Hi Parthiban, Sorry for the late answer, I was quite busy the last few days. On Mon, 2024-03-25 at 13:24 +0000, Parthiban.Veerasooran@xxxxxxxxxxxxx wrote: > Hi Benjamin Bigler, > > Thank you for your testing and feedback. It would be really helpful to > bring the driver to a good shape. We really appreciate your efforts on this. > > On 24/03/24 5:25 pm, Benjamin Bigler wrote: > > [Some people who received this message don't often get email from benjamin@xxxxxxxxxx. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ] > > > > EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe > > > > Hi Parthiban > > > > I hope I send this in the right context as it is not related to just one patch or > > some specific code. > > > > I conducted UDP load testing using three i.MX8MM boards in conjunction with the > > LAN8651. The setup involved one board functioning as a server, which is just > > echoing back received data, while the remaining two boards acted as clients, > > sending UDP packets of different sizes in various bursts to the server. > > Due to hardware constraints, the SPI bus speed was limited to 15 MHz, which might > > have influenced the results. > > > > During the tests I experienced some issues: > > > > - The boards just start receiving after first sending something (ping another board). > > Some measurements showed that the irq stays asserted after init. This makes sense > > as far as I understand the chapter 7.7 of the specification, the irq is deasserted > > on reception of the first data header following CSn being asserted. As a workaround > > I trigger the thread at the end of oa_tc6_init. > It looks like the IRQ is asserted on RESET completion and expects a data > chunk from host to deassert the IRQ. I used to test the driver in RPI 4 > using iperf3. For some reason I never faced this issue, may be when the > network device is being registered there might be some packet > transmission which leads to deliver a data chunk so that the IRQ is > deasserted. Thanks for the workaround. I think that would be the > solution to solve this issue. Adding the below lines in the end of the > function oa_tc6_init() will trigger the oa_tc6_spi_thread_handler() to > perform an empty data chunk transfer which will deassert the IRQ before > starting the actual data transfer. I have ipv6 disabled and use static ipv4 addresses. That could be the reason why on my side no packet is sent. > > /* oa_tc6_sw_reset_macphy() function resets and clears the MAC-PHY reset > * complete status. IRQ is also asserted on reset completion and it is > * remain asserted until MAC-PHY receives a data chunk. So performing an > * empty data chunk transmission will deassert the IRQ. Refer section > * 7.7 and 9.2.8.8 in the OPEN Alliance specification for more details. > */ > tc6->int_flag = true; > wake_up_interruptible(&tc6->spi_wq); Perfect, thats the same I added and also works on my side. > > > > - If there is a lot of traffic, the receive buffer overflow error spams the log. > > > > - If there is a lot of traffic, I got various kernel panics in oa_tc6_update_rx_skb. > > Mostly because more data to rx_skb is added than allocated and sometimes because > > rx_skb is null in oa_tc6_update_rx_skb or oa_tc6_prcs_rx_frame_end. Some debugging > > with a logic analyzer showed that the chip is not behave correctly. There is more > > bytes between start_valid and end_valid than there should be. Also there > > seems to be 2 end_valid without a start_valid between. What is common is that the incorrect > > frame starts in a chunk where end_valid and start_valid is set. > > In my opinion its a problem in the chip (maybe related to the errata in the next point) > > but the driver should be resilent and just drop the packet and not cause a kernel panic. > Usually I run into this issue "receive buffer overflow" when I run RPI 4 > with default cpu governor setting which is "ondemand". In this case, > even though if I set SPI clock speed as 15 MHz the RPI 4 core clock is > clocking down when it is idle which leads delivering half of the > configured SPI clock speed around 5.9 MHz. So the systems like RPI 4 > need performance mode enabled to get the proper clock speed for SPI. > Refer below link for more details. > > https://github.com/raspberrypi/linux/issues/3381#issuecomment-1144723750 > > I used to enable performance mode using the below command. > > echo performance | sudo tee > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor > /dev/null > > So please ensure the SPI clock speed using a logic analyzer to get the > maximum throughput without receive buffer overflow. > > Of course, I agree that the driver should not crash in case of receive > buffer overflow. By referring your investigations, I understand that the > buffers in the MAC-PHY is being continuously overwritten again and again > as the host is very slow to read the data from the MAC-PHY buffers > through SPI which alters the descriptors. There might be two reasons why > we run into this situation. > 1. The host is busy doing something else and delays to initiate SPI even > though SPI clock speed is 15 MHz. > 2. The SPI clock speed is less than 15 MHz. Sorry there is a missunderstanding between us. The receive buffer overflow is not causing any harm except filling the log. In my setup I get in one day about 35000 entries. I am not sure if its appropriate to log these errors. The SPI Frequency is at 14.8 MHz. If I just have 2 boards connected, I am not able to reproduce this. Only with 3 boards when 2 boards sends multiple big ethernet frames (1512 byte per Frame) to one, I get these log entries. The latency seems to be quite low, from IRQ to start reading first frame it takes always less than 500us. Also the boards are just running the udp test. > > I use the below iperf3 setup for my testing and never faced the driver > crash issue even though faced "receive buffer overflow" error when I run > RPI 4 with "ondemand" default mode. > > Node 0 - Raspberry Pi 4 with LAN8650 MAC-PHY > $ iperf3 -s > Node 1 - Raspberry Pi 4 with EVB-LAN8670-USB USB Stick > $ iperf3 -c 192.168.5.100 -u -b 10M -i 1 -t 0 > > and vice versa. > > I never faced "receive buffer overflow" error when I run RPI 4 with > "performance" mode enabled and even though all the cores are stressed > using the below command, > > $ yes >/dev/null & yes >/dev/null & yes >/dev/null & yes >/dev/null & > > Can you share more details about your testing setup and applications you > use, so that I will try to reproduce the issue in my setup to debug the > driver? I use a internal tool which does some stress tests using udp. Unfortunately, I am not allowed to publish it, but a colleague works on a rust implementation, which we can publish, but its not fully ready yet. On one board the tool is running in server mode. It just echoes back the received data. On the 2 other boards the tool is running in client mode. It sends various sized udp-packets in different bursts and then checks if it receives the same data in the same order. The crashes only happens when ZARFE is not set (with Rev B0). When the crash happens, I see on the logic analyzer that there are more bytes than mtu + headers between the frame where start_valid is set and the frame where end_valid is set. Then this happens: [ 437.155673] skbuff: skb_over_panic: text:ffff80007a8c2bd8 len:1600 put:64 head:ffff00000de28080 data:ffff00000de280c0 tail:0x680 end:0x640 dev:eth1 [ 437.168987] kernel BUG at net/core/skbuff.c:192! [ 437.173612] Internal error: Oops - BUG: 00000000f2000800 [#1] PREEMPT SMP [ 437.180407] Modules linked in: ppp_async crc_ccitt ppp_generic slhc lan865x oa_tc6 bec_infoo(O) tpm_tis_spi tpm_tis_core spi_imx imx_sdma [ 437.196016] CPU: 1 PID: 455 Comm: oa-tc6-spi-thre Tainted: G O 6.6.11- gce336e2c2bc3-dirty #1 [ 437.205853] Hardware name: Toradex Verdin iMX8M Mini on FUMU (DT) [ 437.212820] pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 437.219790] pc : skb_panic+0x58/0x5c [ 437.223376] lr : skb_panic+0x58/0x5c [ 437.226959] sp : ffff80008362bd90 [ 437.230278] x29: ffff80008362bda0 x28: 0000000000000000 x27: ffff000001066878 [ 437.237426] x26: 000000000000001e x25: 00000000000007f8 x24: ffff0000010cea80 [ 437.244571] x23: 00000000f0f0f0f1 x22: 000000000000001f x21: 0000000000000000 [ 437.251720] x20: ffff0000010ceaa8 x19: 000000003f20003f x18: ffffffffffffffff [ 437.258867] x17: ffff7ffffded9000 x16: ffff800080008000 x15: 073a0764076e0765 [ 437.266015] x14: 0720073007380736 x13: ffff8000823d1f58 x12: 0000000000000534 [ 437.273162] x11: 00000000000001bc x10: ffff800082429f58 x9 : ffff8000823d1f58 [ 437.280310] x8 : 00000000ffffefff x7 : ffff800082429f58 x6 : 0000000000000000 [ 437.287455] x5 : 000000000000bff4 x4 : 0000000000000000 x3 : 0000000000000000 [ 437.294606] x2 : 0000000000000000 x1 : ffff000001223b00 x0 : 0000000000000087 [ 437.301753] Call trace: [ 437.304203] skb_panic+0x58/0x5c [ 437.307436] skb_find_text+0x0/0xf0 [ 437.310933] oa_tc6_spi_thread_handler+0x438/0x880 [oa_tc6] [ 437.316523] kthread+0x118/0x11c [ 437.319758] ret_from_fork+0x10/0x20 [ 437.323343] Code: f90007e9 b940b908 f90003e8 97ca3c34 (d4210000) [ 437.329446] ---[ end trace 0000000000000000 ]--- Sometimes there are 2 end_valid after eachother without a start_valid between. Then this happens: [ 469.737297] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000074 [ 469.746137] Mem abort info: [ 469.748950] ESR = 0x0000000096000004 [ 469.752709] EC = 0x25: DABT (current EL), IL = 32 bits [ 469.758036] SET = 0, FnV = 0 [ 469.761098] EA = 0, S1PTW = 0 [ 469.764252] FSC = 0x04: level 0 translation fault [ 469.769144] Data abort info: [ 469.772033] ISV = 0, ISS = 0x00000004, ISS2 = 0x00000000 [ 469.777529] CM = 0, WnR = 0, TnD = 0, TagAccess = 0 [ 469.782594] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0 [ 469.787921] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000043c32000 [ 469.794377] [0000000000000074] pgd=0000000000000000, p4d=0000000000000000 [ 469.801184] Internal error: Oops: 0000000096000004 [#1] PREEMPT SMP [ 469.807459] Modules linked in: ppp_async crc_ccitt ppp_generic slhc lan865x oa_tc6 bec_infoo(O) tpm_tis_spi tpm_tis_core spi_imx imx_sdma [ 469.823064] CPU: 2 PID: 456 Comm: oa-tc6-spi-thre Tainted: G O 6.6.11- g350ed394a6ca-dirty #1 [ 469.832903] Hardware name: Toradex Verdin iMX8M Mini on FUMU (DT) [ 469.839871] pstate: 00000005 (nzcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 469.846841] pc : skb_put+0xc/0x6c [ 469.850169] lr : oa_tc6_spi_thread_handler+0x438/0x880 [oa_tc6] [ 469.856106] sp : ffff80008376bdb0 [ 469.859424] x29: ffff80008376bdb0 x28: 0000000000000000 x27: ffff00000194c080 [ 469.866573] x26: 0000000000000000 x25: 0000000000000000 x24: ffff000001095c80 [ 469.873720] x23: 00000000f0f0f0f1 x22: 000000000000001f x21: 0000000000000000 [ 469.880870] x20: ffff000001095ca8 x19: 000000003f20003f x18: 0000000000000000 [ 469.888023] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000000 [ 469.895174] x14: 0000031acf8b86d8 x13: 0000000000000000 x12: 0000000000000000 [ 469.902321] x11: 0000000000000002 x10: 0000000000000a60 x9 : ffff80008376b970 [ 469.909467] x8 : ffff00007fb6e580 x7 : 000000000194b080 x6 : 0000000000000000 [ 469.916616] x5 : 0000000000000000 x4 : 0000000000000000 x3 : 000000000000fc80 [ 469.923765] x2 : 0000000000000001 x1 : 0000000000000040 x0 : 0000000000000000 [ 469.930915] Call trace: [ 469.933365] skb_put+0xc/0x6c [ 469.936342] oa_tc6_spi_thread_handler+0x438/0x880 [oa_tc6] [ 469.941929] kthread+0x118/0x11c [ 469.945166] ret_from_fork+0x10/0x20 [ 469.948752] Code: d65f03c0 d503233f a9bf7bfd 910003fd (b9407406) [ 469.954854] ---[ end trace 0000000000000000 ]--- If interested I can try to get a recording with the logic analyzer and send it to you. By the way in the other answer you attached a screenshot of the logic analyzer and you have a very nice HLA for oa_tc6. Are they open-source or are there any plans to publish them? > > > > - Sometimes the chip stops working. It always asserts the irq but there is no data (rca=0) > > and also exst is not active. I found out that there is an errata (DS80001075) point s3 > > that explains this. I set the ZARFE bit in CONFIG0. This also fixes the point above. > > The driver now works since about 2.5 weeks with various load with just one loss of frame > > error where I had to reboot the system after about 4 days. > It is good to hear that the driver works fine with the above changes. As > mentioned in the errata, this continuous interrupt issue is a known > issue with LAN8651 Rev.B0. Switching to LAN8651 Rev.B1 will solve this > issue and no need of any workaround. Setting ZARFE bit in the CONFIG0 > will solve the continuous interrupt issue but don't know how the above > "receive buffer overflow" issue also solved. I think it is a good idea > to test with LAN8651 Rev.B1 without setting ZARFE bit once. It would be > interesting to see the result. I am always using LAN8651 Rev.B1 for my > testing. Unfortunately I just have LAN8651 Rev. B0 Chips. Are you sure that the Rev B1 has the issue fixed? The errata here says that B1 is affected too: https://ww1.microchip.com/downloads/aemDocuments/documents/AIS/ProductDocuments/Errata/LAN8650-1-Errata-80001075.pdf > > I should be able to reproduce the "receive buffer overflow" issue and > consequently kernel crash in my setup with LAN8651 Rev.B1 so that I can > investigate the issue further. As I am not able to reproduce in my RPI > 4, I need your support for the tests and applications you used in your > setup. > > > > Is there a reason why you removed the netdev watchdog which was active in v2? > When the timeout occurs, there is no further action except increasing > tx_errors. Not seeing this except USB-to-Ethernet which can be removed > unexpectedly. But this is SPI interface which will not be removed > unexpectedly as it is a platform device. That's why we removed this. > > Best regards, > Parthiban V > > > > Thanks, > > Benjamin Bigler > > > Thanks, Benjamin Bigler
