linux-next: build failure after merge of the irqchip tree

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 703 bytes --]

Hi Jason,

After merging the irqchip tree, today's linux-next build (x86_64
allmodconfig) failed like this:

kernel/irq/generic-chip.c: In function 'irq_map_generic_chip':
kernel/irq/generic-chip.c:408:2: error: implicit declaration of function 'irq_domain_set_info' [-Werror=implicit-function-declaration]
  irq_domain_set_info(d, virq, hw_irq, chip, gc, ct->handler, NULL, NULL);
  ^

Caused by commit 3f6b2c298240 ("genirq: Generic chip: Support hierarchy
domain").  The only declaration of irq_domain_set_info() depends on
CONFIG_IRQ_DOMAIN_HIERARCHY.

I have used the irqchip tree from next-20150410 fot today.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Stefan Agner]

Hi Jason,

On 2015-04-13 09:05, Stephen Rothwell wrote:
> Hi Jason,
> 
> After merging the irqchip tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> kernel/irq/generic-chip.c: In function 'irq_map_generic_chip':
> kernel/irq/generic-chip.c:408:2: error: implicit declaration of
> function 'irq_domain_set_info' [-Werror=implicit-function-declaration]
>   irq_domain_set_info(d, virq, hw_irq, chip, gc, ct->handler, NULL, NULL);
>   ^

Hm, sorry about that, I did not notice that this function is only
available with domain hierarchy. I guess making domain hierarchy a
dependency of genirq is a bit harsh...

--
Stefan

> 
> Caused by commit 3f6b2c298240 ("genirq: Generic chip: Support hierarchy
> domain").  The only declaration of irq_domain_set_info() depends on
> CONFIG_IRQ_DOMAIN_HIERARCHY.
> 
> I have used the irqchip tree from next-20150410 fot today.

Re: linux-next: build failure after merge of the irqchip tree

[Jason Cooper]

Hi Stephen,

On Mon, Apr 13, 2015 at 05:05:39PM +1000, Stephen Rothwell wrote:
> Hi Jason,
> 
> After merging the irqchip tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> kernel/irq/generic-chip.c: In function 'irq_map_generic_chip':
> kernel/irq/generic-chip.c:408:2: error: implicit declaration of function 'irq_domain_set_info' [-Werror=implicit-function-declaration]
>   irq_domain_set_info(d, virq, hw_irq, chip, gc, ct->handler, NULL, NULL);
>   ^
> 
> Caused by commit 3f6b2c298240 ("genirq: Generic chip: Support hierarchy
> domain").  The only declaration of irq_domain_set_info() depends on
> CONFIG_IRQ_DOMAIN_HIERARCHY.

Sorry about that.  I was rushing a bit in case the merge window was delayed a
week.  It wasn't, so I'll drop this for v4.1.

It's also a gentle reminder that I need to do *more* thorough testing
with genirq changes before pushing. :-/

thx,

Jason.

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1356 bytes --]

Hi Marc,

After merging the irqchip tree, today's linux-next build (x86_64
allmodconfig) failed like this:

In file included from drivers/irqchip/irq-madera.c:21:
include/linux/irqchip/irq-madera.h: In function 'madera_get_irq_mapping':
include/linux/irqchip/irq-madera.h:99:37: error: 'struct madera' has no member named 'irq_data'; did you mean 'irq_dev'?
  return regmap_irq_get_virq(madera->irq_data, irq);
                                     ^~~~~~~~
                                     irq_dev
drivers/irqchip/irq-madera.c: In function 'madera_irq_probe':
drivers/irqchip/irq-madera.c:217:34: error: 'struct madera' has no member named 'irq_data'; did you mean 'irq_dev'?
       &madera_irq_chip, &madera->irq_data);
                                  ^~~~~~~~
                                  irq_dev
drivers/irqchip/irq-madera.c: In function 'madera_irq_remove':
drivers/irqchip/irq-madera.c:238:43: error: 'struct madera' has no member named 'irq_data'; did you mean 'irq_dev'?
  regmap_del_irq_chip(madera->irq, madera->irq_data);
                                           ^~~~~~~~
                                           irq_dev

Caused by commit

  b817ff5c9509 ("irqchip: Add driver for Cirrus Logic Madera codecs")

I have used the irqchip tree from next-20181005 for today.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Marc Zyngier]

Hi Stephen,

On Mon, 08 Oct 2018 04:06:21 +0100,
Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> 
> [1  <text/plain; US-ASCII (quoted-printable)>]
> Hi Marc,
> 
> After merging the irqchip tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> In file included from drivers/irqchip/irq-madera.c:21:
> include/linux/irqchip/irq-madera.h: In function 'madera_get_irq_mapping':
> include/linux/irqchip/irq-madera.h:99:37: error: 'struct madera' has no member named 'irq_data'; did you mean 'irq_dev'?
>   return regmap_irq_get_virq(madera->irq_data, irq);
>                                      ^~~~~~~~
>                                      irq_dev
> drivers/irqchip/irq-madera.c: In function 'madera_irq_probe':
> drivers/irqchip/irq-madera.c:217:34: error: 'struct madera' has no member named 'irq_data'; did you mean 'irq_dev'?
>        &madera_irq_chip, &madera->irq_data);
>                                   ^~~~~~~~
>                                   irq_dev
> drivers/irqchip/irq-madera.c: In function 'madera_irq_remove':
> drivers/irqchip/irq-madera.c:238:43: error: 'struct madera' has no member named 'irq_data'; did you mean 'irq_dev'?
>   regmap_del_irq_chip(madera->irq, madera->irq_data);
>                                            ^~~~~~~~
>                                            irq_dev
> 
> Caused by commit
> 
>   b817ff5c9509 ("irqchip: Add driver for Cirrus Logic Madera codecs")
> 
> I have used the irqchip tree from next-20181005 for today.

Thanks for doing that. I've now dropped this commit from irqchip-next.

Richard, I'll let you route this patch to the appropriate tree.

Cheers,

	M.

-- 
Jazz is not dead, it just smell funny.

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1135 bytes --]

Hi all,

After merging the irqchip tree, today's linux-next build (powerpc
ppc44x_defconfig) failed like this:

drivers/irqchip/irq-xilinx-intc.c: In function 'xil_intc_handle_irq':
drivers/irqchip/irq-xilinx-intc.c:176:10: error: implicit declaration of function 'handle_domain_irq'; did you mean 'handle_bad_irq'? [-Werror=implicit-function-declaration]
  176 |    ret = handle_domain_irq(irqc->root_domain, hwirq, regs);
      |          ^~~~~~~~~~~~~~~~~
      |          handle_bad_irq
drivers/irqchip/irq-xilinx-intc.c: In function 'xilinx_intc_of_init':
drivers/irqchip/irq-xilinx-intc.c:253:3: error: implicit declaration of function 'set_handle_irq'; did you mean 'generic_handle_irq'? [-Werror=implicit-function-declaration]
  253 |   set_handle_irq(xil_intc_handle_irq);
      |   ^~~~~~~~~~~~~~
      |   generic_handle_irq

Caused by commit

  a0789993bf82 ("irqchip/xilinx: Enable generic irq multi handler")

I have reverted that commit (and commit

  9c2d4f525c00 ("irqchip/xilinx: Do not call irq_set_default_host()")

that conflicted with the other revert).

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Michal Simek]

Hi Stephen,

On 23. 03. 20 7:59, Stephen Rothwell wrote:
> Hi all,
> 
> After merging the irqchip tree, today's linux-next build (powerpc
> ppc44x_defconfig) failed like this:
> 
> drivers/irqchip/irq-xilinx-intc.c: In function 'xil_intc_handle_irq':
> drivers/irqchip/irq-xilinx-intc.c:176:10: error: implicit declaration of function 'handle_domain_irq'; did you mean 'handle_bad_irq'? [-Werror=implicit-function-declaration]
>   176 |    ret = handle_domain_irq(irqc->root_domain, hwirq, regs);
>       |          ^~~~~~~~~~~~~~~~~
>       |          handle_bad_irq
> drivers/irqchip/irq-xilinx-intc.c: In function 'xilinx_intc_of_init':
> drivers/irqchip/irq-xilinx-intc.c:253:3: error: implicit declaration of function 'set_handle_irq'; did you mean 'generic_handle_irq'? [-Werror=implicit-function-declaration]
>   253 |   set_handle_irq(xil_intc_handle_irq);
>       |   ^~~~~~~~~~~~~~
>       |   generic_handle_irq
> 
> Caused by commit
> 
>   a0789993bf82 ("irqchip/xilinx: Enable generic irq multi handler")
> 
> I have reverted that commit (and commit
> 
>   9c2d4f525c00 ("irqchip/xilinx: Do not call irq_set_default_host()")
> 
> that conflicted with the other revert).
> 

thanks for reporting this.

Marc: I completely forget about this wiring for very ancient platform.
AFAIK Xilinx is not testing this platform for years and would likely the
best to remove support for it completely.
The second option is to get xintc_get_irq() back with note about powerpc
because it won't be that simple change all PPC platform to generic
domain irq handler.
Any other suggestion?

Thanks,
Michal

Re: linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1303 bytes --]

Hi all,

On Mon, 23 Mar 2020 17:59:46 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> After merging the irqchip tree, today's linux-next build (powerpc
> ppc44x_defconfig) failed like this:
> 
> drivers/irqchip/irq-xilinx-intc.c: In function 'xil_intc_handle_irq':
> drivers/irqchip/irq-xilinx-intc.c:176:10: error: implicit declaration of function 'handle_domain_irq'; did you mean 'handle_bad_irq'? [-Werror=implicit-function-declaration]
>   176 |    ret = handle_domain_irq(irqc->root_domain, hwirq, regs);
>       |          ^~~~~~~~~~~~~~~~~
>       |          handle_bad_irq
> drivers/irqchip/irq-xilinx-intc.c: In function 'xilinx_intc_of_init':
> drivers/irqchip/irq-xilinx-intc.c:253:3: error: implicit declaration of function 'set_handle_irq'; did you mean 'generic_handle_irq'? [-Werror=implicit-function-declaration]
>   253 |   set_handle_irq(xil_intc_handle_irq);
>       |   ^~~~~~~~~~~~~~
>       |   generic_handle_irq
> 
> Caused by commit
> 
>   a0789993bf82 ("irqchip/xilinx: Enable generic irq multi handler")
> 
> I have reverted that commit (and commit
> 
>   9c2d4f525c00 ("irqchip/xilinx: Do not call irq_set_default_host()")
> 
> that conflicted with the other revert).

I am still getting these errors.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Michal Simek]

Hi Stephen,

On 26. 03. 20 7:36, Stephen Rothwell wrote:
> Hi all,
> 
> On Mon, 23 Mar 2020 17:59:46 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>
>> After merging the irqchip tree, today's linux-next build (powerpc
>> ppc44x_defconfig) failed like this:
>>
>> drivers/irqchip/irq-xilinx-intc.c: In function 'xil_intc_handle_irq':
>> drivers/irqchip/irq-xilinx-intc.c:176:10: error: implicit declaration of function 'handle_domain_irq'; did you mean 'handle_bad_irq'? [-Werror=implicit-function-declaration]
>>   176 |    ret = handle_domain_irq(irqc->root_domain, hwirq, regs);
>>       |          ^~~~~~~~~~~~~~~~~
>>       |          handle_bad_irq
>> drivers/irqchip/irq-xilinx-intc.c: In function 'xilinx_intc_of_init':
>> drivers/irqchip/irq-xilinx-intc.c:253:3: error: implicit declaration of function 'set_handle_irq'; did you mean 'generic_handle_irq'? [-Werror=implicit-function-declaration]
>>   253 |   set_handle_irq(xil_intc_handle_irq);
>>       |   ^~~~~~~~~~~~~~
>>       |   generic_handle_irq
>>
>> Caused by commit
>>
>>   a0789993bf82 ("irqchip/xilinx: Enable generic irq multi handler")
>>
>> I have reverted that commit (and commit
>>
>>   9c2d4f525c00 ("irqchip/xilinx: Do not call irq_set_default_host()")
>>
>> that conflicted with the other revert).
> 
> I am still getting these errors.
> 


Marc: Can you please respond my email I sent on Monday when Stephen
reported this issue?

Thanks,
Michal

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 472 bytes --]

Hi all,

After merging the irqchip tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/irqchip/irq-wpcm450-aic.c:9:10: fatal error: asm/exception.h: No such file or directory
    9 | #include <asm/exception.h>
      |          ^~~~~~~~~~~~~~~~~

Caused by commit

  fead4dd49663 ("irqchip: Add driver for WPCM450 interrupt controller")

I have used the irqchip tree from next-20210407 for today.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Marc Zyngier]

Hi Stephen,

On 2021-04-08 07:35, Stephen Rothwell wrote:
> Hi all,
> 
> After merging the irqchip tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> drivers/irqchip/irq-wpcm450-aic.c:9:10: fatal error: asm/exception.h:
> No such file or directory
>     9 | #include <asm/exception.h>
>       |          ^~~~~~~~~~~~~~~~~
> 
> Caused by commit
> 
>   fead4dd49663 ("irqchip: Add driver for WPCM450 interrupt controller")
> 
> I have used the irqchip tree from next-20210407 for today.

Thanks for the heads up. I guess that's the effect of COMPILE_TEST
which was apparently not very well tested... I'll drop it from Kconfig.

Jonathan, feel free to submit something re-enabling COMPILE_TEST once
you've worked out the missing dependencies.

Thanks,

         M.
-- 
Jazz is not dead. It just smells funny...

Re: linux-next: build failure after merge of the irqchip tree

[Jonathan Neuschäfer]

[-- Attachment #1: Type: text/plain, Size: 1157 bytes --]

On Thu, Apr 08, 2021 at 08:56:18AM +0100, Marc Zyngier wrote:
> Hi Stephen,
> 
> On 2021-04-08 07:35, Stephen Rothwell wrote:
> > Hi all,
> > 
> > After merging the irqchip tree, today's linux-next build (x86_64
> > allmodconfig) failed like this:
> > 
> > drivers/irqchip/irq-wpcm450-aic.c:9:10: fatal error: asm/exception.h:
> > No such file or directory
> >     9 | #include <asm/exception.h>
> >       |          ^~~~~~~~~~~~~~~~~
> > 
> > Caused by commit
> > 
> >   fead4dd49663 ("irqchip: Add driver for WPCM450 interrupt controller")
> > 
> > I have used the irqchip tree from next-20210407 for today.
> 
> Thanks for the heads up. I guess that's the effect of COMPILE_TEST
> which was apparently not very well tested... I'll drop it from Kconfig.

Right, sorry about that.

> Jonathan, feel free to submit something re-enabling COMPILE_TEST once
> you've worked out the missing dependencies.

I used __exception_irq_entry from asm/exception.h, like other irqchip
drivers for ARM. This macro is only defined in arch/arm and arch/arm64.
So, AFAICS, there is no right set of dependencies for COMPILE_TEST.


Jonathan

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1862 bytes --]

Hi all,

After merging the irqchip tree, today's linux-next build (powerpc
ppc64_defconfig) failed like this:

arch/powerpc/platforms/pseries/vas.c: In function 'allocate_setup_window':
arch/powerpc/platforms/pseries/vas.c:219:22: error: implicit declaration of function 'irq_create_mapping' [-Werror=implicit-function-declaration]
  219 |  txwin->fault_virq = irq_create_mapping(NULL, txwin->fault_irq);
      |                      ^~~~~~~~~~~~~~~~~~
arch/powerpc/platforms/pseries/vas.c:248:2: error: implicit declaration of function 'irq_dispose_mapping' [-Werror=implicit-function-declaration]
  248 |  irq_dispose_mapping(txwin->fault_virq);
      |  ^~~~~~~~~~~~~~~~~~~

Caused by commit

  582f5aa1dbb3 ("powerpc: Drop dependency between asm/irq.h and linux/irqdomain.h")

interacting with commit

  6d0aaf5e0de0 ("powerpc/pseries/vas: Setup IRQ and fault handling")

from the powerpc tree

I have applied the following merge fix patch for today.  This could, of
course, be applied directly to the powerpc tree ...

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 22 Jun 2021 14:10:44 +1000
Subject: [PATCH] powerpc: fix up for "Drop dependency between asm/irq.h and linux/irqdomain.h"

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/platforms/pseries/vas.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/powerpc/platforms/pseries/vas.c b/arch/powerpc/platforms/pseries/vas.c
index 3385b5400cc6..b5c1cf1bc64d 100644
--- a/arch/powerpc/platforms/pseries/vas.c
+++ b/arch/powerpc/platforms/pseries/vas.c
@@ -12,6 +12,7 @@
 #include <linux/delay.h>
 #include <linux/slab.h>
 #include <linux/interrupt.h>
+#include <linux/irqdomain.h>
 #include <asm/machdep.h>
 #include <asm/hvcall.h>
 #include <asm/plpar_wrappers.h>
-- 
2.30.2

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1668 bytes --]

Hi all,

After merging the irqchip tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/pci/controller/pci-hyperv.c: In function 'hv_compose_msi_req_v2':
drivers/pci/controller/pci-hyperv.c:1654:42: error: passing argument 1 of 'hv_compose_msi_req_get_cpu' discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]
 1654 |         cpu = hv_compose_msi_req_get_cpu(affinity);
      |                                          ^~~~~~~~
drivers/pci/controller/pci-hyperv.c:1638:55: note: expected 'struct cpumask *' but argument is of type 'const struct cpumask *'
 1638 | static int hv_compose_msi_req_get_cpu(struct cpumask *affinity)
      |                                       ~~~~~~~~~~~~~~~~^~~~~~~~
drivers/pci/controller/pci-hyperv.c: In function 'hv_compose_msi_req_v3':
drivers/pci/controller/pci-hyperv.c:1674:42: error: passing argument 1 of 'hv_compose_msi_req_get_cpu' discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]
 1674 |         cpu = hv_compose_msi_req_get_cpu(affinity);
      |                                          ^~~~~~~~
drivers/pci/controller/pci-hyperv.c:1638:55: note: expected 'struct cpumask *' but argument is of type 'const struct cpumask *'
 1638 | static int hv_compose_msi_req_get_cpu(struct cpumask *affinity)
      |                                       ~~~~~~~~~~~~~~~~^~~~~~~~
cc1: all warnings being treated as errors

Caused by commit

  4d0b8298818b ("genirq: Return a const cpumask from irq_data_get_affinity_mask")

I have used the irqchip tree from next-20220707 for today.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Marc Zyngier]

On Fri, 08 Jul 2022 05:13:44 +0100,
Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> 
> [1  <text/plain; US-ASCII (quoted-printable)>]
> Hi all,
> 
> After merging the irqchip tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> drivers/pci/controller/pci-hyperv.c: In function 'hv_compose_msi_req_v2':
> drivers/pci/controller/pci-hyperv.c:1654:42: error: passing argument 1 of 'hv_compose_msi_req_get_cpu' discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]
>  1654 |         cpu = hv_compose_msi_req_get_cpu(affinity);
>       |                                          ^~~~~~~~
> drivers/pci/controller/pci-hyperv.c:1638:55: note: expected 'struct cpumask *' but argument is of type 'const struct cpumask *'
>  1638 | static int hv_compose_msi_req_get_cpu(struct cpumask *affinity)
>       |                                       ~~~~~~~~~~~~~~~~^~~~~~~~
> drivers/pci/controller/pci-hyperv.c: In function 'hv_compose_msi_req_v3':
> drivers/pci/controller/pci-hyperv.c:1674:42: error: passing argument 1 of 'hv_compose_msi_req_get_cpu' discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]
>  1674 |         cpu = hv_compose_msi_req_get_cpu(affinity);
>       |                                          ^~~~~~~~
> drivers/pci/controller/pci-hyperv.c:1638:55: note: expected 'struct cpumask *' but argument is of type 'const struct cpumask *'
>  1638 | static int hv_compose_msi_req_get_cpu(struct cpumask *affinity)
>       |                                       ~~~~~~~~~~~~~~~~^~~~~~~~
> cc1: all warnings being treated as errors

I've now merged Samuel's quick fix. Thanks both.

	M.

-- 
Without deviation from the norm, progress is not possible.

linux-next: build failure after merge of the irqchip tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1383 bytes --]

Hi all,

After merging the irqchip tree, today's linux-next build (x86_64
allmodconfig) failed like this:

In file included from drivers/gpio/gpio-thunderx.c:18:
include/asm-generic/msi.h:32:3: error: conflicting types for 'msi_alloc_info_t'; have 'struct msi_alloc_info'
   32 | } msi_alloc_info_t;
      |   ^~~~~~~~~~~~~~~~
In file included from include/linux/gpio/driver.h:15,
                 from drivers/gpio/gpio-thunderx.c:10:
arch/x86/include/asm/msi.h:7:31: note: previous declaration of 'msi_alloc_info_t' with type 'msi_alloc_info_t' {aka 'struct irq_alloc_info'}
    7 | typedef struct irq_alloc_info msi_alloc_info_t;
      |                               ^~~~~~~~~~~~~~~~
drivers/gpio/gpio-thunderx.c: In function 'thunderx_gpio_populate_parent_alloc_info':
drivers/gpio/gpio-thunderx.c:416:34: error: initialization of 'msi_alloc_info_t *' {aka 'struct msi_alloc_info *'} from incompatible pointer type 'msi_alloc_info_t *' {aka 'struct irq_alloc_info *'} [-Werror=incompatible-pointer-types]
  416 |         msi_alloc_info_t *info = &gfwspec->msiinfo;
      |                                  ^
cc1: all warnings being treated as errors

Caused by commit

  91a29af413de ("gpio: Remove dynamic allocation from populate_parent_alloc_arg()")

I have used the irqchip tree from next-20220707 again for today.

-- 
Cheers,
Stephen Rothwell

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: build failure after merge of the irqchip tree

[Marc Zyngier]

On Mon, 11 Jul 2022 06:42:52 +0100,
Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> 
> Hi all,
> 
> After merging the irqchip tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> In file included from drivers/gpio/gpio-thunderx.c:18:
> include/asm-generic/msi.h:32:3: error: conflicting types for 'msi_alloc_info_t'; have 'struct msi_alloc_info'
>    32 | } msi_alloc_info_t;
>       |   ^~~~~~~~~~~~~~~~
> In file included from include/linux/gpio/driver.h:15,
>                  from drivers/gpio/gpio-thunderx.c:10:
> arch/x86/include/asm/msi.h:7:31: note: previous declaration of 'msi_alloc_info_t' with type 'msi_alloc_info_t' {aka 'struct irq_alloc_info'}
>     7 | typedef struct irq_alloc_info msi_alloc_info_t;
>       |                               ^~~~~~~~~~~~~~~~
> drivers/gpio/gpio-thunderx.c: In function 'thunderx_gpio_populate_parent_alloc_info':
> drivers/gpio/gpio-thunderx.c:416:34: error: initialization of 'msi_alloc_info_t *' {aka 'struct msi_alloc_info *'} from incompatible pointer type 'msi_alloc_info_t *' {aka 'struct irq_alloc_info *'} [-Werror=incompatible-pointer-types]
>   416 |         msi_alloc_info_t *info = &gfwspec->msiinfo;
>       |                                  ^
> cc1: all warnings being treated as errors
> 
> Caused by commit
> 
>   91a29af413de ("gpio: Remove dynamic allocation from populate_parent_alloc_arg()")
> 
> I have used the irqchip tree from next-20220707 again for today.

Huh. Twice sorry about that one. Patch sent to fix the breakage, and
will push this into irqchip-next shortly.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.

linux-next: build failure after merge of the irqchip tree

[broonie]

Hi all,

After merging the irqchip tree, today's linux-next build (arm64
defconfig) failed like this:

/tmp/next/build/drivers/irqchip/irq-imx-mu-msi.c:14:10: fatal error: linux/dma-iommu.h: No such file or directory
   14 | #include <linux/dma-iommu.h>
      |          ^~~~~~~~~~~~~~~~~~~

Caused by commit

  841e6e9f2bc95baff ("irqchip: Add IMX MU MSI controller driver")

interacting with

  f2042ed21da7f8886 ("iommu/dma: Make header private")

I have reverted the driver for today.

Re: linux-next: build failure after merge of the irqchip tree

[Robin Murphy]

On 2022-09-29 15:31, broonie@kernel.org wrote:
> Hi all,
> 
> After merging the irqchip tree, today's linux-next build (arm64
> defconfig) failed like this:
> 
> /tmp/next/build/drivers/irqchip/irq-imx-mu-msi.c:14:10: fatal error: linux/dma-iommu.h: No such file or directory
>     14 | #include <linux/dma-iommu.h>
>        |          ^~~~~~~~~~~~~~~~~~~
> 
> Caused by commit
> 
>    841e6e9f2bc95baff ("irqchip: Add IMX MU MSI controller driver")
> 
> interacting with
> 
>    f2042ed21da7f8886 ("iommu/dma: Make header private")
> 
> I have reverted the driver for today.

Once again it looks like an unused include which could just be 
removed... illustrating a large part of why I took the header away :)

Cheers,
Robin.

Re: linux-next: build failure after merge of the irqchip tree

[Marc Zyngier]

On Thu, 29 Sep 2022 15:40:56 +0100,
Robin Murphy <robin.murphy@arm.com> wrote:
> 
> On 2022-09-29 15:31, broonie@kernel.org wrote:
> > Hi all,
> > 
> > After merging the irqchip tree, today's linux-next build (arm64
> > defconfig) failed like this:
> > 
> > /tmp/next/build/drivers/irqchip/irq-imx-mu-msi.c:14:10: fatal error: linux/dma-iommu.h: No such file or directory
> >     14 | #include <linux/dma-iommu.h>
> >        |          ^~~~~~~~~~~~~~~~~~~
> > 
> > Caused by commit
> > 
> >    841e6e9f2bc95baff ("irqchip: Add IMX MU MSI controller driver")
> > 
> > interacting with
> > 
> >    f2042ed21da7f8886 ("iommu/dma: Make header private")
> > 
> > I have reverted the driver for today.
> 
> Once again it looks like an unused include which could just be
> removed... illustrating a large part of why I took the header away :)

Thanks both. I've now fixed the driver by dropping this include (and
another one) and pushed the -next branch out again.

	M.

-- 
Without deviation from the norm, progress is not possible.