2023-06-19 01:09:36

by Maciej W. Rozycki

[permalink] [raw]
Subject: [PATCH v2] SPARC: Fix parport_pc support for 32-bit platforms

For 32-bit SPARC platforms PC-style parallel ports are only available as
PCI options. Adjust <asm/parport.h> accordingly, fixing build errors:

sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `ecpp_remove':
parport_pc.c:(.text+0x8f0): undefined reference to `ebus_dma_irq_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x8f8): undefined reference to `ebus_dma_unregister'
sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `parport_pc_fifo_write_block_dma':
parport_pc.c:(.text+0x1430): undefined reference to `ebus_dma_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x1444): undefined reference to `ebus_dma_prepare'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x14e4): undefined reference to `ebus_dma_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x14f4): undefined reference to `ebus_dma_request'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x1584): undefined reference to `ebus_dma_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x1594): undefined reference to `ebus_dma_residue'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x1608): undefined reference to `ebus_dma_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x1618): undefined reference to `ebus_dma_residue'
sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `ecpp_probe':
parport_pc.c:(.text+0x33a0): undefined reference to `ebus_dma_register'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x33b8): undefined reference to `ebus_dma_irq_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x34b0): undefined reference to `ebus_dma_irq_enable'
sparc-linux-gnu-ld: parport_pc.c:(.text+0x34bc): undefined reference to `ebus_dma_unregister'

Reported-by: Randy Dunlap <[email protected]>
Closes: https://lore.kernel.org/r/[email protected]/
Signed-off-by: Maciej W. Rozycki <[email protected]>
Fixes: 66bcd06099bb ("parport_pc: Also enable driver for PCI systems")
Cc: [email protected] # v5.18+
---
NB by enabling CONFIG_WERROR I hit:

sparc-linux-gnu-gcc -Wp,-MMD,kernel/.workqueue.o.d -nostdinc -I./arch/sparc/include -I./arch/sparc/include/generated -I./include -I./arch/sparc/include/uapi -I./arch/sparc/include/generated/uapi -I./include/uapi -I./include/generated/uapi -include ./include/linux/compiler-version.h -include ./include/linux/kconfig.h -include ./include/linux/compiler_types.h -D__KERNEL__ -Werror -fmacro-prefix-map=./= -Wall -Wundef -Werror=strict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common -fshort-wchar -fno-PIE -Werror=implicit-function-declaration -Werror=implicit-int -Werror=return-type -Wno-format-security -funsigned-char -std=gnu11 -m32 -mcpu=v8 -pipe -mno-fpu -fcall-used-g5 -fcall-used-g7 -Wa,-Av8 -fno-delete-null-pointer-checks -Wno-frame-address -Wno-format-truncation -Wno-format-overflow -Wno-address-of-packed-member -O2 -fno-allow-store-data-races -Wframe-larger-than=1024 -fno-stack-protector -Wno-main -Wno-unused-but-set-variable -Wno-unused-const-variable -Wno-dangling-p
ointer -fomit-frame-pointer -ftrivial-auto-var-init=zero -fno-stack-clash-protection -Wdeclaration-after-statement -Wvla -Wno-pointer-sign -Wcast-function-type -Wno-stringop-truncation -Wno-stringop-overflow -Wno-restrict -Wno-maybe-uninitialized -Wno-array-bounds -Wno-alloc-size-larger-than -Wimplicit-fallthrough=5 -fno-strict-overflow -fno-stack-check -fconserve-stack -Werror=date-time -Werror=incompatible-pointer-types -Werror=designated-init -Wno-packed-not-aligned -DKBUILD_MODFILE='"kernel/workqueue"' -DKBUILD_BASENAME='"workqueue"' -DKBUILD_MODNAME='"workqueue"' -D__KBUILD_MODNAME=kmod_workqueue -c -o kernel/workqueue.o kernel/workqueue.c
kernel/workqueue.c: In function 'get_work_pwq':
kernel/workqueue.c:713:24: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
713 | return (void *)(data & WORK_STRUCT_WQ_DATA_MASK);
| ^
kernel/workqueue.c: In function 'get_work_pool':
kernel/workqueue.c:741:25: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
741 | return ((struct pool_workqueue *)
| ^
kernel/workqueue.c: In function 'get_work_pool_id':
kernel/workqueue.c:763:25: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
763 | return ((struct pool_workqueue *)
| ^
cc1: all warnings being treated as errors
make[2]: *** [scripts/Makefile.build:252: kernel/workqueue.o] Error 1
make[1]: *** [scripts/Makefile.build:494: kernel] Error 2
make: *** [Makefile:2026: .] Error 2

with top of tree GCC 14 and even came up with a fix, but I can see the
issue is already being handled with an identical change from Arnd here:
<https://lore.kernel.org/r/[email protected]/>, so
there's nothing to do about it; let's just wait for 6.5.

Changes from v1:

- Wrap `dma_spin_lock' into HAS_DMA to fix a "'dma_spin_lock' defined but
not used" warning. I think it's slightly cleaner this way compared to
wrapping the whole part into CONFIG_SPARC64, as this better reflects the
structure of the dependencies here. It also follows the structure of
parport_pc.c, which likewise has `parport_pc_fifo_write_block_dma' put
into CONFIG_PARPORT_PC_FIFO and then HAS_DMA conditionals.

- Add Reported-by tag.
---
arch/sparc/include/asm/parport.h | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)

linux-sparc32-parport-pc.diff
Index: linux-macro/arch/sparc/include/asm/parport.h
===================================================================
--- linux-macro.orig/arch/sparc/include/asm/parport.h
+++ linux-macro/arch/sparc/include/asm/parport.h
@@ -19,9 +19,11 @@
* While sparc64 doesn't have an ISA DMA API, we provide something that looks
* close enough to make parport_pc happy
*/
+#ifdef CONFIG_SPARC64
#define HAS_DMA
+#endif

-#ifdef CONFIG_PARPORT_PC_FIFO
+#if defined(CONFIG_PARPORT_PC_FIFO) && defined(HAS_DMA)
static DEFINE_SPINLOCK(dma_spin_lock);

#define claim_dma_lock() \
@@ -249,7 +251,8 @@ static struct platform_driver ecpp_drive

static int parport_pc_find_nonpci_ports(int autoirq, int autodma)
{
- return platform_driver_register(&ecpp_driver);
+ return (IS_ENABLED(CONFIG_SPARC64) &&
+ platform_driver_register(&ecpp_driver));
}

#endif /* !(_ASM_SPARC64_PARPORT_H */


2023-06-19 01:43:13

by Randy Dunlap

[permalink] [raw]
Subject: Re: [PATCH v2] SPARC: Fix parport_pc support for 32-bit platforms



On 6/18/23 18:01, Maciej W. Rozycki wrote:
> For 32-bit SPARC platforms PC-style parallel ports are only available as
> PCI options. Adjust <asm/parport.h> accordingly, fixing build errors:
>
> sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `ecpp_remove':
> parport_pc.c:(.text+0x8f0): undefined reference to `ebus_dma_irq_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x8f8): undefined reference to `ebus_dma_unregister'
> sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `parport_pc_fifo_write_block_dma':
> parport_pc.c:(.text+0x1430): undefined reference to `ebus_dma_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1444): undefined reference to `ebus_dma_prepare'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x14e4): undefined reference to `ebus_dma_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x14f4): undefined reference to `ebus_dma_request'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1584): undefined reference to `ebus_dma_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1594): undefined reference to `ebus_dma_residue'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1608): undefined reference to `ebus_dma_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1618): undefined reference to `ebus_dma_residue'
> sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `ecpp_probe':
> parport_pc.c:(.text+0x33a0): undefined reference to `ebus_dma_register'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x33b8): undefined reference to `ebus_dma_irq_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x34b0): undefined reference to `ebus_dma_irq_enable'
> sparc-linux-gnu-ld: parport_pc.c:(.text+0x34bc): undefined reference to `ebus_dma_unregister'
>
> Reported-by: Randy Dunlap <[email protected]>
> Closes: https://lore.kernel.org/r/[email protected]/
> Signed-off-by: Maciej W. Rozycki <[email protected]>
> Fixes: 66bcd06099bb ("parport_pc: Also enable driver for PCI systems")
> Cc: [email protected] # v5.18+

Acked-by: Randy Dunlap <[email protected]>
Tested-by: Randy Dunlap <[email protected]> # build-tested

Thanks for fixing this.

> ---

> Changes from v1:
>
> - Wrap `dma_spin_lock' into HAS_DMA to fix a "'dma_spin_lock' defined but
> not used" warning. I think it's slightly cleaner this way compared to
> wrapping the whole part into CONFIG_SPARC64, as this better reflects the
> structure of the dependencies here. It also follows the structure of
> parport_pc.c, which likewise has `parport_pc_fifo_write_block_dma' put
> into CONFIG_PARPORT_PC_FIFO and then HAS_DMA conditionals.
>
> - Add Reported-by tag.
> ---
> arch/sparc/include/asm/parport.h | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> linux-sparc32-parport-pc.diff
> Index: linux-macro/arch/sparc/include/asm/parport.h
> ===================================================================
> --- linux-macro.orig/arch/sparc/include/asm/parport.h
> +++ linux-macro/arch/sparc/include/asm/parport.h
> @@ -19,9 +19,11 @@
> * While sparc64 doesn't have an ISA DMA API, we provide something that looks
> * close enough to make parport_pc happy
> */
> +#ifdef CONFIG_SPARC64
> #define HAS_DMA
> +#endif
>
> -#ifdef CONFIG_PARPORT_PC_FIFO
> +#if defined(CONFIG_PARPORT_PC_FIFO) && defined(HAS_DMA)
> static DEFINE_SPINLOCK(dma_spin_lock);
>
> #define claim_dma_lock() \
> @@ -249,7 +251,8 @@ static struct platform_driver ecpp_drive
>
> static int parport_pc_find_nonpci_ports(int autoirq, int autodma)
> {
> - return platform_driver_register(&ecpp_driver);
> + return (IS_ENABLED(CONFIG_SPARC64) &&
> + platform_driver_register(&ecpp_driver));
> }
>
> #endif /* !(_ASM_SPARC64_PARPORT_H */

--
~Randy

2023-08-08 20:34:54

by Randy Dunlap

[permalink] [raw]
Subject: Re: [PATCH v2] SPARC: Fix parport_pc support for 32-bit platforms

Hi,

Is there some way to have this build fix merged?
Thanks.


On 6/18/23 18:08, Randy Dunlap wrote:
>
>
> On 6/18/23 18:01, Maciej W. Rozycki wrote:
>> For 32-bit SPARC platforms PC-style parallel ports are only available as
>> PCI options. Adjust <asm/parport.h> accordingly, fixing build errors:
>>
>> sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `ecpp_remove':
>> parport_pc.c:(.text+0x8f0): undefined reference to `ebus_dma_irq_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x8f8): undefined reference to `ebus_dma_unregister'
>> sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `parport_pc_fifo_write_block_dma':
>> parport_pc.c:(.text+0x1430): undefined reference to `ebus_dma_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1444): undefined reference to `ebus_dma_prepare'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x14e4): undefined reference to `ebus_dma_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x14f4): undefined reference to `ebus_dma_request'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1584): undefined reference to `ebus_dma_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1594): undefined reference to `ebus_dma_residue'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1608): undefined reference to `ebus_dma_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x1618): undefined reference to `ebus_dma_residue'
>> sparc-linux-gnu-ld: drivers/parport/parport_pc.o: in function `ecpp_probe':
>> parport_pc.c:(.text+0x33a0): undefined reference to `ebus_dma_register'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x33b8): undefined reference to `ebus_dma_irq_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x34b0): undefined reference to `ebus_dma_irq_enable'
>> sparc-linux-gnu-ld: parport_pc.c:(.text+0x34bc): undefined reference to `ebus_dma_unregister'
>>
>> Reported-by: Randy Dunlap <[email protected]>
>> Closes: https://lore.kernel.org/r/[email protected]/
>> Signed-off-by: Maciej W. Rozycki <[email protected]>
>> Fixes: 66bcd06099bb ("parport_pc: Also enable driver for PCI systems")
>> Cc: [email protected] # v5.18+
>
> Acked-by: Randy Dunlap <[email protected]>
> Tested-by: Randy Dunlap <[email protected]> # build-tested
>
> Thanks for fixing this.
>
>> ---
>
>> Changes from v1:
>>
>> - Wrap `dma_spin_lock' into HAS_DMA to fix a "'dma_spin_lock' defined but
>> not used" warning. I think it's slightly cleaner this way compared to
>> wrapping the whole part into CONFIG_SPARC64, as this better reflects the
>> structure of the dependencies here. It also follows the structure of
>> parport_pc.c, which likewise has `parport_pc_fifo_write_block_dma' put
>> into CONFIG_PARPORT_PC_FIFO and then HAS_DMA conditionals.
>>
>> - Add Reported-by tag.
>> ---
>> arch/sparc/include/asm/parport.h | 7 +++++--
>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>
>> linux-sparc32-parport-pc.diff
>> Index: linux-macro/arch/sparc/include/asm/parport.h
>> ===================================================================
>> --- linux-macro.orig/arch/sparc/include/asm/parport.h
>> +++ linux-macro/arch/sparc/include/asm/parport.h
>> @@ -19,9 +19,11 @@
>> * While sparc64 doesn't have an ISA DMA API, we provide something that looks
>> * close enough to make parport_pc happy
>> */
>> +#ifdef CONFIG_SPARC64
>> #define HAS_DMA
>> +#endif
>>
>> -#ifdef CONFIG_PARPORT_PC_FIFO
>> +#if defined(CONFIG_PARPORT_PC_FIFO) && defined(HAS_DMA)
>> static DEFINE_SPINLOCK(dma_spin_lock);
>>
>> #define claim_dma_lock() \
>> @@ -249,7 +251,8 @@ static struct platform_driver ecpp_drive
>>
>> static int parport_pc_find_nonpci_ports(int autoirq, int autodma)
>> {
>> - return platform_driver_register(&ecpp_driver);
>> + return (IS_ENABLED(CONFIG_SPARC64) &&
>> + platform_driver_register(&ecpp_driver));
>> }
>>
>> #endif /* !(_ASM_SPARC64_PARPORT_H */
>

--
~Randy