HEADS-UP: IFLIB implementations of sys/dev/e1000 em, lem, igb pending

[Michael Tuexen]

> On 11 Jan 2017, at 18:44, Sean Bruno <sbruno at freebsd.org> wrote:
> 
> 
> 
> On 01/11/17 10:24, Michael Tuexen wrote:
>>> On 10 Jan 2017, at 04:32, Sean Bruno <sbruno at freebsd.org> wrote:
>>> 
>>> tl;dir --> you get to keep your igbX devices(thanks jhb), no POLA
>>> violations this week.
>>> 
>>> I've updated sys/dev/e1000 at svn R311849 to match Matt Macy's work on
>>> IFLIB in the kernel.
>>> 
>>> At this point, the driver deviates from Intel's code dramatically and
>>> you now get to yell directly into the freebsd-net@ megaphone for things
>>> that I may have broken.
>> My testing shows that sending SCTP/IPv4, SCTP/IPv6, TCP/IPv6, and UDP/IPV6
>> is broken since the checksum offload doesn't work.
>> 
>> The following patch fixes the issue:
>> 
>> Index: if_em.h
>> ===================================================================
>> --- if_em.h     (revision 311887)
>> +++ if_em.h     (working copy)
>> @@ -330,7 +330,7 @@
>> #define EM_MSIX_LINK           0x01000000 /* For 82574 use */
>> #define ETH_ZLEN               60
>> #define ETH_ADDR_LEN           6
>> -#define CSUM_OFFLOAD           7       /* Offload bits in mbuf flag */
>> +#define CSUM_OFFLOAD           0xE0F   /* Offload bits in mbuf flag */
>> 
>> #define IGB_PKTTYPE_MASK       0x0000FFF0
>> #define IGB_DMCTLX_DCFLUSH_DIS 0x80000000  /* Disable DMA Coalesce Flush */
>> 
>> The magic number 7 is CSUM_IP | CSUM_IP_UDP | CSUM_IP_TCP, which means
>> that checksum offloading is working for TCP/IPv4 and UDP/IPv4.
>> It should be 0xE0F, which is
>> CSUM_IP | CSUM_IP_UDP | CSUM_IP_TCP | CSUM_IP_SCTP | CSUM_IP6_UDP | CSUM_IP6_TCP | CSUM_IP6_SCTP.
>> 
>> I have tested the above patch with an igb card which supports all these offloads.
>> 
>> Best regards
>> Michael
>>> 
>>> man page updates are coming up next.  Please let us know if this
>>> revision has made things better, worse or none-of-the above on whatever
>>> Intel Gigabit NIC you happen to have lying around.
>>> 
>>> sean
>>> 
>>> On 01/05/17 20:18, Sean Bruno wrote:
>>>> tl;dr --> igbX devices will become emX devices
>>>> 
>>>> We're about to commit an update to sys/dev/e1000 that will implement and
>>>> activate IFLIB for em(4), lem(4) & igb(4) and would appreciate all folks
>>>> who can test and poke at the drivers to do so this week.  This will have
>>>> some really great changes for performance and standardization that have
>>>> been bouncing around inside of various FreeBSD shops that have been
>>>> collaborating with Matt Macy over the last year.
>>>> 
>>>> This will implement multiple queues for certain em(4) devices that are
>>>> capable of such things and add some new sysctl's for you to poke at in
>>>> your monitoring tools.
>>>> 
>>>> Due to limitations of device registration, igbX devices will become emX
>>>> devices.  So, you'll need to make a minor update to your rc.conf and
>>>> scripts that manipulate the network devices.
>>>> 
>>>> UPDATING will be bumped to reflect these changes.
>>>> 
>>>> MFC to stable/11 will have a legacy implementation that doesn't use
>>>> IFLIB for compatibility reasons.
>>>> 
>>>> A documentation and man page update will follow in the next few days
>>>> explaining how to work with the changed driver.
>>>> 
>>>> sean
>>>> 
>>>> bcc net@ current@ re@
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>> 
>> 
>> 
> 
> 
> Bah, last one was wrong bits.  Try this one.
> 
> Index: em_txrx.c
> ===================================================================
> --- em_txrx.c	(revision 311875)
> +++ em_txrx.c	(working copy)
> @@ -304,7 +304,7 @@
> 	if (do_tso) {
> 		i = em_tso_setup(sc, pi, &txd_upper, &txd_lower);
> 		tso_desc = TRUE;
> -	} else if (csum_flags & CSUM_OFFLOAD) {
> +	} else if (csum_flags & EM_CSUM_OFFLOAD) {
> 		i = em_transmit_checksum_setup(sc, pi, &txd_upper, &txd_lower);
> 	}
> 
> Index: if_em.h
> ===================================================================
> --- if_em.h	(revision 311875)
> +++ if_em.h	(working copy)
> @@ -330,7 +330,8 @@
> #define EM_MSIX_LINK		0x01000000 /* For 82574 use */
> #define ETH_ZLEN		60
> #define ETH_ADDR_LEN		6
> -#define CSUM_OFFLOAD		7	/* Offload bits in mbuf flag */
> +#define EM_CSUM_OFFLOAD		7	/* Offload bits in mbuf flag */
> +#define IGB_CSUM_OFFLOAD	0x0E0F	/* Offload bits in mbuf flag */
> 
> #define IGB_PKTTYPE_MASK	0x0000FFF0
> #define IGB_DMCTLX_DCFLUSH_DIS	0x80000000  /* Disable DMA Coalesce Flush */
> Index: igb_txrx.c
> ===================================================================
> --- igb_txrx.c	(revision 311875)
> +++ igb_txrx.c	(working copy)
> @@ -171,7 +171,7 @@
> 	*/
>         if (pi->ipi_mflags & M_VLANTAG) {
> 		vlan_macip_lens |= (pi->ipi_vtag << E1000_ADVTXD_VLAN_SHIFT);
> -	} else if ((pi->ipi_csum_flags & CSUM_OFFLOAD) == 0) {
> +	} else if ((pi->ipi_csum_flags & IGB_CSUM_OFFLOAD) == 0) {
> 		return (0);
> 	}
Looks good to me...

Best regards
Michael
> 	
>