linux-next: build failure after merge of the pagemap-headers tree

linux-next: build failure after merge of the pagemap-headers tree

[Mark Brown]

Hi all,

After merging the pagemap-headers tree, today's linux-next build
(powerpc ppc4xx_defconfig) failed like this:

In file included from /tmp/next/build/include/linux/preempt.h:11,
                 from /tmp/next/build/include/linux/alloc_tag.h:11,
                 from /tmp/next/build/include/linux/percpu.h:5,
                 from /tmp/next/build/arch/powerpc/include/asm/time.h:14,
                 from /tmp/next/build/arch/powerpc/platforms/44x/sam440ep.c:21:
/tmp/next/build/include/linux/acpi.h: In function '__free_acpi_object_free':
/tmp/next/build/include/linux/cleanup.h:211:92: error: suggest braces around empty body in an 'if' statement [-Werror=empty-body]
  211 |         static __always_inline void __free_##_name(void *p) { _type _T = *(_type *)p; _free; }
      |                                                                                            ^
/tmp/next/build/include/linux/acpi.h:31:1: note: in expansion of macro 'DEFINE_FREE'
   31 | DEFINE_FREE(acpi_object_free, union acpi_object *, if (_T) ACPI_FREE(_T));
      | ^~~~~~~~~~~
cc1: all warnings being treated as errors
make[6]: *** [/tmp/next/build/scripts/Makefile.build:290: arch/powerpc/platforms/44x/sam440ep.o] Error 1
make[6]: *** Waiting for unfinished jobs....
In file included from /tmp/next/build/include/linux/preempt.h:11,
                 from /tmp/next/build/include/linux/sched.h:15,
                 from /tmp/next/build/include/linux/ratelimit.h:6,
                 from /tmp/next/build/include/linux/dev_printk.h:16,
                 from /tmp/next/build/include/linux/device.h:15,
                 from /tmp/next/build/include/linux/platform_device.h:13,
                 from /tmp/next/build/arch/powerpc/platforms/44x/warp.c:11:
/tmp/next/build/include/linux/acpi.h: In function '__free_acpi_object_free':
/tmp/next/build/include/linux/cleanup.h:211:92: error: suggest braces around empty body in an 'if' statement [-Werror=empty-body]
  211 |         static __always_inline void __free_##_name(void *p) { _type _T = *(_type *)p; _free; }
      |                                                                                            ^
/tmp/next/build/include/linux/acpi.h:31:1: note: in expansion of macro 'DEFINE_FREE'
   31 | DEFINE_FREE(acpi_object_free, union acpi_object *, if (_T) ACPI_FREE(_T));
      | ^~~~~~~~~~~
cc1: all warnings being treated as errors

Caused by commit

   f54586fcd661e (ACPI: fan: Use __free() to simplify AML error handling)

(it was also causing lots of warnings in other GCC configurations but
didn't trip -Werror in allmodconfig since it doesn't seem to upset
clang.).  I have reverted that commit for today.

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

[Wysocki, Rafael J]

Hi Mark,

On 9/24/2026 8:17 PM, Mark Brown wrote:
> Hi all,
>
> After merging the pagemap-headers tree, today's linux-next build
> (powerpc ppc4xx_defconfig) failed like this:
>
> In file included from /tmp/next/build/include/linux/preempt.h:11,
>                   from /tmp/next/build/include/linux/alloc_tag.h:11,
>                   from /tmp/next/build/include/linux/percpu.h:5,
>                   from /tmp/next/build/arch/powerpc/include/asm/time.h:14,
>                   from /tmp/next/build/arch/powerpc/platforms/44x/sam440ep.c:21:
> /tmp/next/build/include/linux/acpi.h: In function '__free_acpi_object_free':
> /tmp/next/build/include/linux/cleanup.h:211:92: error: suggest braces around empty body in an 'if' statement [-Werror=empty-body]
>    211 |         static __always_inline void __free_##_name(void *p) { _type _T = *(_type *)p; _free; }
>        |                                                                                            ^
> /tmp/next/build/include/linux/acpi.h:31:1: note: in expansion of macro 'DEFINE_FREE'
>     31 | DEFINE_FREE(acpi_object_free, union acpi_object *, if (_T) ACPI_FREE(_T));
>        | ^~~~~~~~~~~
> cc1: all warnings being treated as errors
> make[6]: *** [/tmp/next/build/scripts/Makefile.build:290: arch/powerpc/platforms/44x/sam440ep.o] Error 1
> make[6]: *** Waiting for unfinished jobs....
> In file included from /tmp/next/build/include/linux/preempt.h:11,
>                   from /tmp/next/build/include/linux/sched.h:15,
>                   from /tmp/next/build/include/linux/ratelimit.h:6,
>                   from /tmp/next/build/include/linux/dev_printk.h:16,
>                   from /tmp/next/build/include/linux/device.h:15,
>                   from /tmp/next/build/include/linux/platform_device.h:13,
>                   from /tmp/next/build/arch/powerpc/platforms/44x/warp.c:11:
> /tmp/next/build/include/linux/acpi.h: In function '__free_acpi_object_free':
> /tmp/next/build/include/linux/cleanup.h:211:92: error: suggest braces around empty body in an 'if' statement [-Werror=empty-body]
>    211 |         static __always_inline void __free_##_name(void *p) { _type _T = *(_type *)p; _free; }
>        |                                                                                            ^
> /tmp/next/build/include/linux/acpi.h:31:1: note: in expansion of macro 'DEFINE_FREE'
>     31 | DEFINE_FREE(acpi_object_free, union acpi_object *, if (_T) ACPI_FREE(_T));
>        | ^~~~~~~~~~~
> cc1: all warnings being treated as errors
>
> Caused by commit
>
>     f54586fcd661e (ACPI: fan: Use __free() to simplify AML error handling)
>
> (it was also causing lots of warnings in other GCC configurations but
> didn't trip -Werror in allmodconfig since it doesn't seem to upset
> clang.).  I have reverted that commit for today.

Sorry for the breakage!

Apparently, I forgot to fix up the patch before applying it (even though 
I update the changelog to that effect).

Should be fixed now.

Thanks, Rafael

linux-next: build failure after merge of the pagemap-headers tree

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 551 bytes --]

Hi all,

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

/tmp/next/build/drivers/media/pci/intel/ipu6/ipu7-fw-isys.c:229:3: error: call to undeclared function 'clflush_cache_range'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
  229 |                 clflush_cache_range(cpu_mapped_buf, size);
      |                 ^

Caused by something in your tree interacting with another change, most
likely a header tweak.  I have applied a fixup.

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

linux-next: build failure after merge of the pagemap-headers tree

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 2973 bytes --]

Hi all,

After merging the pagemap-headers tree, today's linux-next build
(arm multi_v7_defconfig) failed like this:

/tmp/next/build/drivers/dma/ste_dma40.c:492:42: error: field 'tasklet' has incomplete type
  492 |         struct tasklet_struct            tasklet;
      |                                          ^~~~~~~
/tmp/next/build/drivers/dma/ste_dma40.c: In function 'dma_tc_handle':
/tmp/next/build/drivers/dma/ste_dma40.c:1590:9: error: implicit declaration of function 'tasklet_schedule' [-Wimplicit-function-declaration]
 1590 |         tasklet_schedule(&d40c->tasklet);
      |         ^~~~~~~~~~~~~~~~
/tmp/next/build/drivers/dma/ste_dma40.c: In function 'dma_tasklet':
/tmp/next/build/drivers/dma/ste_dma40.c:1596:33: error: implicit declaration of function 'from_tasklet'; did you mean 'dma_tasklet'? [-Wimplicit-function-declaration]
 1596 |         struct d40_chan *d40c = from_tasklet(d40c, t, tasklet);
      |                                 ^~~~~~~~~~~~
      |                                 dma_tasklet
/tmp/next/build/drivers/dma/ste_dma40.c:1596:55: error: 'tasklet' undeclared (first use in this function)
 1596 |         struct d40_chan *d40c = from_tasklet(d40c, t, tasklet);
      |                                                       ^~~~~~~
/tmp/next/build/drivers/dma/ste_dma40.c:1596:55: note: each undeclared identifier is reported only once for each function it appears in
/tmp/next/build/drivers/dma/ste_dma40.c: At top level:
/tmp/next/build/drivers/dma/ste_dma40.c:1659:8: error: unknown type name 'irqreturn_t'
 1659 | static irqreturn_t d40_handle_interrupt(int irq, void *data)
      |        ^~~~~~~~~~~
/tmp/next/build/drivers/dma/ste_dma40.c: In function 'd40_handle_interrupt':
/tmp/next/build/drivers/dma/ste_dma40.c:1718:16: error: 'IRQ_HANDLED' undeclared (first use in this function)
 1718 |         return IRQ_HANDLED;
      |                ^~~~~~~~~~~
/tmp/next/build/drivers/dma/ste_dma40.c: In function 'd40_chan_init':
/tmp/next/build/drivers/dma/ste_dma40.c:2818:17: error: implicit declaration of function 'tasklet_setup'; did you mean 'timer_setup'? [-Wimplicit-function-declaration]
 2818 |                 tasklet_setup(&d40c->tasklet, dma_tasklet);
      |                 ^~~~~~~~~~~~~
      |                 timer_setup
/tmp/next/build/drivers/dma/ste_dma40.c: In function 'd40_probe':
/tmp/next/build/drivers/dma/ste_dma40.c:3588:15: error: implicit declaration of function 'request_irq'; did you mean 'request_key'? [-Wimplicit-function-declaration]
 3588 |         ret = request_irq(base->irq, d40_handle_interrupt, 0, D40_NAME, base);
      |               ^~~~~~~~~~~
      |               request_key
/tmp/next/build/drivers/dma/ste_dma40.c: In function 'd40_handle_interrupt':
/tmp/next/build/drivers/dma/ste_dma40.c:1719:1: error: control reaches end of non-void function [-Werror=return-type]
 1719 | }
      | ^

Doubtless caused by the pagemap-headers tree.  I will not merge the tree
today.

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