CC to kexec ML. On Mon, Jun 29, 2026 at 11:48:37AM -0300, Desnes Nunes wrote: > When translation is already enabled at boot (e.g. kdump), the vt-d driver > copies context tables from the previous kernel's root table. In scalable > mode, buses that only populate the upper root half (UCTP, devfn >= 0x80) > should be written to ctxt_tbls[tbl_idx + 1] through copy_context_table(). > However, the current copy path always uses tbl[tbl_idx + 0] in this situa- > tion. Since idx wraps to 0 at devfn 0x80 due to a zeroed LCTP, new_ce for > LCTP will be NULL and keep pos equals to 0. Thus, UCTP entries will be co- > pied into tbl[tbl_idx + 0] instead of tbl[tbl_idx + 1], and written after- > wards to root_entry[bus].lo instead of .hi in copy_translation_tables(). > > In short, devices on bus 0x80 with devfn >= 0x80 fail DMA with fault 0x39, > which will break drivers running in kernels with translation pre-enabled. > This fixes NO_PASID DMAR faults for UCTP-only buses such as: > > DMAR: [DMA Read NO_PASID] Request device [80:14.0] fault addr 0xe81759000 > [fault reason 0x39] SM: Present bit in Root Entry is clear > > For instance, this fault yielded to locking issues between systemd and > xHCI, blocking a system's reboot after a vmcore was captured with kdump: > > [ 72.987601] systemd-udevd[246]: usb3: Worker [255] processing SEQNUM=2193 > is taking a long time > [ 132.237566] dracut-initqueue[277]: Timed out while waiting for udev queue > to empty. > [ 202.988014] systemd-udevd[246]: usb3: Worker [255] processing SEQNUM=2193 > killed > [ 202.998059] systemd-udevd[246]: usb3: Worker [255] terminated by signal 9 > (KILL). > ... > [ 206.288378] kdump[569]: saving vmcore complete > ... > [ 206.821258] systemd-shutdown[1]: Rebooting. > [ 246.858495] INFO: task kworker/0:1:11 blocked for more than 122 seconds. > [ 246.865319] Not tainted 7.0.0-clean #1 > [ 246.869663] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables > this message. > [ 246.877623] task:kworker/0:1 state:D stack:0 pid:11 tgid:11 > ppid:2 task_flags:0x4208160 flags:0x00080000 > [ 246.888942] Workqueue: usb_hub_wq hub_event > [ 246.893202] Call Trace: > [ 246.895690] <TASK> > [ 246.897828] __schedule+0x299/0x5c0 > [ 246.901378] schedule+0x27/0x80 > [ 246.904572] schedule_timeout+0xbd/0x100 > [ 246.908565] __wait_for_common+0x97/0x1b0 > [ 246.912644] ? __pfx_schedule_timeout+0x10/0x10 > [ 246.917252] xhci_alloc_dev+0x9e/0x2b0 > [ 246.921068] usb_alloc_dev+0x7a/0x3b0 > [ 246.924795] hub_port_connect+0x285/0x960 > [ 246.928873] hub_port_connect_change+0x94/0x290 > [ 246.933482] port_event+0x4bb/0x840 > [ 246.937030] hub_event+0x141/0x460 > [ 246.940489] process_one_work+0x196/0x390 > [ 246.944569] worker_thread+0x1af/0x320 > [ 246.948383] ? __pfx_worker_thread+0x10/0x10 > [ 246.952724] kthread+0xe3/0x120 > [ 246.955921] ? __pfx_kthread+0x10/0x10 > [ 246.959736] ret_from_fork+0x199/0x260 > [ 246.963550] ? __pfx_kthread+0x10/0x10 > [ 246.967362] ret_from_fork_asm+0x1a/0x30 > [ 246.971355] </TASK> > [ 369.738508] INFO: task systemd-shutdow:1 blocked for more than 122 seconds. > [ 369.745593] Not tainted 7.0.0-clean #1 > [ 369.749935] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables > this message. > [ 369.757897] task:systemd-shutdow state:D stack:0 pid:1 tgid:1 ppid:0 > task_flags:0x400100 flags:0x00080000 > [ 369.769128] Call Trace: > [ 369.771616] <TASK> > [ 369.773752] __schedule+0x299/0x5c0 > [ 369.777299] schedule+0x27/0x80 > [ 369.780493] schedule_preempt_disabled+0x15/0x30 > [ 369.785188] __mutex_lock.constprop.0+0x547/0xac0 > [ 369.789974] device_shutdown+0xac/0x1b0 > [ 369.793877] kernel_restart+0x3a/0x70 > [ 369.797603] __do_sys_reboot+0x147/0x240 > [ 369.801595] do_syscall_64+0x11b/0x6a0 > [ 369.805407] ? handle_mm_fault+0x110/0x350 > [ 369.809574] ? do_user_addr_fault+0x206/0x680 > [ 369.814006] ? irqentry_exit+0x7a/0x4d0 > [ 369.817907] entry_SYSCALL_64_after_hwframe+0x76/0x7e > [ 369.823046] RIP: 0033:0x7fe2958da917 > [ 369.826684] RSP: 002b:00007ffc5c458618 EFLAGS: 00000206 ORIG_RAX: > 00000000000000a9 > [ 369.834383] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: > 00007fe2958da917 > [ 369.841639] RDX: 0000000001234567 RSI: 0000000028121969 RDI: > 00000000fee1dead > [ 369.848893] RBP: 00007ffc5c458790 R08: 0000000000000069 R09: > 00000000ffffffff > [ 369.856148] R10: 0000000000000000 R11: 0000000000000206 R12: > 0000000000000000 > [ 369.863402] R13: 0000000000000000 R14: 00007ffc5c4588b8 R15: > 0000000000000000 > [ 369.870659] </TASK> > [ 369.872888] INFO: task systemd-shutdow:1 is blocked on a mutex likely > owned by task kworker/0:1:11. >
Thanks for attaching the kernel stack trace, which is similar to the one I have encountered recently. And I have applied & tested the patch, works good for me. Tested-by: Tao Liu <[email protected]> > Fixes: 091d42e43d21 ("iommu/vt-d: Copy translation tables from old kernel") > Signed-off-by: Desnes Nunes <[email protected]> > --- > V1 -> V2: Updated commit message and added xHCI stack trace as requested > > v1: https://lore.kernel.org/linux-iommu/[email protected]/T/#t > > drivers/iommu/intel/iommu.c | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c > index 4d0e65bc131d..737936f942a0 100644 > --- a/drivers/iommu/intel/iommu.c > +++ b/drivers/iommu/intel/iommu.c > @@ -1443,7 +1443,7 @@ static int copy_context_table(struct intel_iommu *iommu, > struct context_entry **tbl, > int bus, bool ext) > { > - int tbl_idx, pos = 0, idx, devfn, ret = 0, did; > + int tbl_idx, tbl_slot = 0, idx, devfn, ret = 0, did; > struct context_entry *new_ce = NULL, ce; > struct context_entry *old_ce = NULL; > struct root_entry re; > @@ -1459,10 +1459,9 @@ static int copy_context_table(struct intel_iommu > *iommu, > if (idx == 0) { > /* First save what we may have and clean up */ > if (new_ce) { > - tbl[tbl_idx] = new_ce; > + tbl[tbl_idx + tbl_slot] = new_ce; > __iommu_flush_cache(iommu, new_ce, > VTD_PAGE_SIZE); > - pos = 1; > } > > if (old_ce) > @@ -1484,6 +1483,9 @@ static int copy_context_table(struct intel_iommu *iommu, > } > } > > + /* Track if saving UCTP or LCTP entries in scalable > mode */ > + tbl_slot = ext && devfn >= 0x80 ? 1 : 0; > + > ret = -ENOMEM; > old_ce = memremap(old_ce_phys, PAGE_SIZE, > MEMREMAP_WB); > @@ -1512,7 +1514,7 @@ static int copy_context_table(struct intel_iommu *iommu, > new_ce[idx] = ce; > } > > - tbl[tbl_idx + pos] = new_ce; > + tbl[tbl_idx + tbl_slot] = new_ce; > > __iommu_flush_cache(iommu, new_ce, VTD_PAGE_SIZE); > > -- > 2.54.0 >
