Skip to contentSkip to conversation

Commit d7d8ab87e3e ("PCI: vmd: Switch to msi_create_parent_irq_domain()") causes early-stage reboots on my Dell XPS-9320

DISCUSSION Kenneth Crudup 3 participants 6 messages Aug 6, 2025 – Aug 11, 2025

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

I'm running Linus' master (as of today, cca7a0aae8958c9b1).

If I revert the named commit, I can boot OK. Unfortunately there's no real output before the machine reboots, to help identify the problem.

I have a(n enabled) VMD in my Alderlake machine:

[    0.141952] [      T1] smpboot: CPU0: 12th Gen Intel(R) Core(TM) i7-1280P (family: 0x6, model: 0x9a, stepping: 0x3)

If there's something else I can try, let me know.

Thanks,

-K

Signature · 1 line
Kenneth R. Crudup / Sr. SW Engineer, Scott County Consulting, Orange County CA
#1 lore raw
Quoted 11 lines · Kenneth Crudup

I'm running Linus' master (as of today, cca7a0aae8958c9b1).

If I revert the named commit, I can boot OK. Unfortunately there's no real output before the machine reboots, to help identify the problem.

I have a(n enabled) VMD in my Alderlake machine:

[    0.141952] [      T1] smpboot: CPU0: 12th Gen Intel(R) Core(TM) 

i7-1280P (family: 0x6, model: 0x9a, stepping: 0x3)

If there's something else I can try, let me know.

Thanks for the report. It is unfortunate that there is no output to work with :(

Let me stare at the commit again..

Nam

#2 lore raw
Quoted 12 lines · Nam Cao
Quoted 10 lines · Kenneth Crudup

I'm running Linus' master (as of today, cca7a0aae8958c9b1).

If I revert the named commit, I can boot OK. Unfortunately there's no real output before the machine reboots, to help identify the problem.

I have a(n enabled) VMD in my Alderlake machine:

[    0.141952] [      T1] smpboot: CPU0: 12th Gen Intel(R) Core(TM) 

i7-1280P (family: 0x6, model: 0x9a, stepping: 0x3)

Quoted 6 lines
Quoted 1 line

If there's something else I can try, let me know.

Thanks for the report. It is unfortunate that there is no output to work with :(

Let me stare at the commit again..

Another person reported problem with this commit, and that was resolved with the diff below.

Does it fix your case too?

drivers/pci/controller/vmd.c +3 -1
index 9bbb0ff4cc15..b679c7f28f51 100644
@@ -280,10 +280,12 @@ static int vmd_msi_alloc(struct irq_domain *domain, unsigned int virq,
 static void vmd_msi_free(struct irq_domain *domain, unsigned int virq,
 			 unsigned int nr_irqs)
 {
+	struct irq_data *irq_data;
 	struct vmd_irq *vmdirq;
 
 	for (int i = 0; i < nr_irqs; ++i) {
-		vmdirq = irq_get_chip_data(virq + i);
+		irq_data = irq_domain_get_irq_data(domain, virq + i);
+		vmdirq = irq_data->chip_data;
 
 		synchronize_srcu(&vmdirq->irq->srcu);
 
#3 lore raw

I can boot with this patch- thanks!

-K

Quoted 44 lines · Nam Cao
Quoted 12 lines · Nam Cao
Quoted 10 lines · Kenneth Crudup

I'm running Linus' master (as of today, cca7a0aae8958c9b1).

If I revert the named commit, I can boot OK. Unfortunately there's no real output before the machine reboots, to help identify the problem.

I have a(n enabled) VMD in my Alderlake machine:

[    0.141952] [      T1] smpboot: CPU0: 12th Gen Intel(R) Core(TM)

i7-1280P (family: 0x6, model: 0x9a, stepping: 0x3)

Quoted 6 lines
Quoted 1 line

If there's something else I can try, let me know.

Thanks for the report. It is unfortunate that there is no output to work with :(

Let me stare at the commit again..

Another person reported problem with this commit, and that was resolved with the diff below.

Does it fix your case too?

drivers/pci/controller/vmd.c +3 -1
index 9bbb0ff4cc15..b679c7f28f51 100644
@@ -280,10 +280,12 @@ static int vmd_msi_alloc(struct irq_domain *domain, unsigned int virq,
  static void vmd_msi_free(struct irq_domain *domain, unsigned int virq,
  			 unsigned int nr_irqs)
  {
+	struct irq_data *irq_data;
  	struct vmd_irq *vmdirq;
 
  	for (int i = 0; i < nr_irqs; ++i) {
-		vmdirq = irq_get_chip_data(virq + i);
+		irq_data = irq_domain_get_irq_data(domain, virq + i);
+		vmdirq = irq_data->chip_data;
 
  		synchronize_srcu(&vmdirq->irq->srcu);
 
Signature · 1 line
Kenneth R. Crudup / Sr. SW Engineer, Scott County Consulting, Orange County CA
#4 lore raw
In reply to Kenneth Crudup
Quoted 10 lines · Kenneth Crudup

I'm running Linus' master (as of today, cca7a0aae8958c9b1).

If I revert the named commit, I can boot OK. Unfortunately there's no real output before the machine reboots, to help identify the problem.

I have a(n enabled) VMD in my Alderlake machine:

[    0.141952] [      T1] smpboot: CPU0: 12th Gen Intel(R) Core(TM) i7-1280P

(family: 0x6, model: 0x9a, stepping: 0x3)

#regzbot introduced: d7d8ab87e3e7 ("PCI: vmd: Switch to msi_create_parent_irq_domain()")

#5 lore raw
Quoted 7 lines · Bjorn Helgaas
Quoted 6 lines · Kenneth Crudup

I'm running Linus' master (as of today, cca7a0aae8958c9b1).

If I revert the named commit, I can boot OK. Unfortunately there's no real output before the machine reboots, to help identify the problem.

Quoted 6 lines
Quoted 4 lines

I have a(n enabled) VMD in my Alderlake machine:

[    0.141952] [      T1] smpboot: CPU0: 12th Gen Intel(R) Core(TM) i7-1280P

(family: 0x6, model: 0x9a, stepping: 0x3)

#regzbot introduced: d7d8ab87e3e7 ("PCI: vmd: Switch to msi_create_parent_irq_domain()")

#regzbot fix: d5c647b08ee0 ("PCI: vmd: Fix wrong kfree() in vmd_msi_free()")