On Wed Oct 9, 2024 at 6:29 PM EEST, Usama Arif wrote: > Hi, > > We (meta) are seeing the below warning in production machines for all kernels from 5.12 to 6.11 during boot which results in tpm probe failing: > > > [ 6.388599] ------------[ cut here ]------------ > [ 6.397804] memremap attempted on mixed range 0x000000005f54e018 size: 0x23ba3cfb > [ 6.412733] WARNING: CPU: 2 PID: 1 at kernel/iomem.c:82 memremap+0x58/0x1b0 > [ 6.426621] Modules linked in: > [ 6.432705] CPU: 2 PID: 1 Comm: swapper/0 Not tainted 6.4.3-0_fbk12_2624_g7d95a0297d81 #1 > [ 6.449020] Hardware name: Quanta Twin Lakes MP/Twin Lakes Passive MP, BIOS F09_3A23 12/08/2020 > [ 6.466377] RIP: 0010:memremap+0x58/0x1b0 > [ 6.474369] Code: 41 89 c4 83 f8 02 75 26 80 3d fc 33 66 01 00 75 57 c6 05 f3 33 66 01 01 48 89 e6 48 c7 c7 3b 98 34 82 4c 89 f2 e8 38 96 bc ff <0f> 0b eb 3a 41 f6 c7 01 75 04 31 db eb 13 45 85 e4 74 69 48 89 ef > [ 6.511837] RSP: 0000:ffffc900000378f8 EFLAGS: 00010286 > [ 6.522258] RAX: 0000000000000045 RBX: 0000000000000000 RCX: ffffffff82c5fdb8 > [ 6.536490] RDX: 0000000000000000 RSI: c0000000fffeffff RDI: 000000000000ffff > [ 6.550723] RBP: 000000005f54e018 R08: ffffffff82de0190 R09: 0000000000000000 > [ 6.564956] R10: ffffffff82c5fde0 R11: 3fffffffffffffff R12: 0000000000000002 > [ 6.579187] R13: 0000000000000000 R14: 0000000023ba3cfb R15: 0000000000000001 > [ 6.593421] FS: 0000000000000000(0000) GS:ffff88903f880000(0000) knlGS:0000000000000000 > [ 6.609559] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > [ 6.621026] CR2: 0000000000000000 CR3: 0000000062c0a001 CR4: 00000000007706e0 > [ 6.635257] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > [ 6.649489] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > [ 6.663724] PKRU: 55555554 > [ 6.669115] Call Trace: > [ 6.673986] <TASK> > [ 6.678162] ? __warn+0x9f/0x130 > [ 6.684595] ? memremap+0x58/0x1b0 > [ 6.691371] ? report_bug+0xcc/0x150 > [ 6.698499] ? handle_bug+0x3d/0x70 > [ 6.705449] ? exc_invalid_op+0x16/0x40 > [ 6.713093] ? asm_exc_invalid_op+0x16/0x20 > [ 6.721433] ? memremap+0x58/0x1b0 > [ 6.728211] ? memremap+0x58/0x1b0 > [ 6.732866] Freeing initrd memory: 51140K > [ 6.734986] tpm_read_log_efi+0x7d/0x1e0 > [ 6.750795] tpm_bios_log_setup+0x51/0x170 > [ 6.758959] tpm_chip_register+0x36/0x220 > [ 6.766948] tpm_tis_core_init+0x438/0x630 > [ 6.775112] tpm_tis_init+0x103/0x190 > [ 6.782410] tpm_tis_plat_probe+0x87/0x90 > [ 6.790401] platform_probe+0x2f/0x60 > [ 6.797705] really_probe+0x1ec/0x340 > [ 6.805002] driver_probe_device+0x1e/0x80 > [ 6.813184] __driver_attach+0x10e/0x1a0 > [ 6.821018] ? driver_attach+0x20/0x20 > [ 6.828491] bus_for_each_dev+0x64/0xa0 > [ 6.836136] bus_add_driver+0x196/0x210 > [ 6.843781] ? __initstub__kmod_tpm__370_527_tpm_init4+0x100/0x100 > [ 6.856109] driver_register+0x5e/0xf0 > [ 6.863580] __initstub__kmod_tpm_tis__319_478_init_tis6+0x7f/0xc0 > [ 6.875908] ? add_device_randomness+0x62/0x70 > [ 6.884768] do_one_initcall+0xad/0x1f0 > [ 6.892414] ? alloc_pages+0x120/0x2c0 > [ 6.899885] ? security_kernfs_init_security+0x2a/0x40 > [ 6.910132] ? idr_alloc_cyclic+0xaa/0x110 > [ 6.918295] ? idr_alloc_cyclic+0xaa/0x110 > [ 6.926458] ? alloc_pages+0x15c/0x2c0 > [ 6.933928] ? number+0x1da/0x400 > [ 6.940531] ? ida_alloc_range+0x193/0x410 > [ 6.948696] ? parse_args+0x14b/0x380 > [ 6.955996] kernel_init_freeable+0x1b1/0x2a0 > [ 6.964682] ? rest_init+0xb0/0xb0 > [ 6.971461] kernel_init+0x16/0x1a0 > [ 6.978410] ret_from_fork+0x1f/0x30 > [ 6.985536] </TASK> > [ 6.989885] ---[ end trace 0000000000000000 ]--- > [ 6.999089] Could not map UEFI TPM log table payload! > > > The memremap being attempted is from 0x5f54e018 to 0x830f1d13, which looking at /proc/iomem, is overlapping all the below regions. > > cat /proc/iomem > ... > 61000000-61ffffff : Kernel code > 62000000-62bf4fff : Kernel rodata > 62c00000-634e28bf : Kernel data > 63b8e000-641fffff : Kernel bss > 64dbb000-64dbbfff : Reserved > 64dbc000-69c89fff : System RAM > 69c8a000-6bd89fff : Reserved > 6ad6d018-6ad6d027 : APEI EINJ > 6ad6d048-6ad6d067 : APEI EINJ > 6ad6d078-6ad6d07f : APEI EINJ > 6ad6e018-6ad6e018 : APEI ERST > 6ad6e01c-6ad6e021 : APEI ERST > 6ad6e028-6ad6e039 : APEI ERST > 6ad6e040-6ad6e04c : APEI ERST > 6ad6e050-6ad7004f : APEI ERST > 6bd8a000-6be44fff : ACPI Tables > 6be45000-6c99efff : ACPI Non-volatile Storage > 6c99f000-6ebedfff : Reserved > 6ebee000-6fffffff : System RAM > 70000000-7fffffff : Reserved > 80000000-8fffffff : PCI MMCONFIG 0000 [bus 00-ff] > ... > > The kernel is being booted using kexec. 0x23ba3cfb is 571MB which looks too big. Could it be that the log size grew too big in the previous kernel? Or is it a memory corruption similar to the issue we encountered in [1] and solved in [2]? > > [1] https://lore.kernel.org/all/20240910-juicy-festive-sambar-9ad23a@devvm32600/ > [2] https://lore.kernel.org/all/20240912155159.1951792-2-ardb+git@xxxxxxxxxx/ I'd look at it if I had extra bandwidth but I hope someone who knows better kexec/EFI will take this earliest time I have time to investigate is like few weeks from now because I started on new jobs. So right now for few only strictly TPM focused bugs. Just implying my situation. In a normal situation I'd actually find this quite interesting. I'll check where it is after some weeks. > > Thanks! > Usama BR, Jarkko
