> On 12.05.2021 21:01, Marc Veldman wrote: > > > I?m not sure if this is an interesting data point or not, > > but a warm boot without the card inserted succeeds after > > a cold boot with the card inserted. > > It could explain, why my tests with "same code path" gave different results! I am suspect of 2 things here, something the bios does that leaves the card in a state that alters the loader's disk probing, and that probling itself leaving the device in a state that our kernel does not like. I have it very "odd" that I can boot from a rtsx sd card, ie the kernel gets loaded, but fails at "mountroot" phase due to a sd card timeout. If I simply cycle the sd card in and out of its socket and then give the right command string to mountroot it goes on to multiuser without issue. I'll also note that if I am booting from other disks (nvme or usb) that I get these same sd card timeouts and I have to cycle the card in and out of the socket to use it: rtsx0: <2.0c Realtek RTS525A PCI MMC/SD Card Reader> mem 0xe1000000-0xe1000fff irq 18 at device 0.0 on pci3 rtsx0: pci_read_config() error - reg: 0xeeffffaa rtsx0: Card present mmc0: <MMC/SD bus> on rtsx0 rtsx0: CRC error rtsx0: Transfer fail - status: 0x90010080 rtsx0: CRC error rtsx0: Transfer fail - status: 0x90010080 rtsx0: CRC error rtsx0: Transfer fail - status: 0x90010080 rtsx0: CRC error rtsx0: Transfer fail - status: 0x90010080 rtsx0: Interrupt card inserted/removed rtsx0: Card absent rtsx0: Interrupt card inserted/removed rtsx0: Card present mmc0: <MMC/SD bus> on rtsx0 (The last 5 lines caused by me removeing/inserting the card) Further note that this is a different controller chip version, from a Dell E5470: rtsx0_at_pci0:3:0:0: class=0xff0000 rev=0x01 hdr=0x00 vendor=0x10ec device=0x525a subvendor=0x1028 subdevice=0x06de vendor = 'Realtek Semiconductor Co., Ltd.' device = 'RTS525A PCI Express Card Reader' bar [14] = type Memory, range 32, base 0xe1000000, size 4096, enabled cap 01[80] = powerspec 3 supports D0 D1 D2 D3 current D0 cap 05[90] = MSI supports 1 message, 64 bit enabled with 1 message cap 10[b0] = PCI-Express 2 endpoint max data 256(512) RO max read 512 link x1(x1) speed 5.0(5.0) ASPM disabled(L0s/L1) ClockPM enabled ecap 0001[100] = AER 2 0 fatal 0 non-fatal 1 corrected ecap 0003[148] = Serial 1 00000001004ce000 ecap 0018[158] = LTR 1 ecap 001e[160] = L1 PM Substates 1 > // Lev Serebryakov -- Rod Grimes rgrimes_at_freebsd.orgReceived on Thu May 13 2021 - 17:00:46 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:41:28 UTC