Date: Sat, 6 Apr 2019 03:12:24 +0800 From: Kris von Mach <mach@swishmail.com> To: freebsd-stable@freebsd.org Subject: Re: em performs worse than igb (latency wise) in 12? Message-ID: <ce9e5132-534f-1c24-7245-51cd61d660fc@swishmail.com> In-Reply-To: <7673edad-1e50-7e9b-961e-f28ab7a0f41e@ingresso.co.uk> References: <b910baa6-6428-67fa-5df4-49b777e770d1@swishmail.com> <7673edad-1e50-7e9b-961e-f28ab7a0f41e@ingresso.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
On 4/6/2019 2:56 AM, Pete French wrote: > Something odd going on there there - I am using 12-STABLE and I have=20 > igb just fine, and it attaches to the same hardware that 11 did: It does work in 12, throughput is great, just that the latency is higher = than 11. igb0: flags=3D8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1= 500 options=3De527bb<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU,VLAN_HWC= SUM,TSO4,TSO6,LRO,WOL_MAGIC,VLAN_HWFILTER,VLAN_HWTSO,RXCSUM_IPV6,TXCSUM_I= PV6> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 ether 38:ea:a7:8d:c1:6c =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 inet 208.72.56.19 netmask 0xf= ffffc00 broadcast 208.72.59.255 =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 inet6 fe80::3aea:a7ff:fe8d:c1= 6c%igb0 prefixlen 64 scopeid 0x1 =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 inet6 2602:ffb8::208:72:56:9 = prefixlen 64 =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 media: Ethernet autoselect (1= 000baseT <full-duplex>) =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 status: active =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 nd6 options=3D21<PERFORMNUD,A= UTO_LINKLOCAL> > Do you have a custom kernel, and if so did you see this note in UPDATIN= G? Yes I do, but it includes all of GENERIC which includes em drivers,=20 otherwise it wouldn't even work with the network card. my custom kernel: include GENERIC ident=C2=A0=C2=A0 CUSTOM makeoptions WITH_EXTRA_TCP_STACKS=3D1 options TCPHPTS options SC_KERNEL_CONS_ATTR=3D(FG_GREEN|BG_BLACK) options IPSTEALTH options=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 AHC_REG_PRETTY_PRINT=C2=A0 # = Print register bitfields in debug options=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 AHD_REG_PRETTY_PRINT=C2=A0 # = Print register bitfields in debug device cryptodev device aesni I did try without RACK just in case that was the culprit.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ce9e5132-534f-1c24-7245-51cd61d660fc>