[v5] sh: avoid using IRQ0 on SH3/4

  Now that IRQ0 is no longer returned by platform_get_irq() and its ilk (they
now return -EINVAL instead).  However, the code supporting SH3/4 SoCs still
maps the IRQ #s starting at 0 -- modify that code to start the IRQ #s from
16 instead.

The patch should mostly affect the AP-SH4A-3A/AP-SH4AD-0A boards as they
indeed use IRQ0 for the SMSC911x compatible Ethernet chip...

Fixes: ce753ad1549c ("platform: finally disallow IRQ0 in platform_get_irq() and its ilk")
Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>

---
The patch is against Linus Torvalds' 'linux.git' repo.

Changes in version 5:
- updated the patch description and the "Fixes:" tag as the patch disallowing
  the use of IRQ0 was merged meanwhile.

Changes in version 4:
- fixed up the off-chip base IRQ #s for the Dreamcast/Highlander/R2D/SE7724
  boards.

Changes in version 3:
- added an appropriate Fixes: tag and added a passage about it to the patch
  description;
- added actual cases of the boards using IRQ0 to the patch description;
- added Geert Uytterhoeven's and John Paul Adrian Glaubitz's tags;
- updated the link to point to the version 2 of the patch.

Changes in version 2:
- changed cmp/ge to cmp/hs in the assembly code.

 arch/sh/include/mach-common/mach/highlander.h |    2 +-
 arch/sh/include/mach-common/mach/r2d.h        |    2 +-
 arch/sh/include/mach-dreamcast/mach/sysasic.h |    2 +-
 arch/sh/include/mach-se/mach/se7724.h         |    2 +-
 arch/sh/kernel/cpu/sh3/entry.S                |    4 ++--
 include/linux/sh_intc.h                       |    6 +++---
 6 files changed, 9 insertions(+), 9 deletions(-)
  

Hi Sergey!

On Wed, 2023-05-31 at 23:34 +0300, Sergey Shtylyov wrote:
> Now that IRQ0 is no longer returned by platform_get_irq() and its ilk (they
> now return -EINVAL instead).  However, the code supporting SH3/4 SoCs still
> maps the IRQ #s starting at 0 -- modify that code to start the IRQ #s from
> 16 instead.
> 
> The patch should mostly affect the AP-SH4A-3A/AP-SH4AD-0A boards as they
> indeed use IRQ0 for the SMSC911x compatible Ethernet chip...

I'm not sure I understand. Does that mean that the Ethernet controller on
the AP-SH4A-3A/AP-SH4AD-0A boards will no longer work after this patch
has been applied?

Adrian

> 
> Fixes: ce753ad1549c ("platform: finally disallow IRQ0 in platform_get_irq() and its ilk")
> Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
> Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
> Tested-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
> 
> ---
> The patch is against Linus Torvalds' 'linux.git' repo.
> 
> Changes in version 5:
> - updated the patch description and the "Fixes:" tag as the patch disallowing
>   the use of IRQ0 was merged meanwhile.
> 
> Changes in version 4:
> - fixed up the off-chip base IRQ #s for the Dreamcast/Highlander/R2D/SE7724
>   boards.
> 
> Changes in version 3:
> - added an appropriate Fixes: tag and added a passage about it to the patch
>   description;
> - added actual cases of the boards using IRQ0 to the patch description;
> - added Geert Uytterhoeven's and John Paul Adrian Glaubitz's tags;
> - updated the link to point to the version 2 of the patch.
> 
> Changes in version 2:
> - changed cmp/ge to cmp/hs in the assembly code.
> 
>  arch/sh/include/mach-common/mach/highlander.h |    2 +-
>  arch/sh/include/mach-common/mach/r2d.h        |    2 +-
>  arch/sh/include/mach-dreamcast/mach/sysasic.h |    2 +-
>  arch/sh/include/mach-se/mach/se7724.h         |    2 +-
>  arch/sh/kernel/cpu/sh3/entry.S                |    4 ++--
>  include/linux/sh_intc.h                       |    6 +++---
>  6 files changed, 9 insertions(+), 9 deletions(-)
> 
> Index: linux/arch/sh/include/mach-common/mach/highlander.h
> ===================================================================
> --- linux.orig/arch/sh/include/mach-common/mach/highlander.h
> +++ linux/arch/sh/include/mach-common/mach/highlander.h
> @@ -176,7 +176,7 @@
>  #define IVDR_CK_ON	4		/* iVDR Clock ON */
>  #endif
>  
> -#define HL_FPGA_IRQ_BASE	200
> +#define HL_FPGA_IRQ_BASE	(200 + 16)
>  #define HL_NR_IRL		15
>  
>  #define IRQ_AX88796		(HL_FPGA_IRQ_BASE + 0)
> Index: linux/arch/sh/include/mach-common/mach/r2d.h
> ===================================================================
> --- linux.orig/arch/sh/include/mach-common/mach/r2d.h
> +++ linux/arch/sh/include/mach-common/mach/r2d.h
> @@ -47,7 +47,7 @@
>  
>  #define IRLCNTR1	(PA_BCR + 0)	/* Interrupt Control Register1 */
>  
> -#define R2D_FPGA_IRQ_BASE	100
> +#define R2D_FPGA_IRQ_BASE	(100 + 16)
>  
>  #define IRQ_VOYAGER		(R2D_FPGA_IRQ_BASE + 0)
>  #define IRQ_EXT			(R2D_FPGA_IRQ_BASE + 1)
> Index: linux/arch/sh/include/mach-dreamcast/mach/sysasic.h
> ===================================================================
> --- linux.orig/arch/sh/include/mach-dreamcast/mach/sysasic.h
> +++ linux/arch/sh/include/mach-dreamcast/mach/sysasic.h
> @@ -22,7 +22,7 @@
>     takes.
>  */
>  
> -#define HW_EVENT_IRQ_BASE  48
> +#define HW_EVENT_IRQ_BASE  (48 + 16)
>  
>  /* IRQ 13 */
>  #define HW_EVENT_VSYNC     (HW_EVENT_IRQ_BASE +  5) /* VSync */
> Index: linux/arch/sh/include/mach-se/mach/se7724.h
> ===================================================================
> --- linux.orig/arch/sh/include/mach-se/mach/se7724.h
> +++ linux/arch/sh/include/mach-se/mach/se7724.h
> @@ -37,7 +37,7 @@
>  #define IRQ2_IRQ        evt2irq(0x640)
>  
>  /* Bits in IRQ012 registers */
> -#define SE7724_FPGA_IRQ_BASE	220
> +#define SE7724_FPGA_IRQ_BASE	(220 + 16)
>  
>  /* IRQ0 */
>  #define IRQ0_BASE	SE7724_FPGA_IRQ_BASE
> Index: linux/arch/sh/kernel/cpu/sh3/entry.S
> ===================================================================
> --- linux.orig/arch/sh/kernel/cpu/sh3/entry.S
> +++ linux/arch/sh/kernel/cpu/sh3/entry.S
> @@ -470,9 +470,9 @@ ENTRY(handle_interrupt)
>  	mov	r4, r0		! save vector->jmp table offset for later
>  
>  	shlr2	r4		! vector to IRQ# conversion
> -	add	#-0x10, r4
>  
> -	cmp/pz	r4		! is it a valid IRQ?
> +	mov	#0x10, r5
> +	cmp/hs	r5, r4		! is it a valid IRQ?
>  	bt	10f
>  
>  	/*
> Index: linux/include/linux/sh_intc.h
> ===================================================================
> --- linux.orig/include/linux/sh_intc.h
> +++ linux/include/linux/sh_intc.h
> @@ -13,9 +13,9 @@
>  /*
>   * Convert back and forth between INTEVT and IRQ values.
>   */
> -#ifdef CONFIG_CPU_HAS_INTEVT
> -#define evt2irq(evt)		(((evt) >> 5) - 16)
> -#define irq2evt(irq)		(((irq) + 16) << 5)
> +#ifdef CONFIG_CPU_HAS_INTEVT	/* Avoid IRQ0 (invalid for platform devices) */
> +#define evt2irq(evt)		((evt) >> 5)
> +#define irq2evt(irq)		((irq) << 5)
>  #else
>  #define evt2irq(evt)		(evt)
>  #define irq2evt(irq)		(irq)
  

On 6/1/23 9:38 AM, John Paul Adrian Glaubitz wrote:
[...]
>> Now that IRQ0 is no longer returned by platform_get_irq() and its ilk (they
>> now return -EINVAL instead).  However, the code supporting SH3/4 SoCs still
>> maps the IRQ #s starting at 0 -- modify that code to start the IRQ #s from
>> 16 instead.

  Now that I'm re-reading this passage, it seems unfinished... :-/

>> The patch should mostly affect the AP-SH4A-3A/AP-SH4AD-0A boards as they
>> indeed use IRQ0 for the SMSC911x compatible Ethernet chip...
> 
> I'm not sure I understand. Does that mean that the Ethernet controller on
> the AP-SH4A-3A/AP-SH4AD-0A boards will no longer work after this patch
> has been applied?

   Contrariwise, they should start working again! :-)
   Here's the story as I understand it:

1. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=965b2aa78fbcb831acf4f669f494da201f4bcace broke IRQ0 use in the SMSC 9111x
driver.

2. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=5ef9b803a4af0f5e42012176889b40bb2a978b18 fixed IRQ0 use in that driver.

3. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=ce753ad1549cbe9ccaea4c06a1f5fa47432c8289 totally stopped returning IRQ0
from platform_get_irq(), thus breaking that driver again.

4. This patch remaps IRQ0 to IRQ16 and thus fixes the SMSC 911x driver again.

[...]

MBR, Sergey