Skip to contentSkip to conversation

Linux Kernel Regression: HAPS quirk breaks PCIe on i.MX6QP

DISCUSSION Bjorn Helgaas 5 participants 12 messages Jan 31, 2019 – Feb 5, 2019 1 message not in the archive PCI

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

Bjorn Helgaas

author maintainer
#1 lore raw

[+cc linux-pci, linux-kernel]

Quoted 26 lines · Lukas F. Hartmann

Hi Thinh,

I'm writing you because you're the author in this commit:
https://github.com/torvalds/linux/commit/03e6742584af8866ba014874669d0440bed3a623

This quirk workaround breaks the PCIe on i.MX6QP, at least on my test devices. The reason is because the Synposys PCIe IP in i.MX6QP has the same device ID as the HAPS USB3 controller you are targeting in the
commit: 0xabcd.

Definition:

https://github.com/torvalds/linux/blob/master/include/linux/pci_ids.h#L2364

Here is a bit of lspci output on my i.MX6QP machine (without the problem):

mntmn@reform:~/code/linux$ lspci -v
00:00.0 PCI bridge: Synopsys, Inc. Device abcd (rev 01) (prog-if 00
                                          ^^^^
[Normal decode])
        Flags: bus master, fast devsel, latency 0, IRQ 307
        Memory at 01000000 (32-bit, non-prefetchable) [size=1M]
        Bus: primary=00, secondary=01, subordinate=ff, sec-latency=0
...
Quoted 6 lines

The failure mode is that the PCIe controller shows up as a USB controller and my ath9k wireless PCIe card cannot be assigned the proper resources anymore (-ENOMEM even for very small BARs).

Reverting this commit fixes the problem for me. I suggest to make the check more specific to the platform/chipset you are targeting.

So Synopsys apparently re-used Device ID 0xabcd? That's a pretty bad problem.

It looks like we merged 03e6742584af ("PCI: Override Synopsys USB 3.x HAPS device class") for v5.0, and v5.0-final hasn't been released yet, so if we don't hear from Thinh with a resolution, we can still revert it.

I set myself a reminder to revert this on Feb 11, but hopefully we'll have a better resolution before then.

Bjorn

#2 lore raw

Hi Lukas,

Quoted 41 lines · Bjorn Helgaas

[+cc linux-pci, linux-kernel]

Quoted 25 lines · Lukas F. Hartmann

Hi Thinh,

I'm writing you because you're the author in this commit:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_commit_03e6742584af8866ba014874669d0440bed3a623&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=oJoBWRE8_LGAYbX2alh6QnjYZTTmzcgLw4MtOMToNyo&e=

This quirk workaround breaks the PCIe on i.MX6QP, at least on my test devices. The reason is because the Synposys PCIe IP in i.MX6QP has the same device ID as the HAPS USB3 controller you are targeting in the
commit: 0xabcd.

Definition:

https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_blob_master_include_linux_pci-5Fids.h-23L2364&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=-FVCwe81XNjJsYHYk1w-kdAzSOLunTGeNc73azO2QYw&e=

Here is a bit of lspci output on my i.MX6QP machine (without the problem):

mntmn@reform:~/code/linux$ lspci -v
00:00.0 PCI bridge: Synopsys, Inc. Device abcd (rev 01) (prog-if 00
                                          ^^^^
[Normal decode])
        Flags: bus master, fast devsel, latency 0, IRQ 307
        Memory at 01000000 (32-bit, non-prefetchable) [size=1M]
        Bus: primary=00, secondary=01, subordinate=ff, sec-latency=0
...
Quoted 6 lines

The failure mode is that the PCIe controller shows up as a USB controller and my ath9k wireless PCIe card cannot be assigned the proper resources anymore (-ENOMEM even for very small BARs).

Reverting this commit fixes the problem for me. I suggest to make the check more specific to the platform/chipset you are targeting.

So Synopsys apparently re-used Device ID 0xabcd? That's a pretty bad problem.

It looks like we merged 03e6742584af ("PCI: Override Synopsys USB 3.x HAPS device class") for v5.0, and v5.0-final hasn't been released yet, so if we don't hear from Thinh with a resolution, we can still revert it.

Quoted 5 lines

I set myself a reminder to revert this on Feb 11, but hopefully we'll have a better resolution before then.

Bjorn

This is really odd that the PID for the PCIe controller in i.MX6QP is the same as PID Synopsys use for USB controller. We use a different set of PIDs for PCIe controllers and track VID and PID in our releases.

Is the Root Complex (00:00.0) part of the SoC? Or is this a Synopsys prototype connected to your board? If it is the latter, then the FPGA configuration needs to be updated to other PID.

We're investigating the workaround in case this is on a taped-out SoC.

Thanks,
Thinh

#3 lore raw

Hi Lukas,

Quoted 51 lines · Thinh Nguyen
Quoted 41 lines · Bjorn Helgaas

[+cc linux-pci, linux-kernel]

Quoted 25 lines · Lukas F. Hartmann

Hi Thinh,

I'm writing you because you're the author in this commit:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_commit_03e6742584af8866ba014874669d0440bed3a623&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=oJoBWRE8_LGAYbX2alh6QnjYZTTmzcgLw4MtOMToNyo&e=

This quirk workaround breaks the PCIe on i.MX6QP, at least on my test devices. The reason is because the Synposys PCIe IP in i.MX6QP has the same device ID as the HAPS USB3 controller you are targeting in the
commit: 0xabcd.

Definition:

https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_blob_master_include_linux_pci-5Fids.h-23L2364&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=-FVCwe81XNjJsYHYk1w-kdAzSOLunTGeNc73azO2QYw&e=

Here is a bit of lspci output on my i.MX6QP machine (without the problem):

mntmn@reform:~/code/linux$ lspci -v
00:00.0 PCI bridge: Synopsys, Inc. Device abcd (rev 01) (prog-if 00
                                          ^^^^
[Normal decode])
        Flags: bus master, fast devsel, latency 0, IRQ 307
        Memory at 01000000 (32-bit, non-prefetchable) [size=1M]
        Bus: primary=00, secondary=01, subordinate=ff, sec-latency=0
...
Quoted 6 lines

The failure mode is that the PCIe controller shows up as a USB controller and my ath9k wireless PCIe card cannot be assigned the proper resources anymore (-ENOMEM even for very small BARs).

Reverting this commit fixes the problem for me. I suggest to make the check more specific to the platform/chipset you are targeting.

So Synopsys apparently re-used Device ID 0xabcd? That's a pretty bad problem.

It looks like we merged 03e6742584af ("PCI: Override Synopsys USB 3.x HAPS device class") for v5.0, and v5.0-final hasn't been released yet, so if we don't hear from Thinh with a resolution, we can still revert it.

Quoted 5 lines

I set myself a reminder to revert this on Feb 11, but hopefully we'll have a better resolution before then.

Bjorn

This is really odd that the PID for the PCIe controller in i.MX6QP is the same as PID Synopsys use for USB controller. We use a different set of PIDs for PCIe controllers and track VID and PID in our releases.

Quoted 5 lines

Is the Root Complex (00:00.0) part of the SoC? Or is this a Synopsys prototype connected to your board? If it is the latter, then the FPGA configuration needs to be updated to other PID.

We're investigating the workaround in case this is on a taped-out SoC.

Can you try to see if this will help?

drivers/pci/quirks.c +3 -0
index b0a413f3f7ca..64623fbdd1e5 100644
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Thanks,
Thinh

#4 lore raw

Hi Thinh,

this is in a taped out i.MX6QP 1.2GHz SoC, I also confirmed it with the 1.0GHz version, it has the same 0xabcd device id integrated.

I will try your patch ASAP.

Cheers
Lukas

Quoted 83 lines · Thinh Nguyen

Hi Lukas,

Quoted 51 lines · Thinh Nguyen
Quoted 41 lines · Bjorn Helgaas

[+cc linux-pci, linux-kernel]

Quoted 26 lines · Lukas F. Hartmann

Hi Thinh,

I'm writing you because you're the author in this commit:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_commit_03e6742584af8866ba014874669d0440bed3a623&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=oJoBWRE8_LGAYbX2alh6QnjYZTTmzcgLw4MtOMToNyo&e=

This quirk workaround breaks the PCIe on i.MX6QP, at least on my test devices. The reason is because the Synposys PCIe IP in i.MX6QP has the same device ID as the HAPS USB3 controller you are targeting in the
commit: 0xabcd.

Definition:

https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_blob_master_include_linux_pci-5Fids.h-23L2364&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=-FVCwe81XNjJsYHYk1w-kdAzSOLunTGeNc73azO2QYw&e=

Here is a bit of lspci output on my i.MX6QP machine (without the problem):

mntmn@reform:~/code/linux$ lspci -v
00:00.0 PCI bridge: Synopsys, Inc. Device abcd (rev 01) (prog-if 00
                                         ^^^^
[Normal decode])
       Flags: bus master, fast devsel, latency 0, IRQ 307
       Memory at 01000000 (32-bit, non-prefetchable) [size=1M]
       Bus: primary=00, secondary=01, subordinate=ff, sec-latency=0
...
Quoted 6 lines

The failure mode is that the PCIe controller shows up as a USB controller and my ath9k wireless PCIe card cannot be assigned the proper resources anymore (-ENOMEM even for very small BARs).

Reverting this commit fixes the problem for me. I suggest to make the check more specific to the platform/chipset you are targeting.

So Synopsys apparently re-used Device ID 0xabcd? That's a pretty bad problem.

It looks like we merged 03e6742584af ("PCI: Override Synopsys USB 3.x HAPS device class") for v5.0, and v5.0-final hasn't been released yet, so if we don't hear from Thinh with a resolution, we can still revert it.

Quoted 5 lines

I set myself a reminder to revert this on Feb 11, but hopefully we'll have a better resolution before then.

Bjorn

This is really odd that the PID for the PCIe controller in i.MX6QP is the same as PID Synopsys use for USB controller. We use a different set of PIDs for PCIe controllers and track VID and PID in our releases.

Quoted 5 lines

Is the Root Complex (00:00.0) part of the SoC? Or is this a Synopsys prototype connected to your board? If it is the latter, then the FPGA configuration needs to be updated to other PID.

We're investigating the workaround in case this is on a taped-out SoC.

Can you try to see if this will help?

drivers/pci/quirks.c +3 -0
index b0a413f3f7ca..64623fbdd1e5 100644
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
       u32 class = pdev->class;
 
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
       switch (pdev->device) {
       case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
       case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Thanks,
Thinh

#5 lore raw

Hi Lukas,

Quoted 3 lines · Lukas Hartmann

Hi Thinh,

this is in a taped out i.MX6QP 1.2GHz SoC, I also confirmed it with the 1.0GHz version, it has the same 0xabcd device id integrated.

I see.

Quoted 1 line

I will try your patch ASAP.

Thank you.
Thinh

Quoted 61 lines
Quoted 61 lines · Thinh Nguyen

Hi Lukas,

Quoted 55 lines · Thinh Nguyen
Quoted 41 lines · Bjorn Helgaas

[+cc linux-pci, linux-kernel]

Quoted 26 lines · Lukas F. Hartmann

Hi Thinh,

I'm writing you because you're the author in this commit:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_commit_03e6742584af8866ba014874669d0440bed3a623&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=oJoBWRE8_LGAYbX2alh6QnjYZTTmzcgLw4MtOMToNyo&e=

This quirk workaround breaks the PCIe on i.MX6QP, at least on my test devices. The reason is because the Synposys PCIe IP in i.MX6QP has the same device ID as the HAPS USB3 controller you are targeting in the
commit: 0xabcd.

Definition:

https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_torvalds_linux_blob_master_include_linux_pci-5Fids.h-23L2364&d=DwIBaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=VKipRJmm95P2RbIxyKKYrcUCOGNlQtjlV-5zhrVhIik&m=2sOrowYXlsC3rl0LfHQhIZzImag0jFZXGvR6NIQDsh8&s=-FVCwe81XNjJsYHYk1w-kdAzSOLunTGeNc73azO2QYw&e=

Here is a bit of lspci output on my i.MX6QP machine (without the problem):

mntmn@reform:~/code/linux$ lspci -v
00:00.0 PCI bridge: Synopsys, Inc. Device abcd (rev 01) (prog-if 00
                                         ^^^^
[Normal decode])
       Flags: bus master, fast devsel, latency 0, IRQ 307
       Memory at 01000000 (32-bit, non-prefetchable) [size=1M]
       Bus: primary=00, secondary=01, subordinate=ff, sec-latency=0
...
Quoted 6 lines

The failure mode is that the PCIe controller shows up as a USB controller and my ath9k wireless PCIe card cannot be assigned the proper resources anymore (-ENOMEM even for very small BARs).

Reverting this commit fixes the problem for me. I suggest to make the check more specific to the platform/chipset you are targeting.

So Synopsys apparently re-used Device ID 0xabcd? That's a pretty bad problem.

It looks like we merged 03e6742584af ("PCI: Override Synopsys USB 3.x HAPS device class") for v5.0, and v5.0-final hasn't been released yet, so if we don't hear from Thinh with a resolution, we can still revert it.

Quoted 5 lines

I set myself a reminder to revert this on Feb 11, but hopefully we'll have a better resolution before then.

Bjorn

This is really odd that the PID for the PCIe controller in i.MX6QP is the same as PID Synopsys use for USB controller. We use a different set of PIDs for PCIe controllers and track VID and PID in our releases.

Is the Root Complex (00:00.0) part of the SoC? Or is this a Synopsys prototype connected to your board? If it is the latter, then the FPGA configuration needs to be updated to other PID.

#6 lore raw
In reply to Thinh Nguyen

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 12 lines · Thinh Nguyen
drivers/pci/quirks.c +3 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Thanks,

Lukas

#7 lore raw

Hi Lukas,

Quoted 7 lines · Lukas Wunner
Quoted 6 lines · Thinh Nguyen
drivers/pci/quirks.c +0 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
Quoted 7 lines

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 6 lines
drivers/pci/quirks.c +3 -0
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Sure. That's a better option. Can you test this with your setup?

drivers/pci/quirks.c +2 -2
index b0a413f3f7ca..f46e7de9e15d 100644
@@ -639,8 +639,8 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
                break;
        }
 }
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
-                        quirk_synopsys_haps);
+DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
+               PCI_CLASS_SERIAL_USB_XHCI, 0, quirk_synopsys_haps);
 
 /*
  * Let's make the southbridge information explicit instead of having to

Thanks,
Thinh

#8 lore raw
Quoted 25 lines · Thinh Nguyen
Quoted 7 lines · Lukas Wunner
Quoted 6 lines · Thinh Nguyen
drivers/pci/quirks.c +0 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
Quoted 7 lines

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 6 lines
drivers/pci/quirks.c +3 -0
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Sure. That's a better option. Can you test this with your setup?

drivers/pci/quirks.c +0 -0
index b0a413f3f7ca..f46e7de9e15d 100644

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 646

Quoted 10 lines
drivers/pci/quirks.c +2 -2
@@ -639,8 +639,8 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
                break;
        }
 }
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
-                        quirk_synopsys_haps);
+DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
+               PCI_CLASS_SERIAL_USB_XHCI, 0, quirk_synopsys_haps);
 
 /*
  * Let's make the southbridge information explicit instead of having to

Can we get a formal patch, including details about the issue (I assume Synopsys released two different parts with Device ID 0xabcd) and a signed-off-by?

I'd like to get this into for-linus as soon as possible for v5.0.

Bjorn

#9 lore raw

Hi Bjorn,

Quoted 39 lines · Bjorn Helgaas
Quoted 24 lines · Thinh Nguyen
Quoted 7 lines · Lukas Wunner
Quoted 6 lines · Thinh Nguyen
drivers/pci/quirks.c +0 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
Quoted 7 lines

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 6 lines
drivers/pci/quirks.c +3 -0
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Sure. That's a better option. Can you test this with your setup?

drivers/pci/quirks.c +0 -0
index b0a413f3f7ca..f46e7de9e15d 100644

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 646

Quoted 10 lines
drivers/pci/quirks.c +2 -2
@@ -639,8 +639,8 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
                break;
        }
 }
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
-                        quirk_synopsys_haps);
+DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
+               PCI_CLASS_SERIAL_USB_XHCI, 0, quirk_synopsys_haps);
 
 /*
  * Let's make the southbridge information explicit instead of having to

Can we get a formal patch, including details about the issue (I assume Synopsys released two different parts with Device ID 0xabcd) and a signed-off-by?

Quoted 2 lines

I'd like to get this into for-linus as soon as possible for v5.0.

I already submitted a patch for this. Please review patch subject "[PATCH] PCI: Check for USB xHCI class for HAPS platform".

Thanks,
Thinh

#10 lore raw
Quoted 42 lines · Thinh Nguyen

Hi Bjorn,

Quoted 39 lines · Bjorn Helgaas
Quoted 24 lines · Thinh Nguyen
Quoted 7 lines · Lukas Wunner
Quoted 6 lines · Thinh Nguyen
drivers/pci/quirks.c +0 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
Quoted 7 lines

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 6 lines
drivers/pci/quirks.c +3 -0
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Sure. That's a better option. Can you test this with your setup?

drivers/pci/quirks.c +0 -0
index b0a413f3f7ca..f46e7de9e15d 100644

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 646

Quoted 10 lines
drivers/pci/quirks.c +2 -2
@@ -639,8 +639,8 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
                break;
        }
 }
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
-                        quirk_synopsys_haps);
+DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
+               PCI_CLASS_SERIAL_USB_XHCI, 0, quirk_synopsys_haps);
 
 /*
  * Let's make the southbridge information explicit instead of having to

Can we get a formal patch, including details about the issue (I assume Synopsys released two different parts with Device ID 0xabcd) and a signed-off-by?

Quoted 4 lines
Quoted 1 line

I'd like to get this into for-linus as soon as possible for v5.0.

I already submitted a patch for this. Please review patch subject "[PATCH] PCI: Check for USB xHCI class for HAPS platform".

OK, maybe I'm missing something. I don't see it on the linux-pci list
[1] or the PCI patchwork [2] yet. Maybe it's still making its way through the mailing list servers. If you have a URL to the archive or patchwork, let me know.

Bjorn

[1] https://lore.kernel.org/linux-pci/
[2] https://patchwork.ozlabs.org/project/linux-pci/list

#11 lore raw
Quoted 51 lines · Bjorn Helgaas
Quoted 42 lines · Thinh Nguyen

Hi Bjorn,

Quoted 39 lines · Bjorn Helgaas
Quoted 24 lines · Thinh Nguyen
Quoted 7 lines · Lukas Wunner
Quoted 6 lines · Thinh Nguyen
drivers/pci/quirks.c +0 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
Quoted 7 lines

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 6 lines
drivers/pci/quirks.c +3 -0
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Sure. That's a better option. Can you test this with your setup?

drivers/pci/quirks.c +0 -0
index b0a413f3f7ca..f46e7de9e15d 100644

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 646

Quoted 10 lines
drivers/pci/quirks.c +2 -2
@@ -639,8 +639,8 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
                break;
        }
 }
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
-                        quirk_synopsys_haps);
+DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
+               PCI_CLASS_SERIAL_USB_XHCI, 0, quirk_synopsys_haps);
 
 /*
  * Let's make the southbridge information explicit instead of having to

Can we get a formal patch, including details about the issue (I assume Synopsys released two different parts with Device ID 0xabcd) and a signed-off-by?

Quoted 3 lines
Quoted 1 line

I'd like to get this into for-linus as soon as possible for v5.0.

I already submitted a patch for this. Please review patch subject "[PATCH] PCI: Check for USB xHCI class for HAPS platform".

OK, maybe I'm missing something. I don't see it on the linux-pci list
[1] or the PCI patchwork [2] yet. Maybe it's still making its way through the mailing list servers. If you have a URL to the archive or patchwork, let me know.

Quoted 5 lines

Bjorn

[1] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_&d=DwIBAg&c=DPL6_X_6JkXFx7AXWqB0tg&r=u9FYoxKtyhjrGFcyixFYqTjw1ZX0VsG2d8FCmzkTY-w&m=gf8tJEWOt92UvWl-0yPhkFlhOXuG0fn-pG0zPdqsHv8&s=vQgsCLqN63Bd-d7ZMY5HPVmBa4Htnz64oO96cG-6pNA&e=
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__patchwork.ozlabs.org_project_linux-2Dpci_list&d=DwIBAg&c=DPL6_X_6JkXFx7AXWqB0tg&r=u9FYoxKtyhjrGFcyixFYqTjw1ZX0VsG2d8FCmzkTY-w&m=gf8tJEWOt92UvWl-0yPhkFlhOXuG0fn-pG0zPdqsHv8&s=Tbm0SOa0CCxMxzGcwT6YJrYYzVEU6x7sR8JB9l82VGg&e=

That's odd. I sent it yesterday. I'll resend it then.

Thanks,
Thinh

#12 lore raw
Quoted 52 lines · Thinh Nguyen
Quoted 51 lines · Bjorn Helgaas
Quoted 42 lines · Thinh Nguyen

Hi Bjorn,

Quoted 39 lines · Bjorn Helgaas
Quoted 24 lines · Thinh Nguyen
Quoted 7 lines · Lukas Wunner
Quoted 6 lines · Thinh Nguyen
drivers/pci/quirks.c +0 -0
@@ -629,6 +629,9 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
 {
        u32 class = pdev->class;
 
Quoted 7 lines

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 637

Quoted 6 lines
drivers/pci/quirks.c +3 -0
+       if (class != PCI_CLASS_SERIAL_USB_XHCI)
+               return;
+
        switch (pdev->device) {
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3:
        case PCI_DEVICE_ID_SYNOPSYS_HAPSUSB3_AXI:

Please use DECLARE_PCI_FIXUP_CLASS_HEADER() instead.

Sure. That's a better option. Can you test this with your setup?

drivers/pci/quirks.c +0 -0
index b0a413f3f7ca..f46e7de9e15d 100644

drivers/pci/quirks.c static void quirk_synopsys_haps(struct pci_dev *pdev) · line 646

Quoted 10 lines
drivers/pci/quirks.c +2 -2
@@ -639,8 +639,8 @@ static void quirk_synopsys_haps(struct pci_dev *pdev)
                break;
        }
 }
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
-                        quirk_synopsys_haps);
+DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
+               PCI_CLASS_SERIAL_USB_XHCI, 0, quirk_synopsys_haps);
 
 /*
  * Let's make the southbridge information explicit instead of having to

Can we get a formal patch, including details about the issue (I assume Synopsys released two different parts with Device ID 0xabcd) and a signed-off-by?

Quoted 3 lines
Quoted 1 line

I'd like to get this into for-linus as soon as possible for v5.0.

I already submitted a patch for this. Please review patch subject "[PATCH] PCI: Check for USB xHCI class for HAPS platform".

OK, maybe I'm missing something. I don't see it on the linux-pci list
[1] or the PCI patchwork [2] yet. Maybe it's still making its way through the mailing list servers. If you have a URL to the archive or patchwork, let me know.

Quoted 4 lines
Quoted 4 lines

Bjorn

[1] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_&d=DwIBAg&c=DPL6_X_6JkXFx7AXWqB0tg&r=u9FYoxKtyhjrGFcyixFYqTjw1ZX0VsG2d8FCmzkTY-w&m=gf8tJEWOt92UvWl-0yPhkFlhOXuG0fn-pG0zPdqsHv8&s=vQgsCLqN63Bd-d7ZMY5HPVmBa4Htnz64oO96cG-6pNA&e=
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__patchwork.ozlabs.org_project_linux-2Dpci_list&d=DwIBAg&c=DPL6_X_6JkXFx7AXWqB0tg&r=u9FYoxKtyhjrGFcyixFYqTjw1ZX0VsG2d8FCmzkTY-w&m=gf8tJEWOt92UvWl-0yPhkFlhOXuG0fn-pG0zPdqsHv8&s=Tbm0SOa0CCxMxzGcwT6YJrYYzVEU6x7sR8JB9l82VGg&e=

After I removed some names in the CC list, it looks like it appears on PCI patchwork now.
https://patchwork.ozlabs.org/patch/1037211/

Thanks,
Thinh