TNETX4080 datasheet(46/64 Pages) TI1

46 / 64 page

TNETX4080

ThunderSWITCH II™ 8-PORT 10-/100-MBIT/S ETHERNET™ SWITCH

SPWS050A – AUGUST 1998 – REVISED NOVEMBER 1998

POST OFFICE BOX 655303

• DALLAS, TEXAS 75265

IEEE Std 802.3 flow control

The TNETX4080 supports collision-based flow control for ports in half-duplex mode and IEEE Std 802.3x flow

control for ports in full-duplex mode. The flow bit in the SysControl register determines the action that will be

taken when back pressure is needed, that is, when there are insufficient resources to handle an inbound packet.

The holb bit in the SysControl register determines when back pressure is needed.

D If flow = 0, packets are discarded at the ingress port when insufficient resources are available to handle

them.

D If flow = 1, ports in half-duplex mode cause collisions to avoid accepting packets, ports in full-duplex mode

whose link partners negotiated to accept pause packets will send them, otherwise, packets are dropped.

D With holb = 0, back pressure is applied to all ports when the number of buffers in the global pool is down

to the value in the FlowThreshold register (or half of this value if the packet arrives at a port in gigabit mode).

This prevents the reception of more frames at any port until the frame backlog is reduced and the number

of free buffers has risen above the threshold. When this happens, back pressure is removed from all ports

and packets can be received. The value in FlowThreshold should be set so that all ports can complete

reception of a maximum-size frame, that is, each port should have enough time to activate the flow

mechanisms without dumping a frame for which reception has started.

D If holb = 1, back pressure is applied as when holb = 0, or to an individual port when the number of buffers

held in memory for data that arrived on that port is greater than the available pool remaining. This assumes

that FlowThreshold is set small enough that this mechanism does not affect the back pressure in this mode.

An example is:

When port-A’s traffic begins to backlog in memory [no matter to what port(s) it is destined], back pressure will

be applied when the amount of data backed up is greater than the available pool (about half the buffers are

assigned to data from port A). If A’s data stays backlogged and if data arriving at port B also begins to backlog

in memory, back pressure will be applied to port B when its data amounts to one-fourth of the buffer pool, or half

of the half left after port A had back pressure applied. When port-A’s traffic begins to exit the switch, port A stays

back pressured until its data is equal to one-third of the total. As buffers become available, port B is allowed to

consume up to one-third of the buffer pool (each backlogged total is compared with the buffers available). In

this mode, only the stations that have caused their fair share of buffers to be removed from the available pool

are back pressured.

Setting holb = 1 activates circuitry that attempts to prevent a backlogged conversation stalling other port traffic

by using up all the memory buffers. Because the number of buffers charged to a particular port always is

compared with the number of buffers left, there is no threshold register for this mode.

other flow-control mechanisms

hardware flow control

If a port were in MII mode and full duplex, normally, its Mxx_COL would not be needed. Hardware flow control

has been added by preventing the start of transmission of a frame if the Mxx_COL is high.

multicast limit

Because buffer resources for multicast (or broadcast) frames are released only when the last port has

transmitted the frame, multicast packets to fast and slow ports are released at the slow-port rate. Multicast traffic

from a fast port to a slow port could cause back pressure to be applied to the source port, blocking input from

that port. This occurs even if at least one of the ports in the multicast could have kept pace with the inbound

multicast packets. The MCastLimit register can limit the number of multicast packets allowed to be pending at

any output port. When the limit is reached, that port is not added to the routing vector of the next multicast packet

for which it was eligible. Eligibility is restored when the number of backlogged packets again is below the limit.

The MCastLimit register allows the user to set the ports subject to this restriction, and set the limit in eight binary

steps from 2 to 256. The spanning-tree BPDU multicast packet is exempt from this limit.

TNETX4080 数据表(PDF) 46 Page - Texas Instruments

TNETX4080 数据表(HTML) 46 Page - Texas Instruments

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

数据表下载