Skip to contentSkip to conversation

Drop IRDA ISA dependency

PATCH In review Andi Kleen 4 participants 7 messages Oct 29, 2004 – Oct 30, 2004

Subsystems (2)
lore
Share & tools
lore

Andi Kleen

author
#0 lore raw

PATCH 1/1 Drop IRDA ISA dependency

Make IRDA devices are not really ISA devices not depend on CONFIG_ISA. This allows to use them on x86-64

drivers/net/irda/Kconfig +5 -5
diff -urpN -X ../KDIFX linux/drivers/net/irda/Kconfig linux-2.6.8-amd64/drivers/net/irda/Kconfig
@@ -310,7 +310,7 @@ config SIGMATEL_FIR
 
 config NSC_FIR
 	tristate "NSC PC87108/PC87338"
-	depends on IRDA && ISA
+	depends on IRDA
 	help
 	  Say Y here if you want to build support for the NSC PC87108 and
 	  PC87338 IrDA chipsets.  This driver supports SIR,
@@ -321,7 +321,7 @@ config NSC_FIR
 
 config WINBOND_FIR
 	tristate "Winbond W83977AF (IR)"
-	depends on IRDA && ISA
+	depends on IRDA
 	help
 	  Say Y here if you want to build IrDA support for the Winbond
 	  W83977AF super-io chipset.  This driver should be used for the IrDA
@@ -347,7 +347,7 @@ config AU1000_FIR
 
 config SMC_IRCC_FIR
 	tristate "SMSC IrCC (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && IRDA && ISA
+	depends on EXPERIMENTAL && IRDA
 	help
 	  Say Y here if you want to build support for the SMC Infrared
 	  Communications Controller.  It is used in a wide variety of
@@ -357,7 +357,7 @@ config SMC_IRCC_FIR
 
 config ALI_FIR
 	tristate "ALi M5123 FIR (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && IRDA && ISA
+	depends on EXPERIMENTAL && IRDA
 	help
 	  Say Y here if you want to build support for the ALi M5123 FIR
 	  Controller.  The ALi M5123 FIR Controller is embedded in ALi M1543C,
@@ -385,7 +385,7 @@ config SA1100_FIR
 
 config VIA_FIR
 	tristate "VIA VT8231/VT1211 SIR/MIR/FIR"
-	depends on IRDA && ISA && PCI
+	depends on IRDA
 	help
 	  Say Y here if you want to build support for the VIA VT8231
 	  and VIA VT1211 IrDA controllers, found on the motherboards using
include/net/irda/irda_device.h +0 -2
diff -urpN -X ../KDIFX linux/include/net/irda/irda_device.h linux-2.6.8-amd64/include/net/irda/irda_device.h
@@ -237,9 +237,7 @@ int  irda_device_register_dongle(struct 
 dongle_t *irda_device_dongle_init(struct net_device *dev, int type);
 int irda_device_dongle_cleanup(dongle_t *dongle);
 
-#ifdef CONFIG_ISA
 void irda_setup_dma(int channel, dma_addr_t buffer, int count, int mode);
-#endif
 
 void irda_task_delete(struct irda_task *task);
 struct irda_task *irda_task_execute(void *instance, 
net/irda/irda_device.c +1 -3
diff -urpN -X ../KDIFX linux/net/irda/irda_device.c linux-2.6.8-amd64/net/irda/irda_device.c
@@ -529,11 +529,10 @@ int irda_device_set_mode(struct net_devi
 	return ret;
 }
 
-#ifdef CONFIG_ISA
 /*
  * Function setup_dma (idev, buffer, count, mode)
  *
- *    Setup the DMA channel. Commonly used by ISA FIR drivers
+ *    Setup the DMA channel. Commonly used by LPC FIR drivers
  *
  */
 void irda_setup_dma(int channel, dma_addr_t buffer, int count, int mode)
@@ -552,4 +551,3 @@ void irda_setup_dma(int channel, dma_add
 	release_dma_lock(flags);
 }
 EXPORT_SYMBOL(irda_setup_dma);
-#endif
#1 lore raw
Quoted 3 lines · Andi Kleen

Make IRDA devices are not really ISA devices not depend on CONFIG_ISA. This allows to use them on x86-64

but this is bogus. If it's using isa-style DMA it needs CONFIG_ISA.

author
#2 lore raw
Quoted 6 lines · Christoph Hellwig
Quoted 3 lines · Andi Kleen

Make IRDA devices are not really ISA devices not depend on CONFIG_ISA. This allows to use them on x86-64

but this is bogus. If it's using isa-style DMA it needs CONFIG_ISA.

No it doesn't. They work just fine with the patch applied on x86-64.

-Andi

Quoted 1 line
#3 lore raw
Quoted 9 lines · Andi Kleen
Quoted 6 lines · Christoph Hellwig
Quoted 3 lines · Andi Kleen

Make IRDA devices are not really ISA devices not depend on CONFIG_ISA. This allows to use them on x86-64

but this is bogus. If it's using isa-style DMA it needs CONFIG_ISA.

No it doesn't. They work just fine with the patch applied on x86-64.

It doesn't on various other architectures.

#4 lore raw
In reply to Christoph Hellwig
Quoted 1 line · Christoph Hellwig

but this is bogus. If it's using isa-style DMA it needs CONFIG_ISA.

Sounds like there is some confusion over what CONFIG_ISA means. I always understood it to mean 'We have ISA slots on this architecture' regardless of whether theres an ISA style LPC bus.
Its a means of disabling a whole slew of drivers that have no meaning on a particular platform (in Andi's case, x86-64).

Or did I get confused ?

Dave

author
#5 lore raw
Quoted 9 lines · Dave Jones
Quoted 1 line · Christoph Hellwig

but this is bogus. If it's using isa-style DMA it needs CONFIG_ISA.

Sounds like there is some confusion over what CONFIG_ISA means. I always understood it to mean 'We have ISA slots on this architecture' regardless of whether theres an ISA style LPC bus.
Its a means of disabling a whole slew of drivers that have no meaning on a particular platform (in Andi's case, x86-64).

That's exactly my understanding too.

Quoted 1 line

Or did I get confused ?

I don't think so.

-Andi

#6 lore raw
In reply to Dave Jones
Quoted 1 line · Dave Jones

Sounds like there is some confusion over what CONFIG_ISA means.

i think it was vague / different before maybe?

Quoted 2 lines

I always understood it to mean 'We have ISA slots on this architecture' regardless of whether theres an ISA style LPC bus.

or pc card --- but this got separate recently didn't it?

Quoted 2 lines

Its a means of disabling a whole slew of drivers that have no meaning on a particular platform (in Andi's case, x86-64).

maybe we want CONFIG_BUS_ISA and CONFIG_SLOT_ISA, the former meaning we have bus/semantics the latter depends on the former and means physical slots --- which would be used to prune the driver selection choices?