]h"]h$]h&]current_languageEnglishuh1h hh _documenthsourceNlineNubhcomment)}(h SPDX-License-Identifier: GPL-2.0h]h SPDX-License-Identifier: GPL-2.0}hhsbah}(h]h ]h"]h$]h&] xml:spacepreserveuh1hhhhhhD/var/lib/git/docbuild/linux/Documentation/networking/packet_mmap.rsthKubhsection)}(hhh](htitle)}(h Packet MMAPh]h Packet MMAP}(hhhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhhhhhKubh)}(hhh](h)}(hAbstracth]hAbstract}(hhhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhhhhhKubh paragraph)}(htThis file documents the mmap() facility available with the PACKET socket interface. This type of sockets is used forh]htThis file documents the mmap() facility available with the PACKET socket interface. This type of sockets is used for}(hhhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK hhhhubhenumerated_list)}(hhh](h list_item)}(h4capture network traffic with utilities like tcpdump,h]h)}(hhh]h4capture network traffic with utilities like tcpdump,}(hhhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK hhubah}(h]h ]h"]h$]h&]uh1hhhhhhhhNubh)}(hStransmit network traffic, or any other that needs raw access to network interface. h]h)}(hRtransmit network traffic, or any other that needs raw access to network interface.h]hRtransmit network traffic, or any other that needs raw access to network interface.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjubah}(h]h ]h"]h$]h&]uh1hhhhhhhhNubeh}(h]h ]h"]h$]h&]enumtype lowerromanprefixhsuffix)uh1hhhhhhhhK ubh)}(hHowto can be found at:h]hHowto can be found at:}(hj+hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhhhhubh block_quote)}(hUhttps://web.archive.org/web/20220404160947/https://sites.google.com/site/packetmmap/ h]h)}(hThttps://web.archive.org/web/20220404160947/https://sites.google.com/site/packetmmap/h]h reference)}(hjAh]hThttps://web.archive.org/web/20220404160947/https://sites.google.com/site/packetmmap/}(hjEhhhNhNubah}(h]h ]h"]h$]h&]refurijAuh1jChj?ubah}(h]h ]h"]h$]h&]uh1hhhhKhj;ubah}(h]h ]h"]h$]h&]uh1j9hhhKhhhhubhdefinition_list)}(hhh]hdefinition_list_item)}(hePlease send your comments to - Ulisses Alonso Camaró - Johann Baudy h](hterm)}(hPlease send your comments toh]hPlease send your comments to}(hjlhhhNhNubah}(h]h ]h"]h$]h&]uh1jjhhhKhjfubh definition)}(hhh]h bullet_list)}(hhh](h)}(h6Ulisses Alonso Camaró h]h)}(hjh](hUlisses Alonso Camaró <}(hjhhhNhNubjD)}(huaca@i.hate.spam.alumni.uv.esh]huaca@i.hate.spam.alumni.uv.es}(hjhhhNhNubah}(h]h ]h"]h$]h&]refuri$mailto:uaca@i.hate.spam.alumni.uv.esuh1jChjubh>}(hjhhhNhNubeh}(h]h ]h"]h$]h&]uh1hhhhKhjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h Johann Baudy h]h)}(h Johann Baudyh]h Johann Baudy}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjubah}(h]h ]h"]h$]h&]uh1hhjubeh}(h]h ]h"]h$]h&]bullet-uh1jhhhKhj|ubah}(h]h ]h"]h$]h&]uh1jzhjfubeh}(h]h ]h"]h$]h&]uh1jdhhhKhjaubah}(h]h ]h"]h$]h&]uh1j_hhhhhNhNubeh}(h]abstractah ]h"]abstractah$]h&]uh1hhhhhhhhKubh)}(hhh](h)}(hWhy use PACKET_MMAPh]hWhy use PACKET_MMAP}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhKubh)}(hNon PACKET_MMAP capture process (plain AF_PACKET) is very inefficient. It uses very limited buffers and requires one system call to capture each packet, it requires two if you want to get packet's timestamp (like libpcap always does).h]hNon PACKET_MMAP capture process (plain AF_PACKET) is very inefficient. It uses very limited buffers and requires one system call to capture each packet, it requires two if you want to get packet’s timestamp (like libpcap always does).}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubh)}(hXOn the other hand PACKET_MMAP is very efficient. PACKET_MMAP provides a size configurable circular buffer mapped in user space that can be used to either send or receive packets. This way reading packets just needs to wait for them, most of the time there is no need to issue a single system call. Concerning transmission, multiple packets can be sent through one system call to get the highest bandwidth. By using a shared buffer between the kernel and the user also has the benefit of minimizing packet copies.h]hXOn the other hand PACKET_MMAP is very efficient. PACKET_MMAP provides a size configurable circular buffer mapped in user space that can be used to either send or receive packets. This way reading packets just needs to wait for them, most of the time there is no need to issue a single system call. Concerning transmission, multiple packets can be sent through one system call to get the highest bandwidth. By using a shared buffer between the kernel and the user also has the benefit of minimizing packet copies.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK!hjhhubh)}(hX9It's fine to use PACKET_MMAP to improve the performance of the capture and transmission process, but it isn't everything. At least, if you are capturing at high speeds (this is relative to the cpu speed), you should check if the device driver of your network interface card supports some sort of interrupt load mitigation or (even better) if it supports NAPI, also make sure it is enabled. For transmission, check the MTU (Maximum Transmission Unit) used and supported by devices of your network. CPU IRQ pinning of your network interface card can also be an advantage.h]hX=It’s fine to use PACKET_MMAP to improve the performance of the capture and transmission process, but it isn’t everything. At least, if you are capturing at high speeds (this is relative to the cpu speed), you should check if the device driver of your network interface card supports some sort of interrupt load mitigation or (even better) if it supports NAPI, also make sure it is enabled. For transmission, check the MTU (Maximum Transmission Unit) used and supported by devices of your network. CPU IRQ pinning of your network interface card can also be an advantage.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK)hjhhubeh}(h]why-use-packet-mmapah ]h"]why use packet_mmapah$]h&]uh1hhhhhhhhKubh)}(hhh](h)}(h,How to use mmap() to improve capture processh]h,How to use mmap() to improve capture process}(hj/hhhNhNubah}(h]h ]h"]h$]h&]uh1hhj,hhhhhK3ubh)}(hFrom the user standpoint, you should use the higher level libpcap library, which is a de facto standard, portable across nearly all operating systems including Win32.h]hFrom the user standpoint, you should use the higher level libpcap library, which is a de facto standard, portable across nearly all operating systems including Win32.}(hj=hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK5hj,hhubh)}(hPacket MMAP support was integrated into libpcap around the time of version 1.3.0; TPACKET_V3 support was added in version 1.5.0h]hPacket MMAP support was integrated into libpcap around the time of version 1.3.0; TPACKET_V3 support was added in version 1.5.0}(hjKhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK9hj,hhubeh}(h]*how-to-use-mmap-to-improve-capture-processah ]h"],how to use mmap() to improve capture processah$]h&]uh1hhhhhhhhK3ubh)}(hhh](h)}(h5How to use mmap() directly to improve capture processh]h5How to use mmap() directly to improve capture process}(hjdhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjahhhhhK=ubh)}(hZFrom the system calls stand point, the use of PACKET_MMAP involves the following process::h]hYFrom the system calls stand point, the use of PACKET_MMAP involves the following process:}(hjrhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK?hjahhubh literal_block)}(hX[setup] socket() -------> creation of the capture socket setsockopt() ---> allocation of the circular buffer (ring) option: PACKET_RX_RING mmap() ---------> mapping of the allocated buffer to the user process [capture] poll() ---------> to wait for incoming packets [shutdown] close() --------> destruction of the capture socket and deallocation of all associated resources.h]hX[setup] socket() -------> creation of the capture socket setsockopt() ---> allocation of the circular buffer (ring) option: PACKET_RX_RING mmap() ---------> mapping of the allocated buffer to the user process [capture] poll() ---------> to wait for incoming packets [shutdown] close() --------> destruction of the capture socket and deallocation of all associated resources.}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKChjahhubh)}(hksocket creation and destruction is straight forward, and is done the same way with or without PACKET_MMAP::h]hjsocket creation and destruction is straight forward, and is done the same way with or without PACKET_MMAP:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKPhjahhubj)}(h3int fd = socket(PF_PACKET, mode, htons(ETH_P_ALL));h]h3int fd = socket(PF_PACKET, mode, htons(ETH_P_ALL));}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKShjahhubh)}(hwhere mode is SOCK_RAW for the raw interface were link level information can be captured or SOCK_DGRAM for the cooked interface where link level information capture is not supported and a link level pseudo-header is provided by the kernel.h]hwhere mode is SOCK_RAW for the raw interface were link level information can be captured or SOCK_DGRAM for the cooked interface where link level information capture is not supported and a link level pseudo-header is provided by the kernel.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKUhjahhubh)}(haThe destruction of the socket and all associated resources is done by a simple call to close(fd).h]haThe destruction of the socket and all associated resources is done by a simple call to close(fd).}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK[hjahhubh)}(hSimilarly as without PACKET_MMAP, it is possible to use one socket for capture and transmission. This can be done by mapping the allocated RX and TX buffer ring with a single mmap() call. See "Mapping and use of the circular buffer (ring)".h]hSimilarly as without PACKET_MMAP, it is possible to use one socket for capture and transmission. This can be done by mapping the allocated RX and TX buffer ring with a single mmap() call. See “Mapping and use of the circular buffer (ring)”.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK^hjahhubh)}(hNext I will describe PACKET_MMAP settings and its constraints, also the mapping of the circular buffer in the user process and the use of this buffer.h]hNext I will describe PACKET_MMAP settings and its constraints, also the mapping of the circular buffer in the user process and the use of this buffer.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKchjahhubeh}(h]3how-to-use-mmap-directly-to-improve-capture-processah ]h"]5how to use mmap() directly to improve capture processah$]h&]uh1hhhhhhhhK=ubh)}(hhh](h)}(h:How to use mmap() directly to improve transmission processh]h:How to use mmap() directly to improve transmission process}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhKhubh)}(h;Transmission process is similar to capture as shown below::h]h:Transmission process is similar to capture as shown below:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKihjhhubj)}(hXr[setup] socket() -------> creation of the transmission socket setsockopt() ---> allocation of the circular buffer (ring) option: PACKET_TX_RING bind() ---------> bind transmission socket with a network interface mmap() ---------> mapping of the allocated buffer to the user process [transmission] poll() ---------> wait for free packets (optional) send() ---------> send all packets that are set as ready in the ring The flag MSG_DONTWAIT can be used to return before end of transfer. [shutdown] close() --------> destruction of the transmission socket and deallocation of all associated resources.h]hXr[setup] socket() -------> creation of the transmission socket setsockopt() ---> allocation of the circular buffer (ring) option: PACKET_TX_RING bind() ---------> bind transmission socket with a network interface mmap() ---------> mapping of the allocated buffer to the user process [transmission] poll() ---------> wait for free packets (optional) send() ---------> send all packets that are set as ready in the ring The flag MSG_DONTWAIT can be used to return before end of transfer. [shutdown] close() --------> destruction of the transmission socket and deallocation of all associated resources.}hj sbah}(h]h ]h"]h$]h&]hhuh1jhhhKkhjhhubh)}(hSocket creation and destruction is also straight forward, and is done the same way as in capturing described in the previous paragraph::h]hSocket creation and destruction is also straight forward, and is done the same way as in capturing described in the previous paragraph:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhK{hjhhubj)}(h$int fd = socket(PF_PACKET, mode, 0);h]h$int fd = socket(PF_PACKET, mode, 0);}hj'sbah}(h]h ]h"]h$]h&]hhuh1jhhhK~hjhhubh)}(hXThe protocol can optionally be 0 in case we only want to transmit via this socket, which avoids an expensive call to packet_rcv(). In this case, you also need to bind(2) the TX_RING with sll_protocol = 0 set. Otherwise, htons(ETH_P_ALL) or any other protocol, for example.h]hXThe protocol can optionally be 0 in case we only want to transmit via this socket, which avoids an expensive call to packet_rcv(). In this case, you also need to bind(2) the TX_RING with sll_protocol = 0 set. Otherwise, htons(ETH_P_ALL) or any other protocol, for example.}(hj5hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubh)}(hBinding the socket to your network interface is mandatory (with zero copy) to know the header size of frames used in the circular buffer.h]hBinding the socket to your network interface is mandatory (with zero copy) to know the header size of frames used in the circular buffer.}(hjChhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubh)}(h+As capture, each frame contains two parts::h]h*As capture, each frame contains two parts:}(hjQhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(hX& -------------------- | struct tpacket_hdr | Header. It contains the status of | | of this frame |--------------------| | data buffer | . . Data that will be sent over the network interface. . . -------------------- bind() associates the socket to your network interface thanks to sll_ifindex parameter of struct sockaddr_ll. Initialization example:: struct sockaddr_ll my_addr; struct ifreq s_ifr; ... strscpy_pad (s_ifr.ifr_name, "eth0", sizeof(s_ifr.ifr_name)); /* get interface index of eth0 */ ioctl(this->socket, SIOCGIFINDEX, &s_ifr); /* fill sockaddr_ll struct to prepare binding */ my_addr.sll_family = AF_PACKET; my_addr.sll_protocol = htons(ETH_P_ALL); my_addr.sll_ifindex = s_ifr.ifr_ifindex; /* bind socket to eth0 */ bind(this->socket, (struct sockaddr *)&my_addr, sizeof(struct sockaddr_ll)); A complete tutorial is available at: https://web.archive.org/web/20220404160947/https://sites.google.com/site/packetmmap/h]hX& -------------------- | struct tpacket_hdr | Header. It contains the status of | | of this frame |--------------------| | data buffer | . . Data that will be sent over the network interface. . . -------------------- bind() associates the socket to your network interface thanks to sll_ifindex parameter of struct sockaddr_ll. Initialization example:: struct sockaddr_ll my_addr; struct ifreq s_ifr; ... strscpy_pad (s_ifr.ifr_name, "eth0", sizeof(s_ifr.ifr_name)); /* get interface index of eth0 */ ioctl(this->socket, SIOCGIFINDEX, &s_ifr); /* fill sockaddr_ll struct to prepare binding */ my_addr.sll_family = AF_PACKET; my_addr.sll_protocol = htons(ETH_P_ALL); my_addr.sll_ifindex = s_ifr.ifr_ifindex; /* bind socket to eth0 */ bind(this->socket, (struct sockaddr *)&my_addr, sizeof(struct sockaddr_ll)); A complete tutorial is available at: https://web.archive.org/web/20220404160947/https://sites.google.com/site/packetmmap/}hj_sbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(h)By default, the user should put data at::h]h(By default, the user should put data at:}(hjmhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(h8frame base + TPACKET_HDRLEN - sizeof(struct sockaddr_ll)h]h8frame base + TPACKET_HDRLEN - sizeof(struct sockaddr_ll)}hj{sbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(hqSo, whatever you choose for the socket mode (SOCK_DGRAM or SOCK_RAW), the beginning of the user data will be at::h]hpSo, whatever you choose for the socket mode (SOCK_DGRAM or SOCK_RAW), the beginning of the user data will be at:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(h6frame base + TPACKET_ALIGN(sizeof(struct tpacket_hdr))h]h6frame base + TPACKET_ALIGN(sizeof(struct tpacket_hdr))}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(hX4If you wish to put user data at a custom offset from the beginning of the frame (for payload alignment with SOCK_RAW mode for instance) you can set tp_net (with SOCK_DGRAM) or tp_mac (with SOCK_RAW). In order to make this work it must be enabled previously with setsockopt() and the PACKET_TX_HAS_OFF option.h]hX4If you wish to put user data at a custom offset from the beginning of the frame (for payload alignment with SOCK_RAW mode for instance) you can set tp_net (with SOCK_DGRAM) or tp_mac (with SOCK_RAW). In order to make this work it must be enabled previously with setsockopt() and the PACKET_TX_HAS_OFF option.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubeh}(h]8how-to-use-mmap-directly-to-improve-transmission-processah ]h"]:how to use mmap() directly to improve transmission processah$]h&]uh1hhhhhhhhKhubh)}(hhh](h)}(hPACKET_MMAP settingsh]hPACKET_MMAP settings}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhKubh)}(hBTo setup PACKET_MMAP from user level code is done with a call likeh]hBTo setup PACKET_MMAP from user level code is done with a call like}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj:)}(h- Capture process:: setsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req, sizeof(req)) - Transmission process:: setsockopt(fd, SOL_PACKET, PACKET_TX_RING, (void *) &req, sizeof(req)) h]j)}(hhh](h)}(h\Capture process:: setsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req, sizeof(req)) h](h)}(hCapture process::h]hCapture process:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjubj)}(hFsetsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req, sizeof(req))h]hFsetsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req, sizeof(req))}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjubeh}(h]h ]h"]h$]h&]uh1hhjubh)}(haTransmission process:: setsockopt(fd, SOL_PACKET, PACKET_TX_RING, (void *) &req, sizeof(req)) h](h)}(hTransmission process::h]hTransmission process:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjubj)}(hFsetsockopt(fd, SOL_PACKET, PACKET_TX_RING, (void *) &req, sizeof(req))h]hFsetsockopt(fd, SOL_PACKET, PACKET_TX_RING, (void *) &req, sizeof(req))}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjubeh}(h]h ]h"]h$]h&]uh1hhjubeh}(h]h ]h"]h$]h&]jjuh1jhhhKhjubah}(h]h ]h"]h$]h&]uh1j9hhhKhjhhubh)}(h~The most significant argument in the previous call is the req parameter, this parameter must to have the following structure::h]h}The most significant argument in the previous call is the req parameter, this parameter must to have the following structure:}(hj9hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(hXstruct tpacket_req { unsigned int tp_block_size; /* Minimal size of contiguous block */ unsigned int tp_block_nr; /* Number of blocks */ unsigned int tp_frame_size; /* Size of frame */ unsigned int tp_frame_nr; /* Total number of frames */ };h]hXstruct tpacket_req { unsigned int tp_block_size; /* Minimal size of contiguous block */ unsigned int tp_block_nr; /* Number of blocks */ unsigned int tp_frame_size; /* Size of frame */ unsigned int tp_frame_nr; /* Total number of frames */ };}hjGsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(hXThis structure is defined in /usr/include/linux/if_packet.h and establishes a circular buffer (ring) of unswappable memory. Being mapped in the capture process allows reading the captured frames and related meta-information like timestamps without requiring a system call.h]hXThis structure is defined in /usr/include/linux/if_packet.h and establishes a circular buffer (ring) of unswappable memory. Being mapped in the capture process allows reading the captured frames and related meta-information like timestamps without requiring a system call.}(hjUhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubh)}(hFrames are grouped in blocks. Each block is a physically contiguous region of memory and holds tp_block_size/tp_frame_size frames. The total number of blocks is tp_block_nr. Note that tp_frame_nr is a redundant parameter because::h]hFrames are grouped in blocks. Each block is a physically contiguous region of memory and holds tp_block_size/tp_frame_size frames. The total number of blocks is tp_block_nr. Note that tp_frame_nr is a redundant parameter because:}(hjchhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(h.frames_per_block = tp_block_size/tp_frame_sizeh]h.frames_per_block = tp_block_size/tp_frame_size}hjqsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(hEindeed, packet_set_ring checks that the following condition is true::h]hDindeed, packet_set_ring checks that the following condition is true:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(h-frames_per_block * tp_block_nr == tp_frame_nrh]h-frames_per_block * tp_block_nr == tp_frame_nr}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(h0Lets see an example, with the following values::h]h/Lets see an example, with the following values:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(hItp_block_size= 4096 tp_frame_size= 2048 tp_block_nr = 4 tp_frame_nr = 8h]hItp_block_size= 4096 tp_frame_size= 2048 tp_block_nr = 4 tp_frame_nr = 8}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(h,we will get the following buffer structure::h]h+we will get the following buffer structure:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubj)}(hXn block #1 block #2 +---------+---------+ +---------+---------+ | frame 1 | frame 2 | | frame 3 | frame 4 | +---------+---------+ +---------+---------+ block #3 block #4 +---------+---------+ +---------+---------+ | frame 5 | frame 6 | | frame 7 | frame 8 | +---------+---------+ +---------+---------+h]hXn block #1 block #2 +---------+---------+ +---------+---------+ | frame 1 | frame 2 | | frame 3 | frame 4 | +---------+---------+ +---------+---------+ block #3 block #4 +---------+---------+ +---------+---------+ | frame 5 | frame 6 | | frame 7 | frame 8 | +---------+---------+ +---------+---------+}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhKhjhhubh)}(hXGA frame can be of any size with the only condition it can fit in a block. A block can only hold an integer number of frames, or in other words, a frame cannot be spawned across two blocks, so there are some details you have to take into account when choosing the frame_size. See "Mapping and use of the circular buffer (ring)".h]hXKA frame can be of any size with the only condition it can fit in a block. A block can only hold an integer number of frames, or in other words, a frame cannot be spawned across two blocks, so there are some details you have to take into account when choosing the frame_size. See “Mapping and use of the circular buffer (ring)”.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubeh}(h]packet-mmap-settingsah ]h"]packet_mmap settingsah$]h&]uh1hhhhhhhhKubh)}(hhh](h)}(hPACKET_MMAP setting constraintsh]hPACKET_MMAP setting constraints}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhKubh)}(hIn kernel versions prior to 2.4.26 (for the 2.4 branch) and 2.6.5 (2.6 branch), the PACKET_MMAP buffer could hold only 32768 frames in a 32 bit architecture or 16384 in a 64 bit architecture.h]hIn kernel versions prior to 2.4.26 (for the 2.4 branch) and 2.6.5 (2.6 branch), the PACKET_MMAP buffer could hold only 32768 frames in a 32 bit architecture or 16384 in a 64 bit architecture.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhKhjhhubh)}(hhh](h)}(hBlock size limith]hBlock size limit}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMubh)}(hX$As stated earlier, each block is a contiguous physical region of memory. These memory regions are allocated with calls to the __get_free_pages() function. As the name indicates, this function allocates pages of memory, and the second argument is "order" or a power of two number of pages, that is (for PAGE_SIZE == 4096) order=0 ==> 4096 bytes, order=1 ==> 8192 bytes, order=2 ==> 16384 bytes, etc. The maximum size of a region allocated by __get_free_pages is determined by the MAX_PAGE_ORDER macro. More precisely the limit can be calculated as::h]hX'As stated earlier, each block is a contiguous physical region of memory. These memory regions are allocated with calls to the __get_free_pages() function. As the name indicates, this function allocates pages of memory, and the second argument is “order” or a power of two number of pages, that is (for PAGE_SIZE == 4096) order=0 ==> 4096 bytes, order=1 ==> 8192 bytes, order=2 ==> 16384 bytes, etc. The maximum size of a region allocated by __get_free_pages is determined by the MAX_PAGE_ORDER macro. More precisely the limit can be calculated as:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj)}(hPAGE_SIZE << MAX_PAGE_ORDER In a i386 architecture PAGE_SIZE is 4096 bytes In a 2.4/i386 kernel MAX_PAGE_ORDER is 10 In a 2.6/i386 kernel MAX_PAGE_ORDER is 11h]hPAGE_SIZE << MAX_PAGE_ORDER In a i386 architecture PAGE_SIZE is 4096 bytes In a 2.4/i386 kernel MAX_PAGE_ORDER is 10 In a 2.6/i386 kernel MAX_PAGE_ORDER is 11}hj'sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjhhubh)}(hqSo get_free_pages can allocate as much as 4MB or 8MB in a 2.4/2.6 kernel respectively, with an i386 architecture.h]hqSo get_free_pages can allocate as much as 4MB or 8MB in a 2.4/2.6 kernel respectively, with an i386 architecture.}(hj5hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hUser space programs can include /usr/include/sys/user.h and /usr/include/linux/mmzone.h to get PAGE_SIZE MAX_PAGE_ORDER declarations.h]hUser space programs can include /usr/include/sys/user.h and /usr/include/linux/mmzone.h to get PAGE_SIZE MAX_PAGE_ORDER declarations.}(hjChhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hUThe pagesize can also be determined dynamically with the getpagesize (2) system call.h]hUThe pagesize can also be determined dynamically with the getpagesize (2) system call.}(hjQhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubeh}(h]block-size-limitah ]h"]block size limitah$]h&]uh1hhjhhhhhMubh)}(hhh](h)}(hBlock number limith]hBlock number limit}(hjjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjghhhhhMubh)}(hsTo understand the constraints of PACKET_MMAP, we have to see the structure used to hold the pointers to each block.h]hsTo understand the constraints of PACKET_MMAP, we have to see the structure used to hold the pointers to each block.}(hjxhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM hjghhubh)}(hCurrently, this structure is a dynamically allocated vector with kmalloc called pg_vec, its size limits the number of blocks that can be allocated::h]hCurrently, this structure is a dynamically allocated vector with kmalloc called pg_vec, its size limits the number of blocks that can be allocated:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM#hjghhubj)}(h+---+---+---+---+ | x | x | x | x | +---+---+---+---+ | | | | | | | v | | v block #4 | v block #3 v block #2 block #1h]h+---+---+---+---+ | x | x | x | x | +---+---+---+---+ | | | | | | | v | | v block #4 | v block #3 v block #2 block #1}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhM&hjghhubh)}(hX%kmalloc allocates any number of bytes of physically contiguous memory from a pool of pre-determined sizes. This pool of memory is maintained by the slab allocator which is at the end the responsible for doing the allocation and hence which imposes the maximum memory that kmalloc can allocate.h]hX%kmalloc allocates any number of bytes of physically contiguous memory from a pool of pre-determined sizes. This pool of memory is maintained by the slab allocator which is at the end the responsible for doing the allocation and hence which imposes the maximum memory that kmalloc can allocate.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM0hjghhubh)}(hIn a 2.4/2.6 kernel and the i386 architecture, the limit is 131072 bytes. The predetermined sizes that kmalloc uses can be checked in the "size-" entries of /proc/slabinfoh]hIn a 2.4/2.6 kernel and the i386 architecture, the limit is 131072 bytes. The predetermined sizes that kmalloc uses can be checked in the “size-” entries of /proc/slabinfo}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM5hjghhubh)}(hcIn a 32 bit architecture, pointers are 4 bytes long, so the total number of pointers to blocks is::h]hbIn a 32 bit architecture, pointers are 4 bytes long, so the total number of pointers to blocks is:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM9hjghhubj)}(h131072/4 = 32768 blocksh]h131072/4 = 32768 blocks}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhM<hjghhubeh}(h]block-number-limitah ]h"]block number limitah$]h&]uh1hhjhhhhhMubeh}(h]packet-mmap-setting-constraintsah ]h"]packet_mmap setting constraintsah$]h&]uh1hhhhhhhhKubh)}(hhh](h)}(h"PACKET_MMAP buffer size calculatorh]h"PACKET_MMAP buffer size calculator}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhM?ubh)}(h Definitions:h]h Definitions:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMAhjhhubhtable)}(hhh]htgroup)}(hhh](hcolspec)}(hhh]h}(h]h ]h"]h$]h&]colwidthKuh1jhjubj)}(hhh]h}(h]h ]h"]h$]h&]colwidthK@uh1jhjubhtbody)}(hhh](hrow)}(hhh](hentry)}(hhh]h)}(h h]h }(hj8hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMDhj5ubah}(h]h ]h"]h$]h&]uh1j3hj0ubj4)}(hhh]h)}(hBis the maximum size of allocable with kmalloc (see /proc/slabinfo)h]hBis the maximum size of allocable with kmalloc (see /proc/slabinfo)}(hjOhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMDhjLubah}(h]h ]h"]h$]h&]uh1j3hj0ubeh}(h]h ]h"]h$]h&]uh1j.hj+ubj/)}(hhh](j4)}(hhh]h)}(hh]h}(hjohhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMFhjlubah}(h]h ]h"]h$]h&]uh1j3hjiubj4)}(hhh]h)}(h1depends on the architecture -- ``sizeof(void *)``h](hdepends on the architecture -- }(hjhhhNhNubhliteral)}(h``sizeof(void *)``h]hsizeof(void *)}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1jhjubeh}(h]h ]h"]h$]h&]uh1hhhhMFhjubah}(h]h ]h"]h$]h&]uh1j3hjiubeh}(h]h ]h"]h$]h&]uh1j.hj+ubj/)}(hhh](j4)}(hhh]h)}(h h]h }(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMGhjubah}(h]h ]h"]h$]h&]uh1j3hjubj4)}(hhh]h)}(h;depends on the architecture -- PAGE_SIZE or getpagesize (2)h]h;depends on the architecture -- PAGE_SIZE or getpagesize (2)}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMGhjubah}(h]h ]h"]h$]h&]uh1j3hjubeh}(h]h ]h"]h$]h&]uh1j.hj+ubj/)}(hhh](j4)}(hhh]h)}(h h]h }(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMHhjubah}(h]h ]h"]h$]h&]uh1j3hjubj4)}(hhh]h)}(h(is the value defined with MAX_PAGE_ORDERh]h(is the value defined with MAX_PAGE_ORDER}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMHhjubah}(h]h ]h"]h$]h&]uh1j3hjubeh}(h]h ]h"]h$]h&]uh1j.hj+ubj/)}(hhh](j4)}(hhh]h)}(h h]h }(hj$hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMIhj!ubah}(h]h ]h"]h$]h&]uh1j3hjubj4)}(hhh]h)}(h@it's an upper bound of frame's capture size (more on this later)h]hDit’s an upper bound of frame’s capture size (more on this later)}(hj;hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMIhj8ubah}(h]h ]h"]h$]h&]uh1j3hjubeh}(h]h ]h"]h$]h&]uh1j.hj+ubeh}(h]h ]h"]h$]h&]uh1j)hjubeh}(h]h ]h"]h$]h&]colsKuh1jhj ubah}(h]h ]h"]h$]h&]uh1j hjhhhhhNubh)}(h'from these definitions we will derive::h]h&from these definitions we will derive:}(hjhhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMLhjhhubj)}(hS = / = << h]hS = / = << }hjvsbah}(h]h ]h"]h$]h&]hhuh1jhhhMNhjhhubh)}(hso, the max buffer size is::h]hso, the max buffer size is:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMQhjhhubj)}(h * h]h * }hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhMShjhhubh)}(hand, the number of frames be::h]hand, the number of frames be:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMUhjhhubj)}(h, * / h]h, * / }hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhMWhjhhubh)}(hWSuppose the following parameters, which apply for 2.6 kernel and an i386 architecture::h]hVSuppose the following parameters, which apply for 2.6 kernel and an i386 architecture:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMYhjhhubj)}(h[ = 131072 bytes = 4 bytes = 4096 bytes = 11h]h[ = 131072 bytes = 4 bytes = 4096 bytes = 11}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhM\hjhhubh)}(hIand a value for of 2048 bytes. These parameters will yield::h]hHand a value for of 2048 bytes. These parameters will yield:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMahjhhubj)}(hK = 131072/4 = 32768 blocks = 4096 << 11 = 8 MiB.h]hK = 131072/4 = 32768 blocks = 4096 << 11 = 8 MiB.}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhMchjhhubh)}(hkand hence the buffer will have a 262144 MiB size. So it can hold 262144 MiB / 2048 bytes = 134217728 framesh]hkand hence the buffer will have a 262144 MiB size. So it can hold 262144 MiB / 2048 bytes = 134217728 frames}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMfhjhhubh)}(hActually, this buffer size is not possible with an i386 architecture. Remember that the memory is allocated in kernel space, in the case of an i386 kernel's memory size is limited to 1GiB.h]hActually, this buffer size is not possible with an i386 architecture. Remember that the memory is allocated in kernel space, in the case of an i386 kernel’s memory size is limited to 1GiB.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMihjhhubh)}(hXAll memory allocations are not freed until the socket is closed. The memory allocations are done with GFP_KERNEL priority, this basically means that the allocation can wait and swap other process' memory in order to allocate the necessary memory, so normally limits can be reached.h]hXAll memory allocations are not freed until the socket is closed. The memory allocations are done with GFP_KERNEL priority, this basically means that the allocation can wait and swap other process’ memory in order to allocate the necessary memory, so normally limits can be reached.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMmhjhhubh)}(hhh](h)}(hOther constraintsh]hOther constraints}(hj!hhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMsubh)}(hXaIf you check the source code you will see that what I draw here as a frame is not only the link level frame. At the beginning of each frame there is a header called struct tpacket_hdr used in PACKET_MMAP to hold link level's frame meta information like timestamp. So what we draw here a frame it's really the following (from include/linux/if_packet.h)::h]hXdIf you check the source code you will see that what I draw here as a frame is not only the link level frame. At the beginning of each frame there is a header called struct tpacket_hdr used in PACKET_MMAP to hold link level’s frame meta information like timestamp. So what we draw here a frame it’s really the following (from include/linux/if_packet.h):}(hj/hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMuhjhhubj)}(hX/* Frame structure: - Start. Frame must be aligned to TPACKET_ALIGNMENT=16 - struct tpacket_hdr - pad to TPACKET_ALIGNMENT=16 - struct sockaddr_ll - Gap, chosen so that packet data (Start+tp_net) aligns to TPACKET_ALIGNMENT=16 - Start+tp_mac: [ Optional MAC header ] - Start+tp_net: Packet data, aligned to TPACKET_ALIGNMENT=16. - Pad to align to TPACKET_ALIGNMENT=16 */h]hX/* Frame structure: - Start. Frame must be aligned to TPACKET_ALIGNMENT=16 - struct tpacket_hdr - pad to TPACKET_ALIGNMENT=16 - struct sockaddr_ll - Gap, chosen so that packet data (Start+tp_net) aligns to TPACKET_ALIGNMENT=16 - Start+tp_mac: [ Optional MAC header ] - Start+tp_net: Packet data, aligned to TPACKET_ALIGNMENT=16. - Pad to align to TPACKET_ALIGNMENT=16 */}hj=sbah}(h]h ]h"]h$]h&]hhuh1jhhhM{hjhhubh)}(h@The following are conditions that are checked in packet_set_ringh]h@The following are conditions that are checked in packet_set_ring}(hjKhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj:)}(h- tp_block_size must be a multiple of PAGE_SIZE (1) - tp_frame_size must be greater than TPACKET_HDRLEN (obvious) - tp_frame_size must be a multiple of TPACKET_ALIGNMENT - tp_frame_nr must be exactly frames_per_block*tp_block_nr h]j)}(hhh](h)}(h1tp_block_size must be a multiple of PAGE_SIZE (1)h]h)}(hjbh]h1tp_block_size must be a multiple of PAGE_SIZE (1)}(hjdhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj`ubah}(h]h ]h"]h$]h&]uh1hhj]ubh)}(h;tp_frame_size must be greater than TPACKET_HDRLEN (obvious)h]h)}(hjyh]h;tp_frame_size must be greater than TPACKET_HDRLEN (obvious)}(hj{hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjwubah}(h]h ]h"]h$]h&]uh1hhj]ubh)}(h5tp_frame_size must be a multiple of TPACKET_ALIGNMENTh]h)}(hjh]h5tp_frame_size must be a multiple of TPACKET_ALIGNMENT}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjubah}(h]h ]h"]h$]h&]uh1hhj]ubh)}(h;tp_frame_nr must be exactly frames_per_block*tp_block_nr h]h)}(h:tp_frame_nr must be exactly frames_per_block*tp_block_nrh]h:tp_frame_nr must be exactly frames_per_block*tp_block_nr}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjubah}(h]h ]h"]h$]h&]uh1hhj]ubeh}(h]h ]h"]h$]h&]jjuh1jhhhMhjYubah}(h]h ]h"]h$]h&]uh1j9hhhMhjhhubh)}(haNote that tp_block_size should be chosen to be a power of two or there will be a waste of memory.h]haNote that tp_block_size should be chosen to be a power of two or there will be a waste of memory.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubeh}(h]other-constraintsah ]h"]other constraintsah$]h&]uh1hhjhhhhhMsubh)}(hhh](h)}(h-Mapping and use of the circular buffer (ring)h]h-Mapping and use of the circular buffer (ring)}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMubh)}(hXThe mapping of the buffer in the user process is done with the conventional mmap function. Even the circular buffer is compound of several physically discontiguous blocks of memory, they are contiguous to the user space, hence just one call to mmap is needed::h]hXThe mapping of the buffer in the user process is done with the conventional mmap function. Even the circular buffer is compound of several physically discontiguous blocks of memory, they are contiguous to the user space, hence just one call to mmap is needed:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj)}(h7mmap(0, size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);h]h7mmap(0, size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjhhubh)}(hIf tp_frame_size is a divisor of tp_block_size frames will be contiguously spaced by tp_frame_size bytes. If not, each tp_block_size/tp_frame_size frames there will be a gap between the frames. This is because a frame cannot be spawn across two blocks.h]hIf tp_frame_size is a divisor of tp_block_size frames will be contiguously spaced by tp_frame_size bytes. If not, each tp_block_size/tp_frame_size frames there will be a gap between the frames. This is because a frame cannot be spawn across two blocks.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hTo use one socket for capture and transmission, the mapping of both the RX and TX buffer ring has to be done with one call to mmap::h]hTo use one socket for capture and transmission, the mapping of both the RX and TX buffer ring has to be done with one call to mmap:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj)}(h... setsockopt(fd, SOL_PACKET, PACKET_RX_RING, &foo, sizeof(foo)); setsockopt(fd, SOL_PACKET, PACKET_TX_RING, &bar, sizeof(bar)); ... rx_ring = mmap(0, size * 2, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0); tx_ring = rx_ring + size;h]h... setsockopt(fd, SOL_PACKET, PACKET_RX_RING, &foo, sizeof(foo)); setsockopt(fd, SOL_PACKET, PACKET_TX_RING, &bar, sizeof(bar)); ... rx_ring = mmap(0, size * 2, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0); tx_ring = rx_ring + size;}hj( sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjhhubh)}(hRRX must be the first as the kernel maps the TX ring memory right after the RX one.h]hRRX must be the first as the kernel maps the TX ring memory right after the RX one.}(hj6 hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hAt the beginning of each frame there is an status field (see struct tpacket_hdr). If this field is 0 means that the frame is ready to be used for the kernel, If not, there is a frame the user can read and the following flags apply:h]hAt the beginning of each frame there is an status field (see struct tpacket_hdr). If this field is 0 means that the frame is ready to be used for the kernel, If not, there is a frame the user can read and the following flags apply:}(hjD hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hhh](h)}(hCapture processh]hCapture process}(hjU hhhNhNubah}(h]h ]h"]h$]h&]uh1hhjR hhhhhMubh)}(h From include/linux/if_packet.h::h]hFrom include/linux/if_packet.h:}(hjc hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjR hhubj)}(h#define TP_STATUS_COPY (1 << 1) #define TP_STATUS_LOSING (1 << 2) #define TP_STATUS_CSUMNOTREADY (1 << 3) #define TP_STATUS_CSUM_VALID (1 << 7)h]h#define TP_STATUS_COPY (1 << 1) #define TP_STATUS_LOSING (1 << 2) #define TP_STATUS_CSUMNOTREADY (1 << 3) #define TP_STATUS_CSUM_VALID (1 << 7)}hjq sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjR hhubj )}(hhh]j)}(hhh](j)}(hhh]h}(h]h ]h"]h$]h&]colwidthKuh1jhj ubj)}(hhh]h}(h]h ]h"]h$]h&]colwidthK7uh1jhj ubj*)}(hhh](j/)}(hhh](j4)}(hhh]h)}(hTP_STATUS_COPYh]hTP_STATUS_COPY}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj ubah}(h]h ]h"]h$]h&]uh1j3hj ubj4)}(hhh](h)}(hThis flag indicates that the frame (and associated meta information) has been truncated because it's larger than tp_frame_size. This packet can be read entirely with recvfrom().h]hThis flag indicates that the frame (and associated meta information) has been truncated because it’s larger than tp_frame_size. This packet can be read entirely with recvfrom().}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj ubh)}(hpIn order to make this work it must to be enabled previously with setsockopt() and the PACKET_COPY_THRESH option.h]hpIn order to make this work it must to be enabled previously with setsockopt() and the PACKET_COPY_THRESH option.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj ubh)}(hThe number of frames that can be buffered to be read with recvfrom is limited like a normal socket. See the SO_RCVBUF option in the socket (7) man page.h]hThe number of frames that can be buffered to be read with recvfrom is limited like a normal socket. See the SO_RCVBUF option in the socket (7) man page.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj ubeh}(h]h ]h"]h$]h&]uh1j3hj ubeh}(h]h ]h"]h$]h&]uh1j.hj ubj/)}(hhh](j4)}(hhh]h)}(hTP_STATUS_LOSINGh]hTP_STATUS_LOSING}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj ubah}(h]h ]h"]h$]h&]uh1j3hj ubj4)}(hhh]h)}(h}indicates there were packet drops from last time statistics where checked with getsockopt() and the PACKET_STATISTICS option.h]h}indicates there were packet drops from last time statistics where checked with getsockopt() and the PACKET_STATISTICS option.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj ubah}(h]h ]h"]h$]h&]uh1j3hj ubeh}(h]h ]h"]h$]h&]uh1j.hj ubj/)}(hhh](j4)}(hhh]h)}(hTP_STATUS_CSUMNOTREADYh]hTP_STATUS_CSUMNOTREADY}(hj, hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj) ubah}(h]h ]h"]h$]h&]uh1j3hj& ubj4)}(hhh]h)}(hcurrently it's used for outgoing IP packets which its checksum will be done in hardware. So while reading the packet we should not try to check the checksum.h]hcurrently it’s used for outgoing IP packets which its checksum will be done in hardware. So while reading the packet we should not try to check the checksum.}(hjC hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj@ ubah}(h]h ]h"]h$]h&]uh1j3hj& ubeh}(h]h ]h"]h$]h&]uh1j.hj ubj/)}(hhh](j4)}(hhh]h)}(hTP_STATUS_CSUM_VALIDh]hTP_STATUS_CSUM_VALID}(hjc hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj` ubah}(h]h ]h"]h$]h&]uh1j3hj] ubj4)}(hhh]h)}(hThis flag indicates that at least the transport header checksum of the packet has been already validated on the kernel side. If the flag is not set then we are free to check the checksum by ourselves provided that TP_STATUS_CSUMNOTREADY is also not set.h]hThis flag indicates that at least the transport header checksum of the packet has been already validated on the kernel side. If the flag is not set then we are free to check the checksum by ourselves provided that TP_STATUS_CSUMNOTREADY is also not set.}(hjz hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjw ubah}(h]h ]h"]h$]h&]uh1j3hj] ubeh}(h]h ]h"]h$]h&]uh1j.hj ubeh}(h]h ]h"]h$]h&]uh1j)hj ubeh}(h]h ]h"]h$]h&]colsKuh1jhj ubah}(h]h ]h"]h$]h&]uh1j hjR hhhhhNubh)}(h6for convenience there are also the following defines::h]h5for convenience there are also the following defines:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjR hhubj)}(hC#define TP_STATUS_KERNEL 0 #define TP_STATUS_USER 1h]hC#define TP_STATUS_KERNEL 0 #define TP_STATUS_USER 1}hj sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjR hhubh)}(hX;The kernel initializes all frames to TP_STATUS_KERNEL, when the kernel receives a packet it puts in the buffer and updates the status with at least the TP_STATUS_USER flag. Then the user can read the packet, once the packet is read the user must zero the status field, so the kernel can use again that frame buffer.h]hX;The kernel initializes all frames to TP_STATUS_KERNEL, when the kernel receives a packet it puts in the buffer and updates the status with at least the TP_STATUS_USER flag. Then the user can read the packet, once the packet is read the user must zero the status field, so the kernel can use again that frame buffer.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjR hhubh)}(hdThe user can use poll (any other variant should apply too) to check if new packets are in the ring::h]hcThe user can use poll (any other variant should apply too) to check if new packets are in the ring:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjR hhubj)}(hstruct pollfd pfd; pfd.fd = fd; pfd.revents = 0; pfd.events = POLLIN|POLLRDNORM|POLLERR; if (status == TP_STATUS_KERNEL) retval = poll(&pfd, 1, timeout);h]hstruct pollfd pfd; pfd.fd = fd; pfd.revents = 0; pfd.events = POLLIN|POLLRDNORM|POLLERR; if (status == TP_STATUS_KERNEL) retval = poll(&pfd, 1, timeout);}hj sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjR hhubh)}(h^It doesn't incur in a race condition to first check the status value and then poll for frames.h]h`It doesn’t incur in a race condition to first check the status value and then poll for frames.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjR hhubeh}(h]capture-processah ]h"]capture processah$]h&]uh1hhjhhhhhMubh)}(hhh](h)}(hTransmission processh]hTransmission process}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhj hhhhhMubh)}(h.Those defines are also used for transmission::h]h-Those defines are also used for transmission:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj hhubj)}(hX#define TP_STATUS_AVAILABLE 0 // Frame is available #define TP_STATUS_SEND_REQUEST 1 // Frame will be sent on next send() #define TP_STATUS_SENDING 2 // Frame is currently in transmission #define TP_STATUS_WRONG_FORMAT 4 // Frame format is not correcth]hX#define TP_STATUS_AVAILABLE 0 // Frame is available #define TP_STATUS_SEND_REQUEST 1 // Frame will be sent on next send() #define TP_STATUS_SENDING 2 // Frame is currently in transmission #define TP_STATUS_WRONG_FORMAT 4 // Frame format is not correct}hj" sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhj hhubh)}(hXFirst, the kernel initializes all frames to TP_STATUS_AVAILABLE. To send a packet, the user fills a data buffer of an available frame, sets tp_len to current data buffer size and sets its status field to TP_STATUS_SEND_REQUEST. This can be done on multiple frames. Once the user is ready to transmit, it calls send(). Then all buffers with status equal to TP_STATUS_SEND_REQUEST are forwarded to the network device. The kernel updates each status of sent frames with TP_STATUS_SENDING until the end of transfer.h]hXFirst, the kernel initializes all frames to TP_STATUS_AVAILABLE. To send a packet, the user fills a data buffer of an available frame, sets tp_len to current data buffer size and sets its status field to TP_STATUS_SEND_REQUEST. This can be done on multiple frames. Once the user is ready to transmit, it calls send(). Then all buffers with status equal to TP_STATUS_SEND_REQUEST are forwarded to the network device. The kernel updates each status of sent frames with TP_STATUS_SENDING until the end of transfer.}(hj0 hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj hhubh)}(hJAt the end of each transfer, buffer status returns to TP_STATUS_AVAILABLE.h]hJAt the end of each transfer, buffer status returns to TP_STATUS_AVAILABLE.}(hj> hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM hj hhubj)}(hpheader->tp_len = in_i_size; header->tp_status = TP_STATUS_SEND_REQUEST; retval = send(this->socket, NULL, 0, 0);h]hpheader->tp_len = in_i_size; header->tp_status = TP_STATUS_SEND_REQUEST; retval = send(this->socket, NULL, 0, 0);}hjL sbah}(h]h ]h"]h$]h&]hhuh1jhhhM hj hhubh)}(h?The user can also use poll() to check if a buffer is available:h]h?The user can also use poll() to check if a buffer is available:}(hjZ hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj hhubh)}(h(status == TP_STATUS_SENDING)h]h(status == TP_STATUS_SENDING)}(hjh hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj hhubj)}(hgstruct pollfd pfd; pfd.fd = fd; pfd.revents = 0; pfd.events = POLLOUT; retval = poll(&pfd, 1, timeout);h]hgstruct pollfd pfd; pfd.fd = fd; pfd.revents = 0; pfd.events = POLLOUT; retval = poll(&pfd, 1, timeout);}hjv sbah}(h]h ]h"]h$]h&]hhuh1jhhhMhj hhubeh}(h]transmission-processah ]h"]transmission processah$]h&]uh1hhjhhhhhMubeh}(h]+mapping-and-use-of-the-circular-buffer-ringah ]h"]-mapping and use of the circular buffer (ring)ah$]h&]uh1hhjhhhhhMubeh}(h]"packet-mmap-buffer-size-calculatorah ]h"]"packet_mmap buffer size calculatorah$]h&]uh1hhhhhhhhM?ubh)}(hhh](h)}(h9What TPACKET versions are available and when to use them?h]h9What TPACKET versions are available and when to use them?}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhj hhhhhMubj)}(hint val = tpacket_version; setsockopt(fd, SOL_PACKET, PACKET_VERSION, &val, sizeof(val)); getsockopt(fd, SOL_PACKET, PACKET_VERSION, &val, sizeof(val));h]hint val = tpacket_version; setsockopt(fd, SOL_PACKET, PACKET_VERSION, &val, sizeof(val)); getsockopt(fd, SOL_PACKET, PACKET_VERSION, &val, sizeof(val));}hj sbah}(h]h ]h"]h$]h&]hhuh1jhhhM"hj hhubh)}(hLwhere 'tpacket_version' can be TPACKET_V1 (default), TPACKET_V2, TPACKET_V3.h]hPwhere ‘tpacket_version’ can be TPACKET_V1 (default), TPACKET_V2, TPACKET_V3.}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM&hj hhubj`)}(hhh](je)}(h_TPACKET_V1: - Default if not otherwise specified by setsockopt(2) - RX_RING, TX_RING available h](jk)}(h TPACKET_V1:h]h TPACKET_V1:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1jjhhhM*hj ubj{)}(hhh]j)}(hhh](h)}(h3Default if not otherwise specified by setsockopt(2)h]h)}(hj h]h3Default if not otherwise specified by setsockopt(2)}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM)hj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hRX_RING, TX_RING available h]h)}(hRX_RING, TX_RING availableh]hRX_RING, TX_RING available}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM*hj ubah}(h]h ]h"]h$]h&]uh1hhj ubeh}(h]h ]h"]h$]h&]jjuh1jhhhM)hj ubah}(h]h ]h"]h$]h&]uh1jzhj ubeh}(h]h ]h"]h$]h&]uh1jdhhhM*hj ubje)}(hX'TPACKET_V1 --> TPACKET_V2: - Made 64 bit clean due to unsigned long usage in TPACKET_V1 structures, thus this also works on 64 bit kernel with 32 bit userspace and the like - Timestamp resolution in nanoseconds instead of microseconds - RX_RING, TX_RING available - VLAN metadata information available for packets (TP_STATUS_VLAN_VALID, TP_STATUS_VLAN_TPID_VALID), in the tpacket2_hdr structure: - TP_STATUS_VLAN_VALID bit being set into the tp_status field indicates that the tp_vlan_tci field has valid VLAN TCI value - TP_STATUS_VLAN_TPID_VALID bit being set into the tp_status field indicates that the tp_vlan_tpid field has valid VLAN TPID value - How to switch to TPACKET_V2: 1. Replace struct tpacket_hdr by struct tpacket2_hdr 2. Query header len and save 3. Set protocol version to 2, set up ring as usual 4. For getting the sockaddr_ll, use ``(void *)hdr + TPACKET_ALIGN(hdrlen)`` instead of ``(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))`` h](jk)}(hTPACKET_V1 --> TPACKET_V2:h]hTPACKET_V1 --> TPACKET_V2:}(hj) hhhNhNubah}(h]h ]h"]h$]h&]uh1jjhhhMBhj% ubj{)}(hhh]j)}(hhh](h)}(hMade 64 bit clean due to unsigned long usage in TPACKET_V1 structures, thus this also works on 64 bit kernel with 32 bit userspace and the likeh]h)}(hMade 64 bit clean due to unsigned long usage in TPACKET_V1 structures, thus this also works on 64 bit kernel with 32 bit userspace and the likeh]hMade 64 bit clean due to unsigned long usage in TPACKET_V1 structures, thus this also works on 64 bit kernel with 32 bit userspace and the like}(hjA hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM-hj= ubah}(h]h ]h"]h$]h&]uh1hhj: ubh)}(h;Timestamp resolution in nanoseconds instead of microsecondsh]h)}(hjW h]h;Timestamp resolution in nanoseconds instead of microseconds}(hjY hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM0hjU ubah}(h]h ]h"]h$]h&]uh1hhj: ubh)}(hRX_RING, TX_RING availableh]h)}(hjn h]hRX_RING, TX_RING available}(hjp hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM1hjl ubah}(h]h ]h"]h$]h&]uh1hhj: ubh)}(hXVLAN metadata information available for packets (TP_STATUS_VLAN_VALID, TP_STATUS_VLAN_TPID_VALID), in the tpacket2_hdr structure: - TP_STATUS_VLAN_VALID bit being set into the tp_status field indicates that the tp_vlan_tci field has valid VLAN TCI value - TP_STATUS_VLAN_TPID_VALID bit being set into the tp_status field indicates that the tp_vlan_tpid field has valid VLAN TPID value h](h)}(hVLAN metadata information available for packets (TP_STATUS_VLAN_VALID, TP_STATUS_VLAN_TPID_VALID), in the tpacket2_hdr structure:h]hVLAN metadata information available for packets (TP_STATUS_VLAN_VALID, TP_STATUS_VLAN_TPID_VALID), in the tpacket2_hdr structure:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM2hj ubj:)}(hX- TP_STATUS_VLAN_VALID bit being set into the tp_status field indicates that the tp_vlan_tci field has valid VLAN TCI value - TP_STATUS_VLAN_TPID_VALID bit being set into the tp_status field indicates that the tp_vlan_tpid field has valid VLAN TPID value h]j)}(hhh](h)}(hyTP_STATUS_VLAN_VALID bit being set into the tp_status field indicates that the tp_vlan_tci field has valid VLAN TCI valueh]h)}(hyTP_STATUS_VLAN_VALID bit being set into the tp_status field indicates that the tp_vlan_tci field has valid VLAN TCI valueh]hyTP_STATUS_VLAN_VALID bit being set into the tp_status field indicates that the tp_vlan_tci field has valid VLAN TCI value}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM6hj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hTP_STATUS_VLAN_TPID_VALID bit being set into the tp_status field indicates that the tp_vlan_tpid field has valid VLAN TPID value h]h)}(hTP_STATUS_VLAN_TPID_VALID bit being set into the tp_status field indicates that the tp_vlan_tpid field has valid VLAN TPID valueh]hTP_STATUS_VLAN_TPID_VALID bit being set into the tp_status field indicates that the tp_vlan_tpid field has valid VLAN TPID value}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM8hj ubah}(h]h ]h"]h$]h&]uh1hhj ubeh}(h]h ]h"]h$]h&]jjuh1jhhhM6hj ubah}(h]h ]h"]h$]h&]uh1j9hhhM6hj ubeh}(h]h ]h"]h$]h&]uh1hhj: ubh)}(hX`How to switch to TPACKET_V2: 1. Replace struct tpacket_hdr by struct tpacket2_hdr 2. Query header len and save 3. Set protocol version to 2, set up ring as usual 4. For getting the sockaddr_ll, use ``(void *)hdr + TPACKET_ALIGN(hdrlen)`` instead of ``(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))`` h](h)}(hHow to switch to TPACKET_V2:h]hHow to switch to TPACKET_V2:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM;hj ubj:)}(hX1. Replace struct tpacket_hdr by struct tpacket2_hdr 2. Query header len and save 3. Set protocol version to 2, set up ring as usual 4. For getting the sockaddr_ll, use ``(void *)hdr + TPACKET_ALIGN(hdrlen)`` instead of ``(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))`` h]h)}(hhh](h)}(h1Replace struct tpacket_hdr by struct tpacket2_hdrh]h)}(hj h]h1Replace struct tpacket_hdr by struct tpacket2_hdr}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM=hj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hQuery header len and saveh]h)}(hj h]hQuery header len and save}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM>hj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(h/Set protocol version to 2, set up ring as usualh]h)}(hj' h]h/Set protocol version to 2, set up ring as usual}(hj) hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM?hj% ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hFor getting the sockaddr_ll, use ``(void *)hdr + TPACKET_ALIGN(hdrlen)`` instead of ``(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))`` h]h)}(hFor getting the sockaddr_ll, use ``(void *)hdr + TPACKET_ALIGN(hdrlen)`` instead of ``(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))``h](h!For getting the sockaddr_ll, use }(hj@ hhhNhNubj)}(h'``(void *)hdr + TPACKET_ALIGN(hdrlen)``h]h#(void *)hdr + TPACKET_ALIGN(hdrlen)}(hjH hhhNhNubah}(h]h ]h"]h$]h&]uh1jhj@ ubh instead of }(hj@ hhhNhNubj)}(h;``(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))``h]h7(void *)hdr + TPACKET_ALIGN(sizeof(struct tpacket_hdr))}(hjZ hhhNhNubah}(h]h ]h"]h$]h&]uh1jhj@ ubeh}(h]h ]h"]h$]h&]uh1hhhhM@hj< ubah}(h]h ]h"]h$]h&]uh1hhj ubeh}(h]h ]h"]h$]h&]j&arabicj(hj).uh1hhj ubah}(h]h ]h"]h$]h&]uh1j9hhhM=hj ubeh}(h]h ]h"]h$]h&]uh1hhj: ubeh}(h]h ]h"]h$]h&]jjuh1jhhhM-hj7 ubah}(h]h ]h"]h$]h&]uh1jzhj% ubeh}(h]h ]h"]h$]h&]uh1jdhhhMBhj hhubje)}(hXFTPACKET_V2 --> TPACKET_V3: - Flexible buffer implementation for RX_RING: 1. Blocks can be configured with non-static frame-size 2. Read/poll is at a block-level (as opposed to packet-level) 3. Added poll timeout to avoid indefinite user-space wait on idle links 4. Added user-configurable knobs: 4.1 block::timeout 4.2 tpkt_hdr::sk_rxhash - RX Hash data available in user space - TX_RING semantics are conceptually similar to TPACKET_V2; use tpacket3_hdr instead of tpacket2_hdr, and TPACKET3_HDRLEN instead of TPACKET2_HDRLEN. In the current implementation, the tp_next_offset field in the tpacket3_hdr MUST be set to zero, indicating that the ring does not hold variable sized frames. Packets with non-zero values of tp_next_offset will be dropped. h](jk)}(hTPACKET_V2 --> TPACKET_V3:h]hTPACKET_V2 --> TPACKET_V3:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1jjhhhMUhj ubj{)}(hhh]j)}(hhh](h)}(hXtFlexible buffer implementation for RX_RING: 1. Blocks can be configured with non-static frame-size 2. Read/poll is at a block-level (as opposed to packet-level) 3. Added poll timeout to avoid indefinite user-space wait on idle links 4. Added user-configurable knobs: 4.1 block::timeout 4.2 tpkt_hdr::sk_rxhash h]j`)}(hhh]je)}(hXJFlexible buffer implementation for RX_RING: 1. Blocks can be configured with non-static frame-size 2. Read/poll is at a block-level (as opposed to packet-level) 3. Added poll timeout to avoid indefinite user-space wait on idle links 4. Added user-configurable knobs: 4.1 block::timeout 4.2 tpkt_hdr::sk_rxhash h](jk)}(h+Flexible buffer implementation for RX_RING:h]h+Flexible buffer implementation for RX_RING:}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1jjhhhMMhj ubj{)}(hhh]h)}(hhh](h)}(h3Blocks can be configured with non-static frame-sizeh]h)}(hj h]h3Blocks can be configured with non-static frame-size}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMFhj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(h:Read/poll is at a block-level (as opposed to packet-level)h]h)}(hj h]h:Read/poll is at a block-level (as opposed to packet-level)}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMGhj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hDAdded poll timeout to avoid indefinite user-space wait on idle linksh]h)}(hDAdded poll timeout to avoid indefinite user-space wait on idle linksh]hDAdded poll timeout to avoid indefinite user-space wait on idle links}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMHhj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hUAdded user-configurable knobs: 4.1 block::timeout 4.2 tpkt_hdr::sk_rxhash h](h)}(hAdded user-configurable knobs:h]hAdded user-configurable knobs:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMJhjubj:)}(h+4.1 block::timeout 4.2 tpkt_hdr::sk_rxhash h]h)}(h*4.1 block::timeout 4.2 tpkt_hdr::sk_rxhashh]h*4.1 block::timeout 4.2 tpkt_hdr::sk_rxhash}(hj-hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMLhj)ubah}(h]h ]h"]h$]h&]uh1j9hhhMLhjubeh}(h]h ]h"]h$]h&]uh1hhj ubeh}(h]h ]h"]h$]h&]j&jz j(hj)j{ uh1hhj ubah}(h]h ]h"]h$]h&]uh1jzhj ubeh}(h]h ]h"]h$]h&]uh1jdhhhMMhj ubah}(h]h ]h"]h$]h&]uh1j_hj ubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(h$RX Hash data available in user spaceh]h)}(hjgh]h$RX Hash data available in user space}(hjihhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMOhjeubah}(h]h ]h"]h$]h&]uh1hhj ubh)}(hXsTX_RING semantics are conceptually similar to TPACKET_V2; use tpacket3_hdr instead of tpacket2_hdr, and TPACKET3_HDRLEN instead of TPACKET2_HDRLEN. In the current implementation, the tp_next_offset field in the tpacket3_hdr MUST be set to zero, indicating that the ring does not hold variable sized frames. Packets with non-zero values of tp_next_offset will be dropped. h]h)}(hXrTX_RING semantics are conceptually similar to TPACKET_V2; use tpacket3_hdr instead of tpacket2_hdr, and TPACKET3_HDRLEN instead of TPACKET2_HDRLEN. In the current implementation, the tp_next_offset field in the tpacket3_hdr MUST be set to zero, indicating that the ring does not hold variable sized frames. Packets with non-zero values of tp_next_offset will be dropped.h]hXrTX_RING semantics are conceptually similar to TPACKET_V2; use tpacket3_hdr instead of tpacket2_hdr, and TPACKET3_HDRLEN instead of TPACKET2_HDRLEN. In the current implementation, the tp_next_offset field in the tpacket3_hdr MUST be set to zero, indicating that the ring does not hold variable sized frames. Packets with non-zero values of tp_next_offset will be dropped.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMPhj|ubah}(h]h ]h"]h$]h&]uh1hhj ubeh}(h]h ]h"]h$]h&]jjuh1jhhhMEhj ubah}(h]h ]h"]h$]h&]uh1jzhj ubeh}(h]h ]h"]h$]h&]uh1jdhhhMUhj hhubeh}(h]h ]h"]h$]h&]uh1j_hj hhhNhNubeh}(h]8what-tpacket-versions-are-available-and-when-to-use-themah ]h"]9what tpacket versions are available and when to use them?ah$]h&]uh1hhhhhhhhMubh)}(hhh](h)}(hAF_PACKET fanout modeh]hAF_PACKET fanout mode}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMXubh)}(hIn the AF_PACKET fanout mode, packet reception can be load balanced among processes. This also works in combination with mmap(2) on packet sockets.h]hIn the AF_PACKET fanout mode, packet reception can be load balanced among processes. This also works in combination with mmap(2) on packet sockets.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMZhjhhubh)}(h*Currently implemented fanout policies are:h]h*Currently implemented fanout policies are:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM]hjhhubj:)}(hX|- PACKET_FANOUT_HASH: schedule to socket by skb's packet hash - PACKET_FANOUT_LB: schedule to socket by round-robin - PACKET_FANOUT_CPU: schedule to socket by CPU packet arrives on - PACKET_FANOUT_RND: schedule to socket by random selection - PACKET_FANOUT_ROLLOVER: if one socket is full, rollover to another - PACKET_FANOUT_QM: schedule to socket by skbs recorded queue_mapping h]j)}(hhh](h)}(h;PACKET_FANOUT_HASH: schedule to socket by skb's packet hashh]h)}(hjh]h=PACKET_FANOUT_HASH: schedule to socket by skb’s packet hash}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM_hjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h3PACKET_FANOUT_LB: schedule to socket by round-robinh]h)}(hjh]h3PACKET_FANOUT_LB: schedule to socket by round-robin}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM`hjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h>PACKET_FANOUT_CPU: schedule to socket by CPU packet arrives onh]h)}(hjh]h>PACKET_FANOUT_CPU: schedule to socket by CPU packet arrives on}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMahjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h9PACKET_FANOUT_RND: schedule to socket by random selectionh]h)}(hj/h]h9PACKET_FANOUT_RND: schedule to socket by random selection}(hj1hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMbhj-ubah}(h]h ]h"]h$]h&]uh1hhjubh)}(hBPACKET_FANOUT_ROLLOVER: if one socket is full, rollover to anotherh]h)}(hjFh]hBPACKET_FANOUT_ROLLOVER: if one socket is full, rollover to another}(hjHhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMchjDubah}(h]h ]h"]h$]h&]uh1hhjubh)}(hDPACKET_FANOUT_QM: schedule to socket by skbs recorded queue_mapping h]h)}(hCPACKET_FANOUT_QM: schedule to socket by skbs recorded queue_mappingh]hCPACKET_FANOUT_QM: schedule to socket by skbs recorded queue_mapping}(hj_hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMdhj[ubah}(h]h ]h"]h$]h&]uh1hhjubeh}(h]h ]h"]h$]h&]jjuh1jhhhM_hjubah}(h]h ]h"]h$]h&]uh1j9hhhM_hjhhubh)}(hfMinimal example code by David S. Miller (try things like "./test eth0 hash", "./test eth0 lb", etc.)::h]hmMinimal example code by David S. Miller (try things like “./test eth0 hash”, “./test eth0 lb”, etc.):}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMfhjhhubj)}(hX #include #include #include #include #include #include #include #include #include #include #include #include static const char *device_name; static int fanout_type; static int fanout_id; #ifndef PACKET_FANOUT # define PACKET_FANOUT 18 # define PACKET_FANOUT_HASH 0 # define PACKET_FANOUT_LB 1 #endif static int setup_socket(void) { int err, fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_IP)); struct sockaddr_ll ll; struct ifreq ifr; int fanout_arg; if (fd < 0) { perror("socket"); return EXIT_FAILURE; } memset(&ifr, 0, sizeof(ifr)); strcpy(ifr.ifr_name, device_name); err = ioctl(fd, SIOCGIFINDEX, &ifr); if (err < 0) { perror("SIOCGIFINDEX"); return EXIT_FAILURE; } memset(&ll, 0, sizeof(ll)); ll.sll_family = AF_PACKET; ll.sll_ifindex = ifr.ifr_ifindex; err = bind(fd, (struct sockaddr *) &ll, sizeof(ll)); if (err < 0) { perror("bind"); return EXIT_FAILURE; } fanout_arg = (fanout_id | (fanout_type << 16)); err = setsockopt(fd, SOL_PACKET, PACKET_FANOUT, &fanout_arg, sizeof(fanout_arg)); if (err) { perror("setsockopt"); return EXIT_FAILURE; } return fd; } static void fanout_thread(void) { int fd = setup_socket(); int limit = 10000; if (fd < 0) exit(fd); while (limit-- > 0) { char buf[1600]; int err; err = read(fd, buf, sizeof(buf)); if (err < 0) { perror("read"); exit(EXIT_FAILURE); } if ((limit % 10) == 0) fprintf(stdout, "(%d) \n", getpid()); } fprintf(stdout, "%d: Received 10000 packets\n", getpid()); close(fd); exit(0); } int main(int argc, char **argp) { int fd, err; int i; if (argc != 3) { fprintf(stderr, "Usage: %s INTERFACE {hash|lb}\n", argp[0]); return EXIT_FAILURE; } if (!strcmp(argp[2], "hash")) fanout_type = PACKET_FANOUT_HASH; else if (!strcmp(argp[2], "lb")) fanout_type = PACKET_FANOUT_LB; else { fprintf(stderr, "Unknown fanout type [%s]\n", argp[2]); exit(EXIT_FAILURE); } device_name = argp[1]; fanout_id = getpid() & 0xffff; for (i = 0; i < 4; i++) { pid_t pid = fork(); switch (pid) { case 0: fanout_thread(); case -1: perror("fork"); exit(EXIT_FAILURE); } } for (i = 0; i < 4; i++) { int status; wait(&status); } return 0; }h]hX #include #include #include #include #include #include #include #include #include #include #include #include static const char *device_name; static int fanout_type; static int fanout_id; #ifndef PACKET_FANOUT # define PACKET_FANOUT 18 # define PACKET_FANOUT_HASH 0 # define PACKET_FANOUT_LB 1 #endif static int setup_socket(void) { int err, fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_IP)); struct sockaddr_ll ll; struct ifreq ifr; int fanout_arg; if (fd < 0) { perror("socket"); return EXIT_FAILURE; } memset(&ifr, 0, sizeof(ifr)); strcpy(ifr.ifr_name, device_name); err = ioctl(fd, SIOCGIFINDEX, &ifr); if (err < 0) { perror("SIOCGIFINDEX"); return EXIT_FAILURE; } memset(&ll, 0, sizeof(ll)); ll.sll_family = AF_PACKET; ll.sll_ifindex = ifr.ifr_ifindex; err = bind(fd, (struct sockaddr *) &ll, sizeof(ll)); if (err < 0) { perror("bind"); return EXIT_FAILURE; } fanout_arg = (fanout_id | (fanout_type << 16)); err = setsockopt(fd, SOL_PACKET, PACKET_FANOUT, &fanout_arg, sizeof(fanout_arg)); if (err) { perror("setsockopt"); return EXIT_FAILURE; } return fd; } static void fanout_thread(void) { int fd = setup_socket(); int limit = 10000; if (fd < 0) exit(fd); while (limit-- > 0) { char buf[1600]; int err; err = read(fd, buf, sizeof(buf)); if (err < 0) { perror("read"); exit(EXIT_FAILURE); } if ((limit % 10) == 0) fprintf(stdout, "(%d) \n", getpid()); } fprintf(stdout, "%d: Received 10000 packets\n", getpid()); close(fd); exit(0); } int main(int argc, char **argp) { int fd, err; int i; if (argc != 3) { fprintf(stderr, "Usage: %s INTERFACE {hash|lb}\n", argp[0]); return EXIT_FAILURE; } if (!strcmp(argp[2], "hash")) fanout_type = PACKET_FANOUT_HASH; else if (!strcmp(argp[2], "lb")) fanout_type = PACKET_FANOUT_LB; else { fprintf(stderr, "Unknown fanout type [%s]\n", argp[2]); exit(EXIT_FAILURE); } device_name = argp[1]; fanout_id = getpid() & 0xffff; for (i = 0; i < 4; i++) { pid_t pid = fork(); switch (pid) { case 0: fanout_thread(); case -1: perror("fork"); exit(EXIT_FAILURE); } } for (i = 0; i < 4; i++) { int status; wait(&status); } return 0; }}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhMihjhhubeh}(h]af-packet-fanout-modeah ]h"]af_packet fanout modeah$]h&]uh1hhhhhhhhMXubh)}(hhh](h)}(hAF_PACKET TPACKET_V3 exampleh]hAF_PACKET TPACKET_V3 example}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMubh)}(hAF_PACKET's TPACKET_V3 ring buffer can be configured to use non-static frame sizes by doing its own memory management. It is based on blocks where polling works on a per block basis instead of per ring as in TPACKET_V2 and predecessor.h]hAF_PACKET’s TPACKET_V3 ring buffer can be configured to use non-static frame sizes by doing its own memory management. It is based on blocks where polling works on a per block basis instead of per ring as in TPACKET_V2 and predecessor.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(h9It is said that TPACKET_V3 brings the following benefits:h]h9It is said that TPACKET_V3 brings the following benefits:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj:)}(h* ~15% - 20% reduction in CPU-usage * ~20% increase in packet capture rate * ~2x increase in packet density * Port aggregation analysis * Non static frame size to capture entire packet payload h]j)}(hhh](h)}(h!~15% - 20% reduction in CPU-usageh]h)}(hjh]h!~15% - 20% reduction in CPU-usage}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h$~20% increase in packet capture rateh]h)}(hjh]h$~20% increase in packet capture rate}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h~2x increase in packet densityh]h)}(hjh]h~2x increase in packet density}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(hPort aggregation analysish]h)}(hjh]hPort aggregation analysis}(hj hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjubah}(h]h ]h"]h$]h&]uh1hhjubh)}(h7Non static frame size to capture entire packet payload h]h)}(h6Non static frame size to capture entire packet payloadh]h6Non static frame size to capture entire packet payload}(hj7hhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhj3ubah}(h]h ]h"]h$]h&]uh1hhjubeh}(h]h ]h"]h$]h&]j*uh1jhhhMhjubah}(h]h ]h"]h$]h&]uh1j9hhhMhjhhubh)}(hASo it seems to be a good candidate to be used with packet fanout.h]hASo it seems to be a good candidate to be used with packet fanout.}(hjXhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hMinimal example code by Daniel Borkmann based on Chetan Loke's lolpcap (compile it with gcc -Wall -O2 blob.c, and try things like "./a.out eth0", etc.)::h]hMinimal example code by Daniel Borkmann based on Chetan Loke’s lolpcap (compile it with gcc -Wall -O2 blob.c, and try things like “./a.out eth0”, etc.):}(hjfhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj)}(hX/* Written from scratch, but kernel-to-user space API usage * dissected from lolpcap: * Copyright 2011, Chetan Loke * License: GPL, version 2.0 */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifndef likely # define likely(x) __builtin_expect(!!(x), 1) #endif #ifndef unlikely # define unlikely(x) __builtin_expect(!!(x), 0) #endif struct block_desc { uint32_t version; uint32_t offset_to_priv; struct tpacket_hdr_v1 h1; }; struct ring { struct iovec *rd; uint8_t *map; struct tpacket_req3 req; }; static unsigned long packets_total = 0, bytes_total = 0; static sig_atomic_t sigint = 0; static void sighandler(int num) { sigint = 1; } static int setup_socket(struct ring *ring, char *netdev) { int err, i, fd, v = TPACKET_V3; struct sockaddr_ll ll; unsigned int blocksiz = 1 << 22, framesiz = 1 << 11; unsigned int blocknum = 64; fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_ALL)); if (fd < 0) { perror("socket"); exit(1); } err = setsockopt(fd, SOL_PACKET, PACKET_VERSION, &v, sizeof(v)); if (err < 0) { perror("setsockopt"); exit(1); } memset(&ring->req, 0, sizeof(ring->req)); ring->req.tp_block_size = blocksiz; ring->req.tp_frame_size = framesiz; ring->req.tp_block_nr = blocknum; ring->req.tp_frame_nr = (blocksiz * blocknum) / framesiz; ring->req.tp_retire_blk_tov = 60; ring->req.tp_feature_req_word = TP_FT_REQ_FILL_RXHASH; err = setsockopt(fd, SOL_PACKET, PACKET_RX_RING, &ring->req, sizeof(ring->req)); if (err < 0) { perror("setsockopt"); exit(1); } ring->map = mmap(NULL, ring->req.tp_block_size * ring->req.tp_block_nr, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_LOCKED, fd, 0); if (ring->map == MAP_FAILED) { perror("mmap"); exit(1); } ring->rd = malloc(ring->req.tp_block_nr * sizeof(*ring->rd)); assert(ring->rd); for (i = 0; i < ring->req.tp_block_nr; ++i) { ring->rd[i].iov_base = ring->map + (i * ring->req.tp_block_size); ring->rd[i].iov_len = ring->req.tp_block_size; } memset(&ll, 0, sizeof(ll)); ll.sll_family = PF_PACKET; ll.sll_protocol = htons(ETH_P_ALL); ll.sll_ifindex = if_nametoindex(netdev); ll.sll_hatype = 0; ll.sll_pkttype = 0; ll.sll_halen = 0; err = bind(fd, (struct sockaddr *) &ll, sizeof(ll)); if (err < 0) { perror("bind"); exit(1); } return fd; } static void display(struct tpacket3_hdr *ppd) { struct ethhdr *eth = (struct ethhdr *) ((uint8_t *) ppd + ppd->tp_mac); struct iphdr *ip = (struct iphdr *) ((uint8_t *) eth + ETH_HLEN); if (eth->h_proto == htons(ETH_P_IP)) { struct sockaddr_in ss, sd; char sbuff[NI_MAXHOST], dbuff[NI_MAXHOST]; memset(&ss, 0, sizeof(ss)); ss.sin_family = PF_INET; ss.sin_addr.s_addr = ip->saddr; getnameinfo((struct sockaddr *) &ss, sizeof(ss), sbuff, sizeof(sbuff), NULL, 0, NI_NUMERICHOST); memset(&sd, 0, sizeof(sd)); sd.sin_family = PF_INET; sd.sin_addr.s_addr = ip->daddr; getnameinfo((struct sockaddr *) &sd, sizeof(sd), dbuff, sizeof(dbuff), NULL, 0, NI_NUMERICHOST); printf("%s -> %s, ", sbuff, dbuff); } printf("rxhash: 0x%x\n", ppd->hv1.tp_rxhash); } static void walk_block(struct block_desc *pbd, const int block_num) { int num_pkts = pbd->h1.num_pkts, i; unsigned long bytes = 0; struct tpacket3_hdr *ppd; ppd = (struct tpacket3_hdr *) ((uint8_t *) pbd + pbd->h1.offset_to_first_pkt); for (i = 0; i < num_pkts; ++i) { bytes += ppd->tp_snaplen; display(ppd); ppd = (struct tpacket3_hdr *) ((uint8_t *) ppd + ppd->tp_next_offset); } packets_total += num_pkts; bytes_total += bytes; } static void flush_block(struct block_desc *pbd) { pbd->h1.block_status = TP_STATUS_KERNEL; } static void teardown_socket(struct ring *ring, int fd) { munmap(ring->map, ring->req.tp_block_size * ring->req.tp_block_nr); free(ring->rd); close(fd); } int main(int argc, char **argp) { int fd, err; socklen_t len; struct ring ring; struct pollfd pfd; unsigned int block_num = 0, blocks = 64; struct block_desc *pbd; struct tpacket_stats_v3 stats; if (argc != 2) { fprintf(stderr, "Usage: %s INTERFACE\n", argp[0]); return EXIT_FAILURE; } signal(SIGINT, sighandler); memset(&ring, 0, sizeof(ring)); fd = setup_socket(&ring, argp[argc - 1]); assert(fd > 0); memset(&pfd, 0, sizeof(pfd)); pfd.fd = fd; pfd.events = POLLIN | POLLERR; pfd.revents = 0; while (likely(!sigint)) { pbd = (struct block_desc *) ring.rd[block_num].iov_base; if ((pbd->h1.block_status & TP_STATUS_USER) == 0) { poll(&pfd, 1, -1); continue; } walk_block(pbd, block_num); flush_block(pbd); block_num = (block_num + 1) % blocks; } len = sizeof(stats); err = getsockopt(fd, SOL_PACKET, PACKET_STATISTICS, &stats, &len); if (err < 0) { perror("getsockopt"); exit(1); } fflush(stdout); printf("\nReceived %u packets, %lu bytes, %u dropped, freeze_q_cnt: %u\n", stats.tp_packets, bytes_total, stats.tp_drops, stats.tp_freeze_q_cnt); teardown_socket(&ring, fd); return 0; }h]hX/* Written from scratch, but kernel-to-user space API usage * dissected from lolpcap: * Copyright 2011, Chetan Loke * License: GPL, version 2.0 */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifndef likely # define likely(x) __builtin_expect(!!(x), 1) #endif #ifndef unlikely # define unlikely(x) __builtin_expect(!!(x), 0) #endif struct block_desc { uint32_t version; uint32_t offset_to_priv; struct tpacket_hdr_v1 h1; }; struct ring { struct iovec *rd; uint8_t *map; struct tpacket_req3 req; }; static unsigned long packets_total = 0, bytes_total = 0; static sig_atomic_t sigint = 0; static void sighandler(int num) { sigint = 1; } static int setup_socket(struct ring *ring, char *netdev) { int err, i, fd, v = TPACKET_V3; struct sockaddr_ll ll; unsigned int blocksiz = 1 << 22, framesiz = 1 << 11; unsigned int blocknum = 64; fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_ALL)); if (fd < 0) { perror("socket"); exit(1); } err = setsockopt(fd, SOL_PACKET, PACKET_VERSION, &v, sizeof(v)); if (err < 0) { perror("setsockopt"); exit(1); } memset(&ring->req, 0, sizeof(ring->req)); ring->req.tp_block_size = blocksiz; ring->req.tp_frame_size = framesiz; ring->req.tp_block_nr = blocknum; ring->req.tp_frame_nr = (blocksiz * blocknum) / framesiz; ring->req.tp_retire_blk_tov = 60; ring->req.tp_feature_req_word = TP_FT_REQ_FILL_RXHASH; err = setsockopt(fd, SOL_PACKET, PACKET_RX_RING, &ring->req, sizeof(ring->req)); if (err < 0) { perror("setsockopt"); exit(1); } ring->map = mmap(NULL, ring->req.tp_block_size * ring->req.tp_block_nr, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_LOCKED, fd, 0); if (ring->map == MAP_FAILED) { perror("mmap"); exit(1); } ring->rd = malloc(ring->req.tp_block_nr * sizeof(*ring->rd)); assert(ring->rd); for (i = 0; i < ring->req.tp_block_nr; ++i) { ring->rd[i].iov_base = ring->map + (i * ring->req.tp_block_size); ring->rd[i].iov_len = ring->req.tp_block_size; } memset(&ll, 0, sizeof(ll)); ll.sll_family = PF_PACKET; ll.sll_protocol = htons(ETH_P_ALL); ll.sll_ifindex = if_nametoindex(netdev); ll.sll_hatype = 0; ll.sll_pkttype = 0; ll.sll_halen = 0; err = bind(fd, (struct sockaddr *) &ll, sizeof(ll)); if (err < 0) { perror("bind"); exit(1); } return fd; } static void display(struct tpacket3_hdr *ppd) { struct ethhdr *eth = (struct ethhdr *) ((uint8_t *) ppd + ppd->tp_mac); struct iphdr *ip = (struct iphdr *) ((uint8_t *) eth + ETH_HLEN); if (eth->h_proto == htons(ETH_P_IP)) { struct sockaddr_in ss, sd; char sbuff[NI_MAXHOST], dbuff[NI_MAXHOST]; memset(&ss, 0, sizeof(ss)); ss.sin_family = PF_INET; ss.sin_addr.s_addr = ip->saddr; getnameinfo((struct sockaddr *) &ss, sizeof(ss), sbuff, sizeof(sbuff), NULL, 0, NI_NUMERICHOST); memset(&sd, 0, sizeof(sd)); sd.sin_family = PF_INET; sd.sin_addr.s_addr = ip->daddr; getnameinfo((struct sockaddr *) &sd, sizeof(sd), dbuff, sizeof(dbuff), NULL, 0, NI_NUMERICHOST); printf("%s -> %s, ", sbuff, dbuff); } printf("rxhash: 0x%x\n", ppd->hv1.tp_rxhash); } static void walk_block(struct block_desc *pbd, const int block_num) { int num_pkts = pbd->h1.num_pkts, i; unsigned long bytes = 0; struct tpacket3_hdr *ppd; ppd = (struct tpacket3_hdr *) ((uint8_t *) pbd + pbd->h1.offset_to_first_pkt); for (i = 0; i < num_pkts; ++i) { bytes += ppd->tp_snaplen; display(ppd); ppd = (struct tpacket3_hdr *) ((uint8_t *) ppd + ppd->tp_next_offset); } packets_total += num_pkts; bytes_total += bytes; } static void flush_block(struct block_desc *pbd) { pbd->h1.block_status = TP_STATUS_KERNEL; } static void teardown_socket(struct ring *ring, int fd) { munmap(ring->map, ring->req.tp_block_size * ring->req.tp_block_nr); free(ring->rd); close(fd); } int main(int argc, char **argp) { int fd, err; socklen_t len; struct ring ring; struct pollfd pfd; unsigned int block_num = 0, blocks = 64; struct block_desc *pbd; struct tpacket_stats_v3 stats; if (argc != 2) { fprintf(stderr, "Usage: %s INTERFACE\n", argp[0]); return EXIT_FAILURE; } signal(SIGINT, sighandler); memset(&ring, 0, sizeof(ring)); fd = setup_socket(&ring, argp[argc - 1]); assert(fd > 0); memset(&pfd, 0, sizeof(pfd)); pfd.fd = fd; pfd.events = POLLIN | POLLERR; pfd.revents = 0; while (likely(!sigint)) { pbd = (struct block_desc *) ring.rd[block_num].iov_base; if ((pbd->h1.block_status & TP_STATUS_USER) == 0) { poll(&pfd, 1, -1); continue; } walk_block(pbd, block_num); flush_block(pbd); block_num = (block_num + 1) % blocks; } len = sizeof(stats); err = getsockopt(fd, SOL_PACKET, PACKET_STATISTICS, &stats, &len); if (err < 0) { perror("getsockopt"); exit(1); } fflush(stdout); printf("\nReceived %u packets, %lu bytes, %u dropped, freeze_q_cnt: %u\n", stats.tp_packets, bytes_total, stats.tp_drops, stats.tp_freeze_q_cnt); teardown_socket(&ring, fd); return 0; }}hjtsbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjhhubeh}(h]af-packet-tpacket-v3-exampleah ]h"]af_packet tpacket_v3 exampleah$]h&]uh1hhhhhhhhMubh)}(hhh](h)}(hPACKET_QDISC_BYPASSh]hPACKET_QDISC_BYPASS}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMubh)}(hIf there is a requirement to load the network with many packets in a similar fashion as pktgen does, you might set the following option after socket creation::h]hIf there is a requirement to load the network with many packets in a similar fashion as pktgen does, you might set the following option after socket creation:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj)}(hPint one = 1; setsockopt(fd, SOL_PACKET, PACKET_QDISC_BYPASS, &one, sizeof(one));h]hPint one = 1; setsockopt(fd, SOL_PACKET, PACKET_QDISC_BYPASS, &one, sizeof(one));}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjhhubh)}(hXThis has the side-effect, that packets sent through PF_PACKET will bypass the kernel's qdisc layer and are forcedly pushed to the driver directly. Meaning, packet are not buffered, tc disciplines are ignored, increased loss can occur and such packets are also not visible to other PF_PACKET sockets anymore. So, you have been warned; generally, this can be useful for stress testing various components of a system.h]hXThis has the side-effect, that packets sent through PF_PACKET will bypass the kernel’s qdisc layer and are forcedly pushed to the driver directly. Meaning, packet are not buffered, tc disciplines are ignored, increased loss can occur and such packets are also not visible to other PF_PACKET sockets anymore. So, you have been warned; generally, this can be useful for stress testing various components of a system.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hdOn default, PACKET_QDISC_BYPASS is disabled and needs to be explicitly enabled on PF_PACKET sockets.h]hdOn default, PACKET_QDISC_BYPASS is disabled and needs to be explicitly enabled on PF_PACKET sockets.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubeh}(h]packet-qdisc-bypassah ]h"]packet_qdisc_bypassah$]h&]uh1hhhhhhhhMubh)}(hhh](h)}(hPACKET_TIMESTAMPh]hPACKET_TIMESTAMP}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhMubh)}(hXThe PACKET_TIMESTAMP setting determines the source of the timestamp in the packet meta information for mmap(2)ed RX_RING and TX_RINGs. If your NIC is capable of timestamping packets in hardware, you can request those hardware timestamps to be used. Note: you may need to enable the generation of hardware timestamps with SIOCSHWTSTAMP (see related information from Documentation/networking/timestamping.rst).h]hXThe PACKET_TIMESTAMP setting determines the source of the timestamp in the packet meta information for mmap(2)ed RX_RING and TX_RINGs. If your NIC is capable of timestamping packets in hardware, you can request those hardware timestamps to be used. Note: you may need to enable the generation of hardware timestamps with SIOCSHWTSTAMP (see related information from Documentation/networking/timestamping.rst).}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hHPACKET_TIMESTAMP accepts the same integer bit field as SO_TIMESTAMPING::h]hGPACKET_TIMESTAMP accepts the same integer bit field as SO_TIMESTAMPING:}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM hjhhubj)}(hqint req = SOF_TIMESTAMPING_RAW_HARDWARE; setsockopt(fd, SOL_PACKET, PACKET_TIMESTAMP, (void *) &req, sizeof(req))h]hqint req = SOF_TIMESTAMPING_RAW_HARDWARE; setsockopt(fd, SOL_PACKET, PACKET_TIMESTAMP, (void *) &req, sizeof(req))}hjsbah}(h]h ]h"]h$]h&]hhuh1jhhhM hjhhubh)}(hXFor the mmap(2)ed ring buffers, such timestamps are stored in the ``tpacket{,2,3}_hdr`` structure's tp_sec and ``tp_{n,u}sec`` members. To determine what kind of timestamp has been reported, the tp_status field is binary or'ed with the following possible bits ...h](hBFor the mmap(2)ed ring buffers, such timestamps are stored in the }(hjhhhNhNubj)}(h``tpacket{,2,3}_hdr``h]htpacket{,2,3}_hdr}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1jhjubh structure’s tp_sec and }(hjhhhNhNubj)}(h``tp_{n,u}sec``h]h tp_{n,u}sec}(hj0hhhNhNubah}(h]h ]h"]h$]h&]uh1jhjubh members. To determine what kind of timestamp has been reported, the tp_status field is binary or’ed with the following possible bits ...}(hjhhhNhNubeh}(h]h ]h"]h$]h&]uh1hhhhMhjhhubj)}(h/TP_STATUS_TS_RAW_HARDWARE TP_STATUS_TS_SOFTWAREh]h/TP_STATUS_TS_RAW_HARDWARE TP_STATUS_TS_SOFTWARE}hjHsbah}(h]h ]h"]h$]h&]hhuh1jhhhMhjhhubh)}(h... that are equivalent to its ``SOF_TIMESTAMPING_*`` counterparts. For the RX_RING, if neither is set (i.e. PACKET_TIMESTAMP is not set), then a software fallback was invoked *within* PF_PACKET's processing code (less precise).h](h... that are equivalent to its }(hjVhhhNhNubj)}(h``SOF_TIMESTAMPING_*``h]hSOF_TIMESTAMPING_*}(hj^hhhNhNubah}(h]h ]h"]h$]h&]uh1jhjVubh{ counterparts. For the RX_RING, if neither is set (i.e. PACKET_TIMESTAMP is not set), then a software fallback was invoked }(hjVhhhNhNubhemphasis)}(h*within*h]hwithin}(hjrhhhNhNubah}(h]h ]h"]h$]h&]uh1jphjVubh. PF_PACKET’s processing code (less precise).}(hjVhhhNhNubeh}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hX$Getting timestamps for the TX_RING works as follows: i) fill the ring frames, ii) call sendto() e.g. in blocking mode, iii) wait for status of relevant frames to be updated resp. the frame handed over to the application, iv) walk through the frames to pick up the individual hw/sw timestamps.h]hX$Getting timestamps for the TX_RING works as follows: i) fill the ring frames, ii) call sendto() e.g. in blocking mode, iii) wait for status of relevant frames to be updated resp. the frame handed over to the application, iv) walk through the frames to pick up the individual hw/sw timestamps.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhMhjhhubh)}(hXtOnly (!) if transmit timestamping is enabled, then these bits are combined with binary | with TP_STATUS_AVAILABLE, so you must check for that in your application (e.g. !(tp_status & (TP_STATUS_SEND_REQUEST | TP_STATUS_SENDING)) in a first step to see if the frame belongs to the application, and then one can extract the type of timestamp in a second step from tp_status)!h]hXtOnly (!) if transmit timestamping is enabled, then these bits are combined with binary | with TP_STATUS_AVAILABLE, so you must check for that in your application (e.g. !(tp_status & (TP_STATUS_SEND_REQUEST | TP_STATUS_SENDING)) in a first step to see if the frame belongs to the application, and then one can extract the type of timestamp in a second step from tp_status)!}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM$hjhhubh)}(hX5If you don't care about them, thus having it disabled, checking for TP_STATUS_AVAILABLE resp. TP_STATUS_WRONG_FORMAT is sufficient. If in the TX_RING part only TP_STATUS_AVAILABLE is set, then the tp_sec and tp_{n,u}sec members do not contain a valid value. For TX_RINGs, by default no timestamp is generated!h]hX7If you don’t care about them, thus having it disabled, checking for TP_STATUS_AVAILABLE resp. TP_STATUS_WRONG_FORMAT is sufficient. If in the TX_RING part only TP_STATUS_AVAILABLE is set, then the tp_sec and tp_{n,u}sec members do not contain a valid value. For TX_RINGs, by default no timestamp is generated!}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM*hjhhubh)}(hySee include/linux/net_tstamp.h and Documentation/networking/timestamping.rst for more information on hardware timestamps.h]hySee include/linux/net_tstamp.h and Documentation/networking/timestamping.rst for more information on hardware timestamps.}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM0hjhhubeh}(h]packet-timestampah ]h"]packet_timestampah$]h&]uh1hhhhhhhhMubh)}(hhh](h)}(hMiscellaneous bitsh]hMiscellaneous bits}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhM4ubj)}(hhh]h)}(hPacket sockets work well together with Linux socket filters, thus you also might want to have a look at Documentation/networking/filter.rst h]h)}(hPacket sockets work well together with Linux socket filters, thus you also might want to have a look at Documentation/networking/filter.rsth]hPacket sockets work well together with Linux socket filters, thus you also might want to have a look at Documentation/networking/filter.rst}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhhhM6hjubah}(h]h ]h"]h$]h&]uh1hhjhhhhhNubah}(h]h ]h"]h$]h&]jjuh1jhhhM6hjhhubeh}(h]miscellaneous-bitsah ]h"]miscellaneous bitsah$]h&]uh1hhhhhhhhM4ubh)}(hhh](h)}(hTHANKSh]hTHANKS}(hjhhhNhNubah}(h]h ]h"]h$]h&]uh1hhjhhhhhM:ubj:)}(h