Skip to contentSkip to conversation

4xx change to core files

DISCUSSION Armin 3 participants 5 messages May 8, 2002

lore
Share & tools
lore

Armin

author
#0 lore raw

I have changed the way the paddrs are used in the 4xx specifically in the manner they are used in the core.c files. This eliminates the need for "externs struct *_ADDR[]" , IIC_PORT_DFNS and in most cases the need for *_NUMS. The ocp API's have changed to take advantage of this new core_ocp[]. All the ocp drivers have been affected in a minor way as well as most core.h files.

The core_ocp[] is don ein the following:

new struct in asm-ppc/ocp.h

struct ocp_def {
	enum ocp_type type;
	int paddr;
	int irq;
};

each 4xx core file now contains something simulary to this ( ibm405gp.c)

struct ocp_def core_ocp[] = {
	{UART, UART0_IO_BASE, UART0_INT},
	{UART, UART1_IO_BASE, UART1_INT},
	{IIC, IIC0_BASE, IIC0_IRQ},
	{GPIO, GPIO0_BASE, OCP_IRQ_NA},
	{PCI, PCIL0_BASE, OCP_IRQ_NA},
	{OPB, OPB0_BASE, OCP_IRQ_NA},
	{EMAC, EMAC0_BASE, OCP_IRQ_NA},
	{OCP_NULL_TYPE, 0x0, OCP_IRQ_NA},

};

new ocp APIs:

unsigned long get_ocp_paddr(int type, int dev_num);
returns the physical address for a given ocp type for the nth one. this is used when the mmu is not completely up such as during pci
bring up.

int ocp_get_max(int type);
returns the max number of ocp device type

This affects the 405gp, np405l, np405h, 440gp , stb03xxx, stb04xxx and
np4gs cores.

the following drivers have been altered, ibm_ocp_enet.c, ibm_ocp_zmii.c, ppc405_pci.c, ibm_ocp_gpio.c, i2c-adapt-ibm_ocp.c

I expect to be done soon.

Armin

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

#1 lore raw
Quoted 9 lines · Armin

The core_ocp[] is don ein the following:

new struct in asm-ppc/ocp.h

struct ocp_def {
	enum ocp_type type;
	int paddr;
	int irq;
};

};

<snip>

Quoted 6 lines

new ocp APIs:

unsigned long get_ocp_paddr(int type, int dev_num);
returns the physical address for a given ocp type for the nth one. this is used when the mmu is not completely up such as during pci
bring up.

It would be helpful for 36-bit 4xx core implementations (440gp/440gx) if the paddr used the phys_addr_t typedef so we could store a native 64-bit address.

It would appear to me that one would expect to be able to do the following:

	ioremap(get_ocp_paddr(<type>, <num>), <size>);

Is that the intention?

--
Matt Porter
MontaVista Software, Inc.
[email protected]

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

#4 lore raw
In reply to Armin
Quoted 11 lines · Armin
struct ocp_def core_ocp[] = {
	{UART, UART0_IO_BASE, UART0_INT},
	{UART, UART1_IO_BASE, UART1_INT},
	{IIC, IIC0_BASE, IIC0_IRQ},
	{GPIO, GPIO0_BASE, OCP_IRQ_NA},
	{PCI, PCIL0_BASE, OCP_IRQ_NA},
	{OPB, OPB0_BASE, OCP_IRQ_NA},
	{EMAC, EMAC0_BASE, OCP_IRQ_NA},
	{OCP_NULL_TYPE, 0x0, OCP_IRQ_NA},

};

It would be nice if the global struct ocp_def core_ocp was just a pointer. Then in a ibm405gp.c and friends could have a static of the entire stucture and set the global pointer to it.

That should be a pretty straight forward change. The only change to code that use core_ocp would be maybe a NULL check, but even that could be avoided.

The great thing about these chips is that you can stamp down different OCP's as needed regaurdless of the actual CPU type.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

author
#2 lore raw
In reply to Matt Porter
Quoted 24 lines · Matt Porter
Quoted 10 lines · Armin

The core_ocp[] is don ein the following:

new struct in asm-ppc/ocp.h

struct ocp_def {
	enum ocp_type type;
	int paddr;
	int irq;
};

};

<snip>

Quoted 7 lines

new ocp APIs:

unsigned long get_ocp_paddr(int type, int dev_num);
returns the physical address for a given ocp type for the nth one. this is used when the mmu is not completely up such as during pci
bring up.

Quoted 3 lines

It would be helpful for 36-bit 4xx core implementations (440gp/440gx) if the paddr used the phys_addr_t typedef so we could store a native 64-bit address.

ok

Quoted 7 lines

It would appear to me that one would expect to be able to do the following:

	ioremap(get_ocp_paddr(<type>, <num>), <size>);

Is that the intention?

Yeap

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

#3 lore raw
Quoted 35 lines · Armin
Quoted 24 lines · Matt Porter
Quoted 10 lines · Armin

The core_ocp[] is don ein the following:

new struct in asm-ppc/ocp.h

struct ocp_def {
	enum ocp_type type;
	int paddr;
	int irq;
};

};

<snip>

Quoted 7 lines

new ocp APIs:

unsigned long get_ocp_paddr(int type, int dev_num); returns the physical address for a given ocp type for the nth one. this is used when the mmu is not completely up such as during pci bring up.

Quoted 3 lines

It would be helpful for 36-bit 4xx core implementations (440gp/440gx) if the paddr used the phys_addr_t typedef so we could store a native 64-bit address.

ok

Quoted 4 lines

It would appear to me that one would expect to be able to do the following:

Quoted 6 lines
Quoted 3 lines
	ioremap(get_ocp_paddr(<type>, <num>), <size>);

Is that the intention?

Yeap

Ok, then the other piece that will be necessary is to change to using a 'ioremap_native' for the ocp drivers. That will resolve to 'ioremap' on 32-bit phys cores and 'ioremap64' on 36+-bit phys cores.

I'm able to "fixup" 32-bit phys addrs in the 440-specific ioremap on the 440gp since the memory map luckily provided unique (least significant 32-bit) addresses. This is may or may not work on upcoming 440+ implementations as the memory maps will differ...possibly having I/O regions overlapping in their least significant 32-bits.

Regards,
--
Matt Porter
MontaVista Software, Inc.
[email protected]

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/