Fast Ethernet Consortium MAC Test Suite version 4.3 InterOperability Lab — 121 Technology Drive, Suite 2 — Durham, NH 03824 — (603) 862-0239 July 26, 2004 Mr. Hugh Cao Actel Corporation, Building 1 2061 Stierlin Court Mountain View, CA 94043-4655 Mr. Cao: Enclosed are the results for MAC conformance testing performed on the Actel Platform 8051 Core10/100 MAC testing board with Actel ProASIC APA600 MAC, and VS31AL PHY. The tests are based on Clause 4 of IEEE Standard 802.3, 2002 edition. The tests performed are a subset of the MAC Conformance Test Suite available at the InterOperability Lab’s web site: ftp://ftp.iol.unh.edu/pub/ethernet/test_suites/mac/mac.pdf As always, we welcome any comments regarding this Test Suite. During the testing process, the following issues were uncovered: Test Result 4.6.2 – Truncated The DUT was not observed to properly implement the standard defined collision backoff Binary Exponential algorithm. BackOff test If you have any questions about the test procedures or results, please contact me via e-mail at [email protected] or by phone at (603) 862-0239. Sincerely Marc Deschenes Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board The following table contains possible results and their meanings: Result Interpretation PASS FAIL PASS with Comments Warning The DUT was observed to exhibit conformant behavior. The DUT was observed to exhibit non-compliant behavior. The DUT was observed to exhibit conformant behavior, however this behavior deviated from previous compliant results. An additional explanation of the situation is included. The DUT was observed to exhibit behavior that is not recommended. Refer to Comments From the observations, a valid pass or fail could not be determined. An additional explanation of the situation is included. Not Appl icable Not Available Not Tested Borderline Informative Ethernet Data Link Layer The DUT does not support the technology required to perform these tests. Due to testing station or time limitations, the tests could not be performed, or were performed in a limited capacity. Not tested due to time constraint of the test period. The observed values of the specified parameter are valid at one extreme, and invalid at the other extreme. Results are for informative purposes only and are not judged on a pass or fail basis. 2 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Group 1: Test # 4.1.1 Test Label Reception of Frames with FCS Errors a b Result PASS PASS Comments on Test Procedure Purpose: To verify that the device under test detects frames with FCS errors and reports a frameCheckError. a. b. The DUT shall detect and discard the frame with an invalid value in the FCS field, and, if clause 30 is implemented, log a FCS error. The reception of test frames with an invalid FCS field should not affect the reception of valid request frames preceding or following the Test Frame by a minimum InterFrameGap. Comments on Test Results a. b. The DUT was observed to properly discard a frame with an invalid CRC. The statistics of the DUT were unavailable during testing. The reception of test frames with invalid FCS fields was not observed to affect the reception of the valid request frames. Test # Test Label 4.1.2 Reception of Fragments and Runts a b c Result PASS PASS PASS Comments on Test Procedure Purpose: To verify that the device under test discards collision fragments. a. b. c. The DUT should discard all fragments and runts less than 64 bytes in length. The DUT should properly discard each of the sequences listed below: 1. 7 octets of preamble 2. 7 octets of preamble and SFD 3. 7 octets of preamble, SFD and the MAC address of the DUT 4. 7 octets of preamble, SFD, the MAC address of the DUT and an arbitrary source address The reception of fragments or runts should not interfere with the reception of valid MAC frames. Comments on Test Results a. b. All fragments and runts were discarded by the DUT. The statistics of the DUT were unavailable during testing. The DUT was observed to behave in the following ways when receiving each of the different test fragments: 1. The DUT discarded the fragment consisting of: 7 octets of preamble. 2. The DUT discarded the fragment consisting of: 7 octets of preamble and SFD. 3. The DUT discarded the fragment consisting of: 7 octets of preamble, SFD and the MAC address of the DUT. 4. The DUT discarded the fragment consisting of: 7 octets of preamble, SFD, the MAC address of the DUT and an arbitrary source address. c. The reception of fragments or runts did not interfere with the reception of valid MAC frames. Ethernet Data Link Layer 3 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # 4.1.3 Test Label Reception of Frames Greater than the Maximum Permitted Frame Size a b c d Result PASS Not Applicable PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) discards frames greater than maxFrameSize. a. b. c. d. DUT should discard or truncate though not transmit, untagged frames greater than 1518 bytes in length DUT should discard or truncate though not transmit, tagged frames greater than 1522 bytes in length The DUT should discard or truncate and respond to the test frames containing two CRC values. The DUT shall not transmit, untagged frames greater than 1518 bytes in length. If the DUT accepted these test frames and discarded the test frames of the same size in part A, then the DUT must truncate and respond to the test frames of the same size with a valid CRC value in bytes 1515 – 1518 and an invalid CRC in the last 4 bytes of the frame. (Otherwise this would indicate that the DUT improperly checked the CRC value twice, once before truncation and once after.) The DUT may optionally log the test frames as frameTooLong errors. All valid frames preceding and following the test frames should be replied to. Comments on Test Results a. b. c. d. The DUT was observed toreply to untagged frames up to 1518 bytes in length. The DUT was observed to discard untagged frames greater than 1518 bytes in length. The DUT was not observed to support VLAN tagging. The DUT was observed to discard the untagged frames that contained a valid 1518 byte frame. The reception of the test frames did not interfere with the reception of valid MAC frames. Test # Test Label 4.1.4 Frames with Length Errors a b Result Not Applicable Not Applicable Comments on Test Procedure Purpose: To verify that the device under test (DUT) discards frames with invalid length values. a. b. The DUT should discard all frames with invalid length values. All valid frames preceding and following the test frames should be replied to. Comments on Test Results Note: The properly encapsulated 64-byte 802.3 SNAP frame was not replied to by the DUT. The remaining frame length tests could not be performed because the DUT did not respond to the 802.3 SNAP frame. Ethernet Data Link Layer 4 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # Test Label 4.1.5 Receive Frames with Excess Pad a b Result Not Applicable Not Applicable Comments on Test Procedure Purpose: To verify that the device under test (DUT) accepts frames with excessive padding. a. b. The DUT should accept each test frame and reply with a frame that does not contain padding. The reception of test frames should not affect the reception of the valid request frames. Comments on Test Results Note: The properly encapsulated 64-byte 802.3 SNAP frame was not replied to by the DUT. The remaining frame length tests could not be performed because the DUT did not respond to the 802.3 SNAP frame. Test # Test Label a 4.1.6 Jabber Frame Reception and Recovery b Result PASS PASS REFER TO COMMENTS Comments on Test Procedure Purpose: To verify that the device under test (DUT) is able to withstand and recover from the reception of worst-case jabber transmissions. a. b. The DUT should discard jabber frames. All valid frames preceding and following the test frames should be replied to. Comments on Test Results b. When receciving a jabber frame followed by a valid frame separated by a minimum inter frame gap the DUT was observed to delay the response to the valid frame until another frame was received by the DUT. When this occurred the DUT was observed to send a reply to both the frame that was being delayed as well as the new frame. Ethernet Data Link Layer 5 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # Test Label 4.1.7 Start of Frame Delimiter Error Reception and Recovery a b c Result PASS PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) discards frames which do not contain a valid Start of Frame Delimiter (SFD). a. b. c. The DUT should discard a test frame where the SFD is replaced by another byte of preamble (bit pattern 10101010) in an otherwise valid frame. The DUT should discard a test frame where the bit pattern 10011011 replaces the SFD in an otherwise valid frame. All valid frames preceding and following the test frames should be replied to. Comments on Test Results a. b. c. The DUT was observed to discard the test frame. The DUT was observed to discard the frame with 7 bytes of preamble and 10011011 in place of an SFD. The reception of the test frames did not interfere with the reception of valid MAC frames. Test # 4.1.8 Test Label Frames with Alignment Errors a b c Result PASS PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) detects and discards frames with alignment errors. a. b. c. The DUT should truncate the test frame to the nearest octet and accept it. The DUT should truncate the test frame to the nearest octet, discard it and report an alignmentError. The reception of valid frames preceding and following the test frames separated by the minimum inter-frame gap should not be affected. Comments on Test Results a. The DUT was observed to accept the test frame. b. The DUT was observed to discard the test frame. The statistics of the DUT were unavailable during testing. c. The reception of the test frames did not interfere with the reception of valid MAC frames. Ethernet Data Link Layer 6 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # 4.1.9 Test Label Preamble Error Reception and Recovery a b c d Result PASS PASS PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) accepts frames which contain errors with preamble. a. b. c. d. The DUT should accept a frame, which has preamble replaced with: 10101010 00000000 00000000 00000000 00000000 00000000 00000000 10101011 The DUT should accept a frame, which has preamble replaced with: 10101010 01111111 11111111 11111111 11111111 11111111 11111111 10101011 The DUT should accept a frame, which has preamble replaced with: 10101010 10101010 10101010 10101010 10101010 10101000 10101111 10101011 All valid frames preceding and following the test frames should be replied to. Comments on Test Results a. b. c. d. The DUT was observed to accept the frame with preamble replaced with above-mentioned sequence. The DUT was observed to accept the frame with preamble replaced with above-mentioned sequence. The DUT was observed to accept the frame with preamble replaced with above-mentioned sequence. The reception of the test frames did not interfere with the reception of valid MAC frames. Ethernet Data Link Layer 7 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Group 2: Test # Test Label 4.2.1 Transmit proper SFD and preamble Result REFER TO COMMENTS Comments on Test Procedure Purpose: To verify that the device under test (DUT) properly encapsulates a frame with preamble and SFD, and that the combination of the two is 8 bytes. The DUT should always encapsulate frames with a combined 8 bytes of preamble and SFD. Comments on Test Results Note: No preamble or SFD problems were observed while testing the DUT. However this test is designed to test the DUT's transmit abilities under stressful traffic loads. When testing End Stations they will typically not transmit rapidly enough for us to determine a Pass/Fail indication. Test # Test Label 4.2.2 Transmission of Minimum interFrameGa p Result REFER TO COMMENTS Comments on Test Procedure Purpose: To verify that the device under test (DUT) enforces the minimum interFrameGap (IFG) of 96 bit times. The DUT should not transmit frames separated by less than minimum IFG, even under the most stressful situations. Comments on Test Results Note: No inter-frame gap problems were observed while testing the DUT. However this test is designed to test the DUT's transmit abilities under stressful traffic loads. When testing End Stations they will typically not transmit rapidly enough for us to determine a Pass/Fail indication. Test # Test Label 4.2.3 Transmit proper length within the Length/Type Field Comments on Test Procedure Result Not Applicable Purpose: To verify that the device under test (DUT) properly calculates the length of the data in the Data/Pad field. The DUT should transmit 802.3 Ethernet frames with a length value equal to the amount of data in the Data/Pad field. If the frame is greater than minFrameSize that frame should not contain pad. Comments on Test Results Note: The properly encapsulated 64-byte 802.3 SNAP frame was not replied to by the DUT. The remaining frame length tests could not be performed because the DUT did not respond to the 802.3 SNAP frame. Ethernet Data Link Layer 8 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # 4.2.4 Comments on Test Procedure Test Label Compute and transmit proper CRC Result PASS Purpose: To verify that the device under test (DUT) correctly calculates the CRC-32 value for the frame being transmitted and assigns it to the FCS field. The CRC-32 value transmitted by the DUT should match the value calculated by the testing station. Comments on Test Results The CRC calculated by the DUT was verified to be correct by the testing platform. Test # 4.2.5 Comments on Test Procedure Test Label Receive variable preamble Result PASS Purpose: To verify that the device under test (DUT) accepts frames with varied amounts of preamble. The DUT should accept frames that contain at least 1 byte of preamble followed by an SFD. Comments on Test Results The DUT was observed to accept frames with 1 to 64 bytes of preamble and an SFD. Test # 4.2.6 Test Label Receive all Frame Sizes a b Result PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) accepts all valid sized frames. a. b. The DUT should accept all valid untagged frames from minFrameSize (64 bytes) to maxUntaggedFrameSize (1518 bytes). If VLAN tagging is supported, the DUT should accept all tagged frames from minFrameSize (64 bytes) to maxUntaggedFrameSize (1518 bytes). The DUT may optionally accept tagged frames 1519 bytes to maxUntaggedFrameSize + qTagPrefixSize (1522 bytes). Comments on Test Results a. b. The DUT was observed to accept all of the different untagged frame sizes. The DUT was observed to accept all of the different tagged frame sizes from 64 bytes to 1522 bytes in length. Ethernet Data Link Layer 9 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # 4.2.7 Comments on Test Procedure Test Label Reception of minimum interFrameGap Result PASS Purpose: To verify that device under test (DUT) is capable of receiving frames separated by a minimum interFrameGap The DUT should accept two frames separated by minimum IFG (96 bit times). Comments on Test Results The DUT was observed to accept both frames that were separated by a minimum IFG. Ethernet Data Link Layer 10 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Group 3: Test # 4.3.1 Comments on Test Procedure Test Label Does not Defer Result PASS Purpose: To verify that the device under test (DUT) does not defer to carrier sense while in full duplex mode. The DUT should not defer the transmission of a frame in waiting to a long carrier event while in full duplex mode. Comments on Test Results The DUT was not observed to defer while in full duplex mode. Test # 4.3.2 Comments on Test Procedure Test Label No Collisions Result PASS Purpose: To verify that no collisions occur when the device under test (DUT) is in full duplex mode. The DUT should not detect and enforce a collision when frames are being transmitted and received simultaneously in full duplex mode. Comments on Test Results No collisions or collision fragments were observed while in full duplex mode. Ethernet Data Link Layer 11 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Group 4: Test # 4.4.1 Test Label Collisions During Preamble and SFD within slotTime a b Result PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) properly detects and enforces collisions that occur during the transmission of preamble and SFD while within slotTime. a. b. The DUT should detect the collision, complete the transmission of preamble and SFD, and transmit a 32-bit jam pattern. The DUT should attempt to retransmit the frame involved in the collision. Comments on Test Results a. b. The DUT was observed to detect the collision, complete the transmission of the preamble and SFD and transmit a 32-bit jam pattern. The DUT was observed to retransmit the frame. Test # 4.4.2 Test Label Collisions in Data within slotTime a b Result PASS PASS Comments on Test Procedure Purpose: To verify that the device under test (DUT) properly detects and enforces collisions that occur during the transmission of data while with slotTime. a. b. The DUT should detect the collision, cease transmission of data and transmit a 32-bit jam pattern. The DUT should attempt to retransmit the frame involved in the collision. Comments on Test Results a. b. The DUT was observed to detect the collision, cease transmission of data and transmit a 32-bit jam pattern. The DUT was observed to retransmit the frame. Ethernet Data Link Layer 12 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Test # Test Label 4.4.3 Collisions in Data Outside of slotTime while Not Bursting A B Result PASS Informative Comments on Test Procedure Purpose: To verify that the device under test (DUT) properly detects and enforces collisions that occur during the transmission of data outside of slotTime while not in burst mode. a. b. The DUT should detect the late collision, cease transmission of data and transmit a 32-bit jam pattern. The DUT may or may not attempt to retransmit a frame involved in a late collision. Comments on Test Results a. b. The DUT was observed to detect the collision, cease transmission of data and transmit a 32-bit jam pattern. The DUT was not observed to retransmit the frame. Ethernet Data Link Layer 13 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Group 5: Test # 4.5.1 Comments on Test Procedure Test Label Defer to Carrier Sense while Frame Waiting Result PASS Purpose: To verify that the device under test (DUT) defers to carrier sense when it has a frame waiting to be transmitted. While in half duplex, the DUT should delay the transmission of a frame in waiting until the completion of a long carrier event Comments on Test Results The DUT was observed to properly defer to carrier events. Test # 4.5.2 Comments on Test Procedure Test Label Deference after Collision Result PASS Purpose: To verify that the device under tes t (DUT) continues to observe the deference process when it is attempting to retransmit after a collision. After a collision has occurred, in half duplex, the DUT should delay any retransmission attempts until carrier sense is no longer asserted. Comments on Test Results The DUT was observed to properly defer to carrier events. Test # 4.5.3 Comments on Test Procedure Test Label InterFrame Spacing Result PASS Purpose: To verify that the device under test (DUT) observes the 2/3 rule for interFrameSpacing. While in half duplex, the DUT should cause a collision to occur if carrier sense is assert during interFrameSpacingPart2, which should be at least the last third of IFS but may be as large as IFS (96 bit times). By observing when the DUT starts causing collisions it is possible to determine the size of interFrameSpacing parts 1 and 2. The interFrameSpacingPart1 of the DUT should not be greater than 2/3 of IFS (64 bit times). Comments on Test Results The value of interFrameSpacingPart1 was observed to be 56 ± 4 bit times. Ethernet Data Link Layer 14 MAC Test Suite Ethernet MAC Test Suite Actel Platform 8051 Co re10/100 MAC testing board Group 6: Test # 4.6.1 Comments on Test Procedure Test Label Retransmission Attempt Limit Result PASS Purpose: To verify that the device under test (DUT) allows a maximum of 15 attempts for retransmission after a collision. After a collision has occurred the DUT should attempt to retransmit the frame involved in the collision. Should multiple collisions occur during the attempted transmission of a frame, the DUT should continue to attempt to retransmit the frame until the number of attempts equals attemptLimit (16). The DUT should, therefore, not attempt to transmit a frame more than 16 times (initial attempt + 15 retransmission attempts). The collision fragments should be properly formed. Comments on Test Results Upon multiple collisions with a frame being transmitted from the DUT, the testing station captured a total of 16 collision fragments. The collision fragments were properly formed. Test # Test Label 4.6.2 Truncated Binary Exponential BackOff Test a b Result PASS FAIL Comments on Test Procedure Purpose: To verify that the device under test (DUT) correctly calculates the time to wait before attempting retransmission of a frame. a. b. The combined average backOff time for each backOff attempt should not be more aggressive than the expected average for that attempt. The variable ‘r’ should be a uniformly distributed random number in the range of [0, 2k – 1] where k=min(n,10). Comments on Test Results a. b. The average backoff for the retransmission attempts was not observed to be significantly more aggressive than the allowable range to over a 99% Confidence level. According to the histogram data we can say with over a 99% Confidence level that the backoff times chosen were not uniformly distributed over the standard specified backoff ranges, specifically retry attempt 12. For further information please refer to the enclosed histograms. If you would like a copy of the gathered data please let us know. Ethernet Data Link Layer 15 MAC Test Suite Histogram of the Backoff of the First Retransmission Attempt 18000 16000 14000 Occurrences 12000 10000 8000 6000 4000 2000 0 0 1 Backoff in Slot Times Histogram of the Backoff of the Second Retransmission Attempt 9000 8000 7000 Occurrences 6000 5000 4000 3000 2000 1000 0 0 1 2 Backoff in Slot Times 3 Histogram of the Backoff of the Third Retransmission Attempt 4500 4000 3500 Occurrences 3000 2500 2000 1500 1000 500 0 0 1 2 3 4 Backoff in Slot Times 5 6 7 Histogram of the Backoff of the Fourth Retransmission Attempt 2500 Occurrences 2000 1500 1000 500 0 0 1 2 3 4 5 6 7 8 Backoff in Slot Times 9 10 11 12 13 14 15 Histogram of the Backoff of the Fifth Retransmission Attempt 1200 1000 Occurrences 800 600 400 200 0 0 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 Backoff in Slot Times Histogram of the Backoff of the Sixth Retransmission Attempt 600 500 300 200 100 Backoff in Slot Times 62 60 58 56 54 52 50 48 46 44 42 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 0 Occurrences 400 Histogram of the Backoff of the Seventh Retransmission Attempt 350 300 200 150 100 50 Backoff in Slot Times 124 120 116 112 108 104 100 96 92 88 84 80 76 72 68 64 60 56 52 48 44 40 36 32 28 24 20 16 12 8 4 0 0 Occurrences 250 Histogram of the Backoff of the Eighth Retransmission Attempt 180 160 140 100 80 60 40 20 Backoff in Slot Times 248 240 232 224 216 208 200 192 184 176 168 160 152 144 136 128 120 112 104 96 88 80 72 64 56 48 40 32 24 16 8 0 0 Occurrences 120 Backoff in Slot Times 510 495 480 465 450 435 420 405 390 375 360 345 330 315 300 285 270 255 240 225 210 195 180 165 150 135 120 105 90 75 60 45 30 15 0 Number of Occurrences Histogram of the Backoff of the Ninth Retransmission Attempt 120 100 80 60 40 20 0 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 Number of Occurrences Histogram of the Backoff of the 10th Retransmission Attempt 70 60 50 40 30 20 10 0 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 Number of Occurrences Histogram of the Backoff of the 11th Retransmission Attempt 70 60 50 40 30 20 10 0 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 Number of Occurrences Histogram of the Backoff of the 12th Retransmission Attempt 60 50 40 30 20 10 0 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 Number of Occurrences Histogram of the Backoff of the 13th Retransmission Attempt 90 80 70 60 50 40 30 20 10 0 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 Number of Occurrences Histogram of the Backoff of the 14th Retransmission Attempt 80 70 60 50 40 30 20 10 0 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 Number of Occurrences Histogram of the Backoff of the 15th Retransmission Attempt 60 50 40 30 20 10 0 Histogram of the Backoff of the Combined 10th-15th Retransmission Attempts 400 350 300 200 150 100 50 Backoff in Slot Times 1015 986 957 928 899 870 841 812 783 754 725 696 667 638 609 580 551 522 493 464 435 406 377 348 319 290 261 232 203 174 145 116 87 58 29 0 0 Occurrences 250