| Both sides previous revision Previous revision Next revision | Previous revision |
| docs:techref:hardware:switch [2022/10/04 22:15] – [Support matrix] Updated kernels & targets borromini | docs:techref:hardware:switch [2024/08/11 20:52] (current) – [Realtek] fix link linus |
|---|
| ====== Ethernet Network Switch ====== | ====== Ethernet Network Switch ====== |
| Many devices supported by OpenWrt contain an Ethernet network switch. Most switches are //configurable// and driver options are available. Typical features include ethernet frame tagging and [[docs:guide-user:network:vlan:switch_configuration|VLAN]] support. We require two software components: | Many devices supported by OpenWrt contain an Ethernet network switch. Most switches are //configurable// and driver options are available. Typical features include ethernet frame tagging and [[docs:guide-user:network:vlan:switch_configuration|VLAN]] support. We require two software components: |
| - a kind of a driver | - A kind of a driver |
| - a utility | - A utility |
| |
| ===== DSA Distributed Network Switch ===== | ===== DSA Distributed Network Switch ===== |
| |
| ^ Target ^ Kernel ^ Notes ^ | ^ Target ^ Kernel ^ Notes ^ |
| | | bcm53xx | 5.10? | | |
| | Gemini | 5.10 | | | | Gemini | 5.10 | | |
| | IMX6 | 5.10 | | | | IMX6 | 5.10 | | |
| * [[http://realtek.info/pdf/rtl8366_8369_datasheet_1-1.pdf|RTL8366/RTL8369 datasheet]], notice that this is quite different from RTL8366RB or RTL8369S | * [[http://realtek.info/pdf/rtl8366_8369_datasheet_1-1.pdf|RTL8366/RTL8369 datasheet]], notice that this is quite different from RTL8366RB or RTL8369S |
| * [[https://web.archive.org/web/20100729202805/http://www.realtek.com.tw/products/productsView.aspx?Langid=1&PNid=18&PFid=15&Level=5&Conn=4&ProdID=197|RealTek RTL8366RB]], a 5-port Gigabit switch e.g. in the [[toh/tp-link/tl-wr1043nd|WR1043ND v1]] or [[toh/d-link/DIR-685]] | * [[https://web.archive.org/web/20100729202805/http://www.realtek.com.tw/products/productsView.aspx?Langid=1&PNid=18&PFid=15&Level=5&Conn=4&ProdID=197|RealTek RTL8366RB]], a 5-port Gigabit switch e.g. in the [[toh/tp-link/tl-wr1043nd|WR1043ND v1]] or [[toh/d-link/DIR-685]] |
| * [[https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/dsa/rtl8366rb.c|Upstream RTL8366RB DSA kernel driver]] this should become the default after incorporating kernel v4.19 or later | * [[https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/dsa/realtek/rtl8366rb.c|Upstream RTL8366RB DSA kernel driver]] this should become the default after incorporating kernel v4.19 or later |
| * see [[https://dev.openwrt.org/search?q=+8366RB&noquickjump=1&ticket=on|all tickets concerning the Realtek RTL8366RB]] | * see [[https://dev.openwrt.org/search?q=+8366RB&noquickjump=1&ticket=on|all tickets concerning the Realtek RTL8366RB]] |
| * e.g. [[https://dev.openwrt.org/ticket/7977|Ticket 7977:Enhance rtl8366rb driver]] | * e.g. [[https://dev.openwrt.org/ticket/7977|Ticket 7977:Enhance rtl8366rb driver]] |
| |
| ==== Features ==== | ==== Features ==== |
| * some Switches contain a functionality called **//"Hardware NAT"//**; support for this features has not yet been included into OpenWrt. Since the [[docs:techref:hardware:cpu]]s used in CPE Routers are relatively slow, expect 2 to 4 times performance gain if you were to use Hardware NAT, e.g. without hardware NAT some router could achieve 400Mbit/s at full CPU load, with hardware NAT it could achieve 900Mbit/s at full CPU load. OpenWrt developer is supposed to have said, that implementing HW-NAT support would be unacceptably hacky to accomplish. | * Some switches contain a functionality called **//"Hardware NAT"//**; support for this features has not yet been included into OpenWrt. Since the [[docs:techref:hardware:cpu]]s used in CPE Routers are relatively slow, expect 2 to 4 times performance gain if you were to use Hardware NAT, e.g. without hardware NAT some router could achieve 400Mbit/s at full CPU load, with hardware NAT it could achieve 900Mbit/s at full CPU load. OpenWrt developer is supposed to have said, that implementing HW-NAT support would be unacceptably hacky to accomplish. |