Skip to contentSkip to conversation

PCI: Disable MSI for ULi M1575 EHCI controller

PATCH Applied Imre Kaloz 4 participants 4 messages Sep 30 – Oct 1

Subsystems (3)
lore
Share & tools
lore

Imre Kaloz

author
#0 lore raw

PATCH 1/1 PCI: Disable MSI for ULi M1575 EHCI controller

With MSI, the ULi M1575 EHCI controller (10b9:5239) intermittently stops executing its asynchronous schedule. Control transfers time out ("device not accepting address", "device descriptor read/64"), the port is handed to the companion OHCI controller and the device ends up at full speed. On a configured hub the stall shows up as hub descriptor and status read failures. Only re-initializing the controller recovers it.

On a Sun Ultra 45, 65 of 800 controller initializations stalled with MSI and none of 200 with INTx. Each initialization is an unbind and bind of ehci-pci on the EHCI function, with MSI or INTx selected through its msi_bus attribute; with this quirk, 0 of 100 cycles of the same loop stalled. The stall reproduces on a kernel that already contains commit 0b60557230ad ("usb: ehci: Prevent missed ehci interrupts with edge-triggered MSI").

The OHCI functions (10b9:5237) keep MSI. They showed no stall, and the hub enumerated through them on MSI during the EHCI failures, so only the EHCI function is quirked.

The illumos EHCI driver also disables MSI on this chip, in ehci_register_intrs_and_init_mutex() in usr/src/uts/common/io/usb/hcd/ehci/ehci_util.c.

Until USB HCDs started trying MSI first in v5.9, the controller always used INTx. Disable MSI so it uses INTx again on this chipset.

Fixes 306c54d0edb6 ("usb: hcd: Try MSI interrupts on PCI devices")

---

Notes:
lspci -nnv of the EHCI function on the Ultra 45, taken with this quirk active, so MSI shows Enable-:

0000:05:1c.3 USB controller [0c03]: ULi Electronics Inc. USB 2.0 Controller [10b9:5239] (rev 01) (prog-if 20 [EHCI])
Subsystem: ULi Electronics Inc. Device [10b9:5238]
Device tree node: /sys/firmware/devicetree/base/pci@1e,600000/pci@0/pci@1/pci@0/usb@1c,3
Flags: bus master, 66MHz, medium devsel, latency 64, IRQ 00000006
Memory at 00206000 (32-bit, non-prefetchable) [size=8K]
Capabilities: [50] Power Management version 2
Capabilities: [58] Debug port: BAR=1 offset=0090
Capabilities: [78] MSI: Enable- Count=1/1 Maskable- 64bit-
Kernel driver in use: ehci-pci
Kernel modules: ehci_pci

 drivers/pci/quirks.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/pci/quirks.c +2 -0
index de9bbccda21f..9fb530a08fda 100644
@@ -1941,6 +1941,8 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x4388, quirk_no_msi);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x4389, quirk_no_msi);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x438a, quirk_no_msi);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x438b, quirk_no_msi);
+/* ULi M1575 EHCI (10b9:5239): MSI stalls the async schedule */
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AL, 0x5239, quirk_no_msi);
 
 static void quirk_pcie_mch(struct pci_dev *pdev)
 {

base-commit: fe2ec83746e501645709761605c2464a44fd2929

Signature · 1 line
2.47.3
#1 lore raw
Sashiko has reviewed this patch and found no issues. It looks great!
Quoted 23 lines

With MSI, the ULi M1575 EHCI controller (10b9:5239) intermittently stops executing its asynchronous schedule. Control transfers time out ("device not accepting address", "device descriptor read/64"), the port is handed to the companion OHCI controller and the device ends up at full speed. On a configured hub the stall shows up as hub descriptor and status read failures. Only re-initializing the controller recovers it.

On a Sun Ultra 45, 65 of 800 controller initializations stalled with MSI and none of 200 with INTx. Each initialization is an unbind and bind of ehci-pci on the EHCI function, with MSI or INTx selected through its msi_bus attribute; with this quirk, 0 of 100 cycles of the same loop stalled. The stall reproduces on a kernel that already contains commit 0b60557230ad ("usb: ehci: Prevent missed ehci interrupts with edge-triggered MSI").

The OHCI functions (10b9:5237) keep MSI. They showed no stall, and the hub enumerated through them on MSI during the EHCI failures, so only the EHCI function is quirked.

The illumos EHCI driver also disables MSI on this chip, in ehci_register_intrs_and_init_mutex() in usr/src/uts/common/io/usb/hcd/ehci/ehci_util.c.

Quoted 6 lines

Until USB HCDs started trying MSI first in v5.9, the controller always used INTx. Disable MSI so it uses INTx again on this chipset.

Fixes 306c54d0edb6 ("usb: hcd: Try MSI interrupts on PCI devices")

Sashiko has reviewed this patch and found no issues. It looks great!

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
Reviewed-by on v1 1/1
#2 lore raw
In reply to Imre Kaloz
Quoted 19 lines · Imre Kaloz

With MSI, the ULi M1575 EHCI controller (10b9:5239) intermittently stops executing its asynchronous schedule. Control transfers time out ("device not accepting address", "device descriptor read/64"), the port is handed to the companion OHCI controller and the device ends up at full speed. On a configured hub the stall shows up as hub descriptor and status read failures. Only re-initializing the controller recovers it.

On a Sun Ultra 45, 65 of 800 controller initializations stalled with MSI and none of 200 with INTx. Each initialization is an unbind and bind of ehci-pci on the EHCI function, with MSI or INTx selected through its msi_bus attribute; with this quirk, 0 of 100 cycles of the same loop stalled. The stall reproduces on a kernel that already contains commit 0b60557230ad ("usb: ehci: Prevent missed ehci interrupts with edge-triggered MSI").

The OHCI functions (10b9:5237) keep MSI. They showed no stall, and the hub enumerated through them on MSI during the EHCI failures, so only the EHCI function is quirked.

Quoted 6 lines

The illumos EHCI driver also disables MSI on this chip, in ehci_register_intrs_and_init_mutex() in usr/src/uts/common/io/usb/hcd/ehci/ehci_util.c.

Until USB HCDs started trying MSI first in v5.9, the controller always used INTx. Disable MSI so it uses INTx again on this chipset.

Since it's tested on real HW, the quirk makes sense. Thanks for report and the fix!

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#3 lore raw
In reply to Imre Kaloz
Quoted 23 lines · Imre Kaloz

With MSI, the ULi M1575 EHCI controller (10b9:5239) intermittently stops executing its asynchronous schedule. Control transfers time out ("device not accepting address", "device descriptor read/64"), the port is handed to the companion OHCI controller and the device ends up at full speed. On a configured hub the stall shows up as hub descriptor and status read failures. Only re-initializing the controller recovers it.

On a Sun Ultra 45, 65 of 800 controller initializations stalled with MSI and none of 200 with INTx. Each initialization is an unbind and bind of ehci-pci on the EHCI function, with MSI or INTx selected through its msi_bus attribute; with this quirk, 0 of 100 cycles of the same loop stalled. The stall reproduces on a kernel that already contains commit 0b60557230ad ("usb: ehci: Prevent missed ehci interrupts with edge-triggered MSI").

The OHCI functions (10b9:5237) keep MSI. They showed no stall, and the hub enumerated through them on MSI during the EHCI failures, so only the EHCI function is quirked.

The illumos EHCI driver also disables MSI on this chip, in ehci_register_intrs_and_init_mutex() in usr/src/uts/common/io/usb/hcd/ehci/ehci_util.c.

Quoted 6 lines

Until USB HCDs started trying MSI first in v5.9, the controller always used INTx. Disable MSI so it uses INTx again on this chipset.

Fixes 306c54d0edb6 ("usb: hcd: Try MSI interrupts on PCI devices")

Applied to pci/misc for v7.4, thanks!

Quoted 38 lines

---

Notes:
lspci -nnv of the EHCI function on the Ultra 45, taken with this quirk active, so MSI shows Enable-:

0000:05:1c.3 USB controller [0c03]: ULi Electronics Inc. USB 2.0 Controller [10b9:5239] (rev 01) (prog-if 20 [EHCI])
Subsystem: ULi Electronics Inc. Device [10b9:5238]
Device tree node: /sys/firmware/devicetree/base/pci@1e,600000/pci@0/pci@1/pci@0/usb@1c,3
Flags: bus master, 66MHz, medium devsel, latency 64, IRQ 00000006
Memory at 00206000 (32-bit, non-prefetchable) [size=8K]
Capabilities: [50] Power Management version 2
Capabilities: [58] Debug port: BAR=1 offset=0090
Capabilities: [78] MSI: Enable- Count=1/1 Maskable- 64bit-
Kernel driver in use: ehci-pci
Kernel modules: ehci_pci

 drivers/pci/quirks.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/pci/quirks.c +2 -0
index de9bbccda21f..9fb530a08fda 100644
@@ -1941,6 +1941,8 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x4388, quirk_no_msi);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x4389, quirk_no_msi);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x438a, quirk_no_msi);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x438b, quirk_no_msi);
+/* ULi M1575 EHCI (10b9:5239): MSI stalls the async schedule */
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AL, 0x5239, quirk_no_msi);
 
 static void quirk_pcie_mch(struct pci_dev *pdev)
 {

base-commit: fe2ec83746e501645709761605c2464a44fd2929

Signature · 1 line
2.47.3