Skip to contentSkip to conversation

PCI: xgene: Fix IB window setup

PATCH Applied Rob Herring 6 participants 14 messages Nov 29, 2021 – Mar 7, 2022

Subsystems (5)
Stable branch PCI PCI native host bridge and endpoint drivers Regressions +1 more subsystems
lore
Share & tools
lore

Rob Herring

author reviewer
#0 lore raw

PATCH 1/1 PCI: xgene: Fix IB window setup

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

Reported-by Stéphane Graber <[email protected]>
Fixes 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup")
Link https://lore.kernel.org/all/CA+enf=v9rY_xnZML01oEgKLmvY1NGBUUhnSJaETmXtDtXfaczA@mail.gmail.com/ open here
Cc [email protected] # v5.5+
Signed-off-by Rob Herring <[email protected]>
---
 drivers/pci/controller/pci-xgene.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/controller/pci-xgene.c +1 -1
index 56d0d50338c8..d83dbd977418 100644
@@ -465,7 +465,7 @@ static int xgene_pcie_select_ib_reg(u8 *ib_reg_mask, u64 size)
 		return 1;
 	}
 
-	if ((size > SZ_1K) && (size < SZ_1T) && !(*ib_reg_mask & (1 << 0))) {
+	if ((size > SZ_1K) && (size < SZ_4G) && !(*ib_reg_mask & (1 << 0))) {
 		*ib_reg_mask |= (1 << 0);
 		return 0;
 	}
Signature · 7 lines
2.32.0


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Tested-by on v1 1/1
#1 lore raw
Quoted 10 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

Quoted 5 lines
Reported-by Stéphane Graber <[email protected]>
Fixes 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup")
Link https://lore.kernel.org/all/CA+enf=v9rY_xnZML01oEgKLmvY1NGBUUhnSJaETmXtDtXfaczA@mail.gmail.com/ open here
Cc [email protected] # v5.5+
Signed-off-by Rob Herring <[email protected]>

I've been running with this exact change on top of the latest 5.12 stable release for a few days now, so can confirm that on my hardware it's behaving perfectly (on 4 different servers).

Tested-by Stéphane Graber <[email protected]>
Quoted 20 lines
---
 drivers/pci/controller/pci-xgene.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/controller/pci-xgene.c +1 -2
index 56d0d50338c8..d83dbd977418 100644
@@ -465,7 +465,7 @@ static int xgene_pcie_select_ib_reg(u8 *ib_reg_mask, u64 size)
                return 1;
        }
 
-       if ((size > SZ_1K) && (size < SZ_1T) && !(*ib_reg_mask & (1 << 0))) {
+       if ((size > SZ_1K) && (size < SZ_4G) && !(*ib_reg_mask & (1 << 0))) {
                *ib_reg_mask |= (1 << 0);
                return 0;
        }
--

2.32.0

Reviewed-by on v1 1/1
#2 lore raw
In reply to Rob Herring

Hi,

Quoted 8 lines

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

A small nitpick: it would be "X-Gene" in the above as per Applied Micro's (or rather MACOM Technology Solutions these days, I suppose) product line naming.

drivers/pci/controller/pci-xgene.c static int xgene_pcie_select_ib_reg(u8 *ib_reg_mask, u64 size) · line 471

Quoted 9 lines
drivers/pci/controller/pci-xgene.c +1 -1
@@ -465,7 +465,7 @@ static int xgene_pcie_select_ib_reg(u8 *ib_reg_mask, u64 size)
 		return 1;
 	}
 
-	if ((size > SZ_1K) && (size < SZ_1T) && !(*ib_reg_mask & (1 << 0))) {
+	if ((size > SZ_1K) && (size < SZ_4G) && !(*ib_reg_mask & (1 << 0))) {
 		*ib_reg_mask |= (1 << 0);
 		return 0;
 	}

Thank you!

Reviewed-by Krzysztof Wilczyński <[email protected]>

Also, thank you Stéphane for testing! Much appreciated!

Krzysztof

Applied
#3 lore raw
In reply to Rob Herring
Quoted 10 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

[...]

Applied to pci/xgene, thanks!

[1/1] PCI: xgene: Fix IB window setup

      https://git.kernel.org/lpieralisi/pci/c/c7a75d0782

Thanks,
Lorenzo

#4 lore raw
In reply to Rob Herring
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

-dann

author reviewer
#5 lore raw
Quoted 28 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

Quoted 6 lines

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Otherwise, we're going to need some debugging added to xgene_pcie_setup_ib_reg() to see if the register setup changed. I can come up with something next week.

Rob

#6 lore raw
Quoted 34 lines · Rob Herring
Quoted 32 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

Quoted 5 lines
Quoted 2 lines

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Sure: http://dannf.org/bugs/fdt

-dann

Quoted 5 lines

Otherwise, we're going to need some debugging added to xgene_pcie_setup_ib_reg() to see if the register setup changed. I can come up with something next week.

Rob

author reviewer
#7 lore raw
Quoted 36 lines · dann frazier
Quoted 34 lines · Rob Herring
Quoted 32 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Sure: http://dannf.org/bugs/fdt

The first fix certainly is a problem. It's going to need something besides size to key off of (originally it was dependent on order of dma-ranges entries).

The 2nd issue is the 'dma-ranges' has a second entry that is now ignored:

dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00>, <0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;

Based on the flags (3rd addr cell: 0x0), we have an inbound config space which the kernel now ignores because inbound config space accesses make no sense. But clearly some setup is needed. Upstream, in contrast, sets up a memory range that includes this region, so the setup does happen:

<0x42000000 0x00 0x00000000 0x00 0x00000000 0x80 0x00000000>

Minimally, I suspect it will work if you change dma-ranges 2nd entry to:

<0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>

While we shouldn't break existing DTs, the moonshot DT doesn't use what's documented upstream. There are multiple differences compared to what's documented. Is upstream supposed to support upstream DTs, downstream DTs, and ACPI for XGene which is an abandoned platform with only a handful of users?

Rob

#8 lore raw
Quoted 60 lines · Rob Herring
Quoted 36 lines · dann frazier
Quoted 34 lines · Rob Herring
Quoted 32 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Sure: http://dannf.org/bugs/fdt

The first fix certainly is a problem. It's going to need something besides size to key off of (originally it was dependent on order of dma-ranges entries).

The 2nd issue is the 'dma-ranges' has a second entry that is now ignored:

dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00>, <0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;

Based on the flags (3rd addr cell: 0x0), we have an inbound config space which the kernel now ignores because inbound config space accesses make no sense. But clearly some setup is needed. Upstream, in contrast, sets up a memory range that includes this region, so the setup does happen:

Quoted 5 lines

<0x42000000 0x00 0x00000000 0x00 0x00000000 0x80 0x00000000>

Minimally, I suspect it will work if you change dma-ranges 2nd entry to:

<0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>

Thanks for looking into this Rob. I tried to test that theory, but it didn't seem to work. This is what I tried:

m400.dts.dmaonly renamed from m400.dts +5 -5
@@ -446,7 +446,7 @@
 			reg = <0x00 0x1f2b0000 0x00 0x10000 0xe0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
 			reg-names = "csr\0cfg\0msi_gen\0msi_term";
 			ranges = <0x1000000 0x00 0x00 0xe0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xe1 0x30000000 0x00 0x80000000>;
-			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
 			interrupts = <0x00 0x10 0x04>;

@@ -471,7 +471,7 @@
 			reg = <0x00 0x1f2c0000 0x00 0x10000 0xd0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
 			reg-names = "csr\0cfg\0msi_gen\0msi_term";
 			ranges = <0x1000000 0x00 0x00 0xd0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xd1 0x30000000 0x00 0x80000000>;
-			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
 			interrupts = <0x00 0x10 0x04>;

@@ -496,7 +496,7 @@
 			reg = <0x00 0x1f2d0000 0x00 0x10000 0x90 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
 			reg-names = "csr\0cfg\0msi_gen\0msi_term";
 			ranges = <0x1000000 0x00 0x00 0x90 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0x91 0x30000000 0x00 0x80000000>;
-			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
 			interrupts = <0x00 0x10 0x04>;

@@ -522,7 +522,7 @@
 			reg = <0x00 0x1f500000 0x00 0x10000 0xa0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
 			reg-names = "csr\0cfg\0msi_gen\0msi_term";
 			ranges = <0x2000000 0x00 0x30000000 0xa1 0x30000000 0x00 0x80000000>;
-			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
 			interrupts = <0x00 0x10 0x04>;

@@ -547,7 +547,7 @@
 			reg = <0x00 0x1f510000 0x00 0x10000 0xc0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
 			reg-names = "csr\0cfg\0msi_gen\0msi_term";
 			ranges = <0x1000000 0x00 0x00 0xc0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xc1 0x30000000 0x00 0x80000000>;
-			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+			dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
 			ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
 			interrupts = <0x00 0x10 0x04>;

And that failed to boot with a 5.17-rc3. Since dma-ranges was previously identical to ib-ranges, I also tried making the same change to ib-ranges, but with no success.

Quoted 5 lines

While we shouldn't break existing DTs, the moonshot DT doesn't use what's documented upstream. There are multiple differences compared to what's documented. Is upstream supposed to support upstream DTs, downstream DTs, and ACPI for XGene which is an abandoned platform with only a handful of users?

That's a fair question, though it's one of a policy, and I feel I'd be overstepping by weighing in. I suppose one option I have is to try and create and upstream a dts for these systems and modify our boot.scr to always load that over the one provided by firmware. While we do have some of these systems in production, they are being retired and replaced with newer kit over time, and it's possible we'll never need to upgrade them to a modern kernel.

-dann

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

author reviewer
#9 lore raw
Quoted 119 lines · dann frazier
Quoted 60 lines · Rob Herring
Quoted 36 lines · dann frazier
Quoted 34 lines · Rob Herring
Quoted 32 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Sure: http://dannf.org/bugs/fdt

The first fix certainly is a problem. It's going to need something besides size to key off of (originally it was dependent on order of dma-ranges entries).

The 2nd issue is the 'dma-ranges' has a second entry that is now ignored:

dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00>, <0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;

Based on the flags (3rd addr cell: 0x0), we have an inbound config space which the kernel now ignores because inbound config space accesses make no sense. But clearly some setup is needed. Upstream, in contrast, sets up a memory range that includes this region, so the setup does happen:

Quoted 5 lines

<0x42000000 0x00 0x00000000 0x00 0x00000000 0x80 0x00000000>

Minimally, I suspect it will work if you change dma-ranges 2nd entry to:

<0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>

Thanks for looking into this Rob. I tried to test that theory, but it didn't seem to work. This is what I tried:

m400.dts.dmaonly 2022-02-08 00:17:54.097132000 +0000 renamed from m400.dts 2022-02-07 20:16:44.840475323 +0000 +5 -5
@@ -446,7 +446,7 @@
                        reg = <0x00 0x1f2b0000 0x00 0x10000 0xe0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xe0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xe1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -471,7 +471,7 @@
                        reg = <0x00 0x1f2c0000 0x00 0x10000 0xd0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xd0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xd1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -496,7 +496,7 @@
                        reg = <0x00 0x1f2d0000 0x00 0x10000 0x90 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0x90 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0x91 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -522,7 +522,7 @@
                        reg = <0x00 0x1f500000 0x00 0x10000 0xa0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x2000000 0x00 0x30000000 0xa1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -547,7 +547,7 @@
                        reg = <0x00 0x1f510000 0x00 0x10000 0xc0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xc0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xc1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;
Quoted 3 lines

And that failed to boot with a 5.17-rc3. Since dma-ranges was previously identical to ib-ranges, I also tried making the same change to ib-ranges, but with no success.

Failed to boot at all or just PCIe still didn't work causing boot to eventually fail? 'ib-ranges' is unknown to the kernel, so the firmware is using it somehow?

You also need to revert the first fix for PCIe to work.

Quoted 15 lines
Quoted 5 lines

While we shouldn't break existing DTs, the moonshot DT doesn't use what's documented upstream. There are multiple differences compared to what's documented. Is upstream supposed to support upstream DTs, downstream DTs, and ACPI for XGene which is an abandoned platform with only a handful of users?

That's a fair question, though it's one of a policy, and I feel I'd be overstepping by weighing in. I suppose one option I have is to try and create and upstream a dts for these systems and modify our boot.scr to always load that over the one provided by firmware. While we do have some of these systems in production, they are being retired and replaced with newer kit over time, and it's possible we'll never need to upgrade them to a modern kernel.

-dann

#10 lore raw
Quoted 120 lines · Rob Herring
Quoted 119 lines · dann frazier
Quoted 60 lines · Rob Herring
Quoted 36 lines · dann frazier
Quoted 34 lines · Rob Herring
Quoted 32 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Sure: http://dannf.org/bugs/fdt

The first fix certainly is a problem. It's going to need something besides size to key off of (originally it was dependent on order of dma-ranges entries).

The 2nd issue is the 'dma-ranges' has a second entry that is now ignored:

dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00>, <0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;

Based on the flags (3rd addr cell: 0x0), we have an inbound config space which the kernel now ignores because inbound config space accesses make no sense. But clearly some setup is needed. Upstream, in contrast, sets up a memory range that includes this region, so the setup does happen:

Quoted 5 lines

<0x42000000 0x00 0x00000000 0x00 0x00000000 0x80 0x00000000>

Minimally, I suspect it will work if you change dma-ranges 2nd entry to:

<0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>

Thanks for looking into this Rob. I tried to test that theory, but it didn't seem to work. This is what I tried:

m400.dts.dmaonly 2022-02-08 00:17:54.097132000 +0000 renamed from m400.dts 2022-02-07 20:16:44.840475323 +0000 +5 -5
@@ -446,7 +446,7 @@
                        reg = <0x00 0x1f2b0000 0x00 0x10000 0xe0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xe0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xe1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -471,7 +471,7 @@
                        reg = <0x00 0x1f2c0000 0x00 0x10000 0xd0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xd0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xd1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -496,7 +496,7 @@
                        reg = <0x00 0x1f2d0000 0x00 0x10000 0x90 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0x90 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0x91 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -522,7 +522,7 @@
                        reg = <0x00 0x1f500000 0x00 0x10000 0xa0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x2000000 0x00 0x30000000 0xa1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -547,7 +547,7 @@
                        reg = <0x00 0x1f510000 0x00 0x10000 0xc0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xc0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xc1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;
Quoted 6 lines
Quoted 3 lines

And that failed to boot with a 5.17-rc3. Since dma-ranges was previously identical to ib-ranges, I also tried making the same change to ib-ranges, but with no success.

Failed to boot at all or just PCIe still didn't work causing boot to eventually fail?

Sorry, I mean PCIe still didn't work, here's the log:
http://dannf.org/bugs/m400-tweaked_dtb.log
(unmodified kernel source w/ above dtb change)

Quoted 4 lines

'ib-ranges' is unknown to the kernel, so the firmware is using it somehow?

You also need to revert the first fix for PCIe to work.

Oh, OK. I misunderstood. I tried reverting commit 6dce5aa59e0b "PCI:
xgene: Use inbound resources for setup" along with a dtb with the dma-ranges change in the diff above, but PCIe still didn't work. Here's the log:

http://dannf.org/bugs/m400-6dce5aa5_reverted+tweaked_dtb.log

-dann

Quoted 16 lines
Quoted 16 lines
Quoted 6 lines

While we shouldn't break existing DTs, the moonshot DT doesn't use what's documented upstream. There are multiple differences compared to what's documented. Is upstream supposed to support upstream DTs, downstream DTs, and ACPI for XGene which is an abandoned platform with only a handful of users?

That's a fair question, though it's one of a policy, and I feel I'd be overstepping by weighing in. I suppose one option I have is to try and create and upstream a dts for these systems and modify our boot.scr to always load that over the one provided by firmware. While we do have some of these systems in production, they are being retired and replaced with newer kit over time, and it's possible we'll never need to upgrade them to a modern kernel.

-dann

#11 lore raw

Hi, this is your Linux kernel regression tracker speaking. Top-posting for once, to make this easy accessible to everyone.

What's the status of this regression and getting it fixed? It looks like there was quite some progress for a while, but then things seem to have come to a halt ten days ago. Could anyone please provide a status update, please?

Ciao, Thorsten (wearing his 'Linux kernel regression tracker' hat)

P.S.: As a Linux kernel regression tracker I'm getting a lot of reports on my table. I can only look briefly into most of them. Unfortunately therefore I sometimes will get things wrong or miss something important. I hope that's not the case here; if you think it is, don't hesitate to tell me about it in a public reply, that's in everyone's interest.

#regzbot poke

Quoted 162 lines · dann frazier
Quoted 120 lines · Rob Herring
Quoted 119 lines · dann frazier
Quoted 60 lines · Rob Herring
Quoted 36 lines · dann frazier
Quoted 34 lines · Rob Herring
Quoted 32 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

The ranges and dma-ranges addresses don't appear to match up with any upstream dts files. Can you send me the DT?

Sure: http://dannf.org/bugs/fdt

The first fix certainly is a problem. It's going to need something besides size to key off of (originally it was dependent on order of dma-ranges entries).

The 2nd issue is the 'dma-ranges' has a second entry that is now ignored:

dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00>, <0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;

Based on the flags (3rd addr cell: 0x0), we have an inbound config space which the kernel now ignores because inbound config space accesses make no sense. But clearly some setup is needed. Upstream, in contrast, sets up a memory range that includes this region, so the setup does happen:

Quoted 5 lines

<0x42000000 0x00 0x00000000 0x00 0x00000000 0x80 0x00000000>

Minimally, I suspect it will work if you change dma-ranges 2nd entry to:

<0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>

Thanks for looking into this Rob. I tried to test that theory, but it didn't seem to work. This is what I tried:

m400.dts.dmaonly 2022-02-08 00:17:54.097132000 +0000 renamed from m400.dts 2022-02-07 20:16:44.840475323 +0000 +5 -5
@@ -446,7 +446,7 @@
                        reg = <0x00 0x1f2b0000 0x00 0x10000 0xe0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xe0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xe1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -471,7 +471,7 @@
                        reg = <0x00 0x1f2c0000 0x00 0x10000 0xd0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xd0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xd1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -496,7 +496,7 @@
                        reg = <0x00 0x1f2d0000 0x00 0x10000 0x90 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0x90 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0x91 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -522,7 +522,7 @@
                        reg = <0x00 0x1f500000 0x00 0x10000 0xa0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x2000000 0x00 0x30000000 0xa1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;

@@ -547,7 +547,7 @@
                        reg = <0x00 0x1f510000 0x00 0x10000 0xc0 0xd0000000 0x00 0x200000 0x00 0x79e00000 0x00 0x2000000 0x00 0x79000000 0x00 0x800000>;
                        reg-names = "csr\0cfg\0msi_gen\0msi_term";
                        ranges = <0x1000000 0x00 0x00 0xc0 0x10000000 0x00 0x10000 0x2000000 0x00 0x30000000 0xc1 0x30000000 0x00 0x80000000>;
-                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
+                       dma-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x42000000 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges = <0x42000000 0x40 0x00 0x40 0x00 0x40 0x00 0x00 0x00 0x79000000 0x00 0x79000000 0x00 0x800000>;
                        ib-ranges-ep = <0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x00 0x00 0x00 0x00 0x400000 0x2000000 0x00 0x79000000 0x00 0x79000000 0x00 0x100000>;
                        interrupts = <0x00 0x10 0x04>;
Quoted 6 lines
Quoted 3 lines

And that failed to boot with a 5.17-rc3. Since dma-ranges was previously identical to ib-ranges, I also tried making the same change to ib-ranges, but with no success.

Failed to boot at all or just PCIe still didn't work causing boot to eventually fail?

Sorry, I mean PCIe still didn't work, here's the log:
http://dannf.org/bugs/m400-tweaked_dtb.log
(unmodified kernel source w/ above dtb change)

Quoted 4 lines

'ib-ranges' is unknown to the kernel, so the firmware is using it somehow?

You also need to revert the first fix for PCIe to work.

Oh, OK. I misunderstood. I tried reverting commit 6dce5aa59e0b "PCI:
xgene: Use inbound resources for setup" along with a dtb with the dma-ranges change in the diff above, but PCIe still didn't work. Here's the log:

http://dannf.org/bugs/m400-6dce5aa5_reverted+tweaked_dtb.log

-dann

Quoted 16 lines
Quoted 16 lines
Quoted 6 lines

While we shouldn't break existing DTs, the moonshot DT doesn't use what's documented upstream. There are multiple differences compared to what's documented. Is upstream supposed to support upstream DTs, downstream DTs, and ACPI for XGene which is an abandoned platform with only a handful of users?

That's a fair question, though it's one of a policy, and I feel I'd be overstepping by weighing in. I suppose one option I have is to try and create and upstream a dts for these systems and modify our boot.scr to always load that over the one provided by firmware. While we do have some of these systems in production, they are being retired and replaced with newer kit over time, and it's possible we'll never need to upgrade them to a modern kernel.

-dann

#12 lore raw
In reply to dann frazier

[TLDR: I'm adding the regression report below to regzbot, the Linux kernel regression tracking bot; nearly all text you find below is compiled from a few templates paragraphs you likely have encountered already already from mails similar to this one.]

Hi, this is your Linux kernel regression tracker speaking.

CCing the regression mailing list, as it should be in the loop for all regressions, as explained here:
https://www.kernel.org/doc/html/latest/admin-guide/reporting-issues.html

Quoted 27 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

Quoted 6 lines

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

Thanks for the report.

To be sure this issue doesn't fall through the cracks unnoticed, I'm adding it to regzbot, my Linux kernel regression tracking bot:

#regzbot ^introduced c7a75d07827a1f33d
#regzbot title Follow-up error for the commit fixing "PCIe regression on APM Merlin (aarch64 dev platform) preventing NVME initialization" #regzbot ignore-activity

Reminder for developers: when fixing the issue, please add a 'Link:' tags pointing to the report (the mail quoted above) using the lore.kernel.org/r/, as explained in 'Documentation/process/submitting-patches.rst' and 'Documentation/process/5.Posting.rst', as this allows the bot to assign any fixes posted or commited with the report to always show the current status of things and automatically close the issue when the fix hits the right tree.

I'm sending this to everyone that got the initial report, to make them aware of the tracking. I also hope that messages like this motivate people to directly get at least the regression mailing list and ideally even regzbot involved when dealing with regressions, as messages like this wouldn't be needed then.

Don't worry, I'll send further messages wrt to this regression just to the lists (with a tag in the subject so people can filter them away), if they are relevant just for regzbot. With a bit of luck no such messages will be needed anyway.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)

P.S.: As the Linux kernel's regression tracker I'm getting a lot of reports on my table. I can only look briefly into most of them and lack knowledge about most of the areas they concern. I thus unfortunately will sometimes get things wrong or miss something important. I hope that's not the case here; if you think it is, don't hesitate to tell me in a public reply, it's in everyone's interest to set the public record straight.

Signature · 25 lines
Additional information about regzbot:

If you want to know more about regzbot, check out its web-interface, the
getting start guide, and the references documentation:

https://linux-regtracking.leemhuis.info/regzbot/
https://gitlab.com/knurd42/regzbot/-/blob/main/docs/getting_started.md
https://gitlab.com/knurd42/regzbot/-/blob/main/docs/reference.md

The last two documents will explain how you can interact with regzbot
yourself if your want to.

Hint for reporters: when reporting a regression it's in your interest to
CC the regression list and tell regzbot about the issue, as that ensures
the regression makes it onto the radar of the Linux kernel's regression
tracker -- that's in your interest, as it ensures your report won't fall
through the cracks unnoticed.

Hint for developers: you normally don't need to care about regzbot once
it's involved. Fix the issue as you normally would, just remember to
include 'Link:' tag in the patch descriptions pointing to all reports
about the issue. This has been expected from developers even before
regzbot showed up for reasons explained in
'Documentation/process/submitting-patches.rst' and
'Documentation/process/5.Posting.rst'.
#13 lore raw

Re: [PATCH] PCI: xgene: Fix IB window setup #forregzbot

TWIMC: this mail is primarily send for documentation purposes and for regzbot, my Linux kernel regression tracking bot. These mails usually contain '#forregzbot' in the subject, to make them easy to spot and filter.

Forwarding a few details from here:
https://lore.kernel.org/regressions/CAL_JsqLHun+X4jMwTbVMmjjETfbP73j52XCwWBj9MJCkpQ41mA@mail.gmail.com/

#regzbot introduced: 6dce5aa59e0b
#regzbot title: c7a75d07827a fixed 6dce5aa59e0b for XGene2, but that *further* broke XGene1
#regzbot backburner: needs more debugging and number of people who care likely small anyway

Quoted 86 lines · Thorsten Leemhuis

[TLDR: I'm adding the regression report below to regzbot, the Linux kernel regression tracking bot; nearly all text you find below is compiled from a few templates paragraphs you likely have encountered already already from mails similar to this one.]

Hi, this is your Linux kernel regression tracker speaking.

CCing the regression mailing list, as it should be in the loop for all regressions, as explained here:
https://www.kernel.org/doc/html/latest/admin-guide/reporting-issues.html

Quoted 27 lines · dann frazier
Quoted 8 lines · Rob Herring

Commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup") broke PCI support on XGene. The cause is the IB resources are now sorted in address order instead of being in DT dma-ranges order. The result is which inbound registers are used for each region are swapped. I don't know the details about this h/w, but it appears that IB region 0 registers can't handle a size greater than 4GB. In any case, limiting the size for region 0 is enough to get back to the original assignment of dma-ranges to regions.

hey Rob!

I've been seeing a panic on HP Moonshoot m400 cartridges (X-Gene1) - only during network installs - that I also bisected down to commit 6dce5aa59e0b ("PCI: xgene: Use inbound resources for setup"). I was hoping that this patch that fixed the issue on Stéphane's X-Gene2 system would also fix my issue, but no luck. In fact, it seems to just makes it fail differently. Reverting both patches is required to get a v5.17-rc kernel to boot.

I've collected the following logs - let me know if anything else would be useful.

1) v5.17-rc2+ (unmodified):
http://dannf.org/bugs/m400-no-reverts.log
Note that the mlx4 driver fails initialization.

Quoted 6 lines

2) v5.17-rc2+, w/o the commit that fixed Stéphane's system:
http://dannf.org/bugs/m400-xgene2-fix-reverted.log
Note the mlx4 MSI-X timeout, and later panic.

3) v5.17-rc2+, w/ both commits reverted (works)
http://dannf.org/bugs/m400-both-reverted.log

Thanks for the report.

To be sure this issue doesn't fall through the cracks unnoticed, I'm adding it to regzbot, my Linux kernel regression tracking bot:

#regzbot ^introduced c7a75d07827a1f33d
#regzbot title Follow-up error for the commit fixing "PCIe regression on APM Merlin (aarch64 dev platform) preventing NVME initialization" #regzbot ignore-activity

Reminder for developers: when fixing the issue, please add a 'Link:' tags pointing to the report (the mail quoted above) using the lore.kernel.org/r/, as explained in 'Documentation/process/submitting-patches.rst' and 'Documentation/process/5.Posting.rst', as this allows the bot to assign any fixes posted or commited with the report to always show the current status of things and automatically close the issue when the fix hits the right tree.

I'm sending this to everyone that got the initial report, to make them aware of the tracking. I also hope that messages like this motivate people to directly get at least the regression mailing list and ideally even regzbot involved when dealing with regressions, as messages like this wouldn't be needed then.

Don't worry, I'll send further messages wrt to this regression just to the lists (with a tag in the subject so people can filter them away), if they are relevant just for regzbot. With a bit of luck no such messages will be needed anyway.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)

P.S.: As the Linux kernel's regression tracker I'm getting a lot of reports on my table. I can only look briefly into most of them and lack knowledge about most of the areas they concern. I thus unfortunately will sometimes get things wrong or miss something important. I hope that's not the case here; if you think it is, don't hesitate to tell me in a public reply, it's in everyone's interest to set the public record straight.