Skip to contentSkip to conversation

PCI/VPD: Remove VPD quirk for QLogic 1077:2261

PATCH In review Arun Easi 2 participants 7 messages Dec 19, 2020 – Feb 25, 2021 PCI

lore
Share & tools
lore

Arun Easi

author
#0 lore raw

PATCH 1/1 PCI/VPD: Remove VPD quirk for QLogic 1077:2261

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://lore.kernel.org/linux-pci/alpine.LRH.2.21.9999.2012161641230.28924@irv1user01.caveonetworks.com/

---
 drivers/pci/vpd.c | 1 -
 1 file changed, 1 deletion(-)
drivers/pci/vpd.c +0 -1
index 7915d10..bd54907 100644
@@ -570,7 +570,6 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_LSI_LOGIC, 0x005d, quirk_blacklist_vpd);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_LSI_LOGIC, 0x005f, quirk_blacklist_vpd);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATTANSIC, PCI_ANY_ID,
 		quirk_blacklist_vpd);
-DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_QLOGIC, 0x2261, quirk_blacklist_vpd);
 /*
  * The Amazon Annapurna Labs 0x0031 device id is reused for other non Root Port
  * device types, so the quirk is registered for the PCI_CLASS_BRIDGE_PCI class.
Signature · 1 line
2.9.5
author
#1 lore raw

Hi Bjorn,

Quoted 6 lines · Arun Easi

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].
Quoted 7 lines

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_alpine.LRH.2.21.9999.2012161641230.28924-40irv1user01.caveonetworks.com_&d=DwIBAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=P-q_Qkt75qFy33SvdD2nAxAyN87eO1d-mFO-lqNOomw&m=Bw8qGbVsETqSibSD8JVMAxZh8BCn1cHuskKjbarfuT8&s=IMvYnIBgaHJkzF2-GgIrGymbRguV287NVLG1_KcP_po&e=

---

Wondering if there is something needed from my side. I could not find this in the v5.12 list.

Regards,
-Arun

#2 lore raw
Quoted 14 lines · Arun Easi

Hi Bjorn,

Quoted 10 lines · Arun Easi

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_alpine.LRH.2.21.9999.2012161641230.28924-40irv1user01.caveonetworks.com_&d=DwIBAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=P-q_Qkt75qFy33SvdD2nAxAyN87eO1d-mFO-lqNOomw&m=Bw8qGbVsETqSibSD8JVMAxZh8BCn1cHuskKjbarfuT8&s=IMvYnIBgaHJkzF2-GgIrGymbRguV287NVLG1_KcP_po&e=

Quoted 6 lines
Quoted 3 lines

---

Wondering if there is something needed from my side. I could not find this in the v5.12 list.

Sorry, I blew it on this one (and other VPD patches from Heiner). It's too late for v5.12, but I'll try again for v5.13.

Bjorn

author
#3 lore raw
Quoted 19 lines · Bjorn Helgaas
Quoted 18 lines · Arun Easi

Hi Bjorn,

Quoted 13 lines · Arun Easi

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_alpine.LRH.2.21.9999.2012161641230.28924-40irv1user01.caveonetworks.com_&d=DwIBAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=P-q_Qkt75qFy33SvdD2nAxAyN87eO1d-mFO-lqNOomw&m=Bw8qGbVsETqSibSD8JVMAxZh8BCn1cHuskKjbarfuT8&s=IMvYnIBgaHJkzF2-GgIrGymbRguV287NVLG1_KcP_po&e=

---

Quoted 6 lines
Quoted 2 lines

Wondering if there is something needed from my side. I could not find this in the v5.12 list.

Sorry, I blew it on this one (and other VPD patches from Heiner). It's too late for v5.12, but I'll try again for v5.13.

No worries. As these are mostly one-liners and straight forward changes, posting during RC is not an option for 5.12 inclusion?

Regards,
-Arun

#4 lore raw
Quoted 21 lines · Arun Easi
Quoted 20 lines · Bjorn Helgaas
Quoted 18 lines · Arun Easi
Quoted 6 lines · Arun Easi

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].
Quoted 7 lines

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_alpine.LRH.2.21.9999.2012161641230.28924-40irv1user01.caveonetworks.com_&d=DwIBAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=P-q_Qkt75qFy33SvdD2nAxAyN87eO1d-mFO-lqNOomw&m=Bw8qGbVsETqSibSD8JVMAxZh8BCn1cHuskKjbarfuT8&s=IMvYnIBgaHJkzF2-GgIrGymbRguV287NVLG1_KcP_po&e=

---

Wondering if there is something needed from my side. I could not find this in the v5.12 list.

Quoted 6 lines
Quoted 3 lines

Sorry, I blew it on this one (and other VPD patches from Heiner). It's too late for v5.12, but I'll try again for v5.13.

No worries. As these are mostly one-liners and straight forward changes, posting during RC is not an option for 5.12 inclusion?

Not unless it's critical or it fixes a bug we introduced during the merge window. This doesn't seem like either, but please clarify if it is.

From the commit log, it looks like this is just a cleanup with no benefit to the user. If that's the case, there's no reason to backport to stable either.

Bjorn

author
#5 lore raw
Quoted 29 lines · Bjorn Helgaas
Quoted 21 lines · Arun Easi
Quoted 20 lines · Bjorn Helgaas
Quoted 18 lines · Arun Easi
Quoted 6 lines · Arun Easi

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].
Quoted 7 lines

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_alpine.LRH.2.21.9999.2012161641230.28924-40irv1user01.caveonetworks.com_&d=DwIBAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=P-q_Qkt75qFy33SvdD2nAxAyN87eO1d-mFO-lqNOomw&m=Bw8qGbVsETqSibSD8JVMAxZh8BCn1cHuskKjbarfuT8&s=IMvYnIBgaHJkzF2-GgIrGymbRguV287NVLG1_KcP_po&e=

---

Wondering if there is something needed from my side. I could not find this in the v5.12 list.

Quoted 6 lines
Quoted 3 lines

Sorry, I blew it on this one (and other VPD patches from Heiner). It's too late for v5.12, but I'll try again for v5.13.

No worries. As these are mostly one-liners and straight forward changes, posting during RC is not an option for 5.12 inclusion?

Quoted 8 lines

Not unless it's critical or it fixes a bug we introduced during the merge window. This doesn't seem like either, but please clarify if it is.

From the commit log, it looks like this is just a cleanup with no benefit to the user. If that's the case, there's no reason to backport to stable either.

Actually, this fixes the issue that block users from seeing VPD data in "lspci -vvv" listing. My bad in not stating that well in the commit log. In retrospect, this was mentioned only in a prior e-mail (link[2] above -- sorry about our exchange servers mangling the URL and redirecting it via proofpoint).

I can re-send the patch to correctly reflect it as a fix. Something like:

Subject: "PCI/VPD: Fix blocking of VPD data in lspci for QLogic 1077:2261"

What do you think?

Regards,
-Arun

#6 lore raw
Quoted 44 lines · Arun Easi
Quoted 29 lines · Bjorn Helgaas
Quoted 21 lines · Arun Easi
Quoted 20 lines · Bjorn Helgaas
Quoted 18 lines · Arun Easi
Quoted 6 lines · Arun Easi

The VPD quirk was added by [0] to avoid a system NMI; this issue has been long fixed in the HBA firmware. In addition, PCI also has the logic to check the VPD size [1], so this quirk can be reverted now. More details in the thread:

    "VPD blacklist of Marvell QLogic 1077/2261" [2].
Quoted 7 lines

[0] 0d5370d1d852 ("PCI: Prevent VPD access for QLogic ISP2722")
[1] 104daa71b396 ("PCI: Determine actual VPD size on first access")
[2] https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_linux-2Dpci_alpine.LRH.2.21.9999.2012161641230.28924-40irv1user01.caveonetworks.com_&d=DwIBAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=P-q_Qkt75qFy33SvdD2nAxAyN87eO1d-mFO-lqNOomw&m=Bw8qGbVsETqSibSD8JVMAxZh8BCn1cHuskKjbarfuT8&s=IMvYnIBgaHJkzF2-GgIrGymbRguV287NVLG1_KcP_po&e=

---

Wondering if there is something needed from my side. I could not find this in the v5.12 list.

Quoted 6 lines
Quoted 3 lines

Sorry, I blew it on this one (and other VPD patches from Heiner). It's too late for v5.12, but I'll try again for v5.13.

No worries. As these are mostly one-liners and straight forward changes, posting during RC is not an option for 5.12 inclusion?

Quoted 7 lines

Not unless it's critical or it fixes a bug we introduced during the merge window. This doesn't seem like either, but please clarify if it is.

From the commit log, it looks like this is just a cleanup with no benefit to the user. If that's the case, there's no reason to backport to stable either.

Actually, this fixes the issue that block users from seeing VPD data in "lspci -vvv" listing. My bad in not stating that well in the commit log. In retrospect, this was mentioned only in a prior e-mail (link[2] above -- sorry about our exchange servers mangling the URL and redirecting it via proofpoint).

Quoted 3 lines

I can re-send the patch to correctly reflect it as a fix. Something like:

Subject: "PCI/VPD: Fix blocking of VPD data in lspci for QLogic 1077:2261"

Yes, that would be useful :) The commit log needs to explain why the patch should be merged.

Bjorn