I still have no LPC dongle, but I am getting closer. I really liked this as it shows off the power of our new config system, the dts. There is a whole function in v2 to make com2 operate at the com1 address. For v3, we just set a value in the mainboard dts and we're done (in theory ; I can't tell if this works). DBE62 is getting through a lot of POST btw but hangs at 13.
I just realized this exampel also shows that you do NOT want all dts values to be visible in Kconfig. It would make no sense to change these values on the dbe62.
Signed-off-by: Ronald G. Minnich rminnich@gmail.com
Index: mainboard/artecgroup/dbe62/dts =================================================================== --- mainboard/artecgroup/dbe62/dts (revision 630) +++ mainboard/artecgroup/dbe62/dts (working copy) @@ -43,6 +43,10 @@ /* GPIO(0-0x20) for INT D:C:B:A, 0xFF=none. * See virtual PIC spec. */ enable_gpio_int_route = "0x0D0C0700"; + /* we use com2 since that is on the dongle */ + com2_enable = "1"; + /* Set com2 address to be COM1 */ + com2_address = "0x3f8"; }; }; };
On 05.03.2008 06:51, ron minnich wrote:
I still have no LPC dongle, but I am getting closer. I really liked this as it shows off the power of our new config system, the dts. There is a whole function in v2 to make com2 operate at the com1 address. For v3, we just set a value in the mainboard dts and we're done (in theory ; I can't tell if this works). DBE62 is getting through a lot of POST btw but hangs at 13.
I just realized this exampel also shows that you do NOT want all dts values to be visible in Kconfig. It would make no sense to change these values on the dbe62.
Signed-off-by: Ronald G. Minnich rminnich@gmail.com
Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
Index: mainboard/artecgroup/dbe62/dts
--- mainboard/artecgroup/dbe62/dts (revision 630) +++ mainboard/artecgroup/dbe62/dts (working copy) @@ -43,6 +43,10 @@ /* GPIO(0-0x20) for INT D:C:B:A, 0xFF=none. * See virtual PIC spec. */ enable_gpio_int_route = "0x0D0C0700";
/* we use com2 since that is on the dongle */
Uppercase the W in "we"?
com2_enable = "1";
/* Set com2 address to be COM1 */
com2_address = "0x3f8"; }; };
};
If you look at the dbe61 dts, you can see the same settings except that dbe61 additionally sets com2_irq = "4".
Regards, Carl-Daniel
On Wed, Mar 5, 2008 at 5:22 AM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
If you look at the dbe61 dts, you can see the same settings except that dbe61 additionally sets com2_irq = "4".
and there is a good question. I might have misread the v2 code, I will have to go back and check.
But will com2 still interrupt on irq 3 no matter what? I'm never sure how things go on this chipset.
ron
On 05.03.2008 17:01, ron minnich wrote:
On Wed, Mar 5, 2008 at 5:22 AM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
If you look at the dbe61 dts, you can see the same settings except that dbe61 additionally sets com2_irq = "4".
and there is a good question. I might have misread the v2 code, I will have to go back and check.
But will com2 still interrupt on irq 3 no matter what? I'm never sure how things go on this chipset.
southbridge/amd/cs5536/cs5536.c: msr.lo |= sb->com2_irq << 28; southbridge/amd/cs5536/dts: com2_irq = "3";
The default is irq 3, can be overridden in the mainboard dts.
Regards, Carl-Daniel
ron minnich wrote:
On Wed, Mar 5, 2008 at 5:22 AM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
If you look at the dbe61 dts, you can see the same settings except that dbe61 additionally sets com2_irq = "4".
and there is a good question. I might have misread the v2 code, I will have to go back and check.
But will com2 still interrupt on irq 3 no matter what? I'm never sure how things go on this chipset.
ron
There is no chipset default. Each piece needs to be setup. If you didn't set it in the mainboard dts it will use the chipset dts default. Marc
ron minnich wrote:
Index: mainboard/artecgroup/dbe62/dts
--- mainboard/artecgroup/dbe62/dts (revision 630) +++ mainboard/artecgroup/dbe62/dts (working copy) @@ -43,6 +43,10 @@ /* GPIO(0-0x20) for INT D:C:B:A, 0xFF=none. * See virtual PIC spec. */ enable_gpio_int_route = "0x0D0C0700";
/* we use com2 since that is on the dongle */
com2_enable = "1";
/* Set com2 address to be COM1 */
com2_address = "0x3f8";
Can't we use 2f8 as base address here and put console on COM2?
Reordering com ports is kind of evil. It will also change what is ttyS0 in Linux..
Stefan
On Wed, Mar 5, 2008 at 3:03 PM, Stefan Reinauer stepan@coresystems.de wrote:
Can't we use 2f8 as base address here and put console on COM2?
Reordering com ports is kind of evil. It will also change what is ttyS0 in Linux..
actually it won't. All I'm doing is making the one available com port that is connected BE COM1 -- base address, irq, everything. Linux won't know. Linux will use ttyS0 and as far as it is concerned, nothing is out of the ordinary. It all looks like COM2. COM1 is not even connected to anything on this platform.
I looked at both ways and this appears (to me) to be the best. It's also what is done on v2.
ron
ron minnich wrote:
On Wed, Mar 5, 2008 at 3:03 PM, Stefan Reinauer stepan@coresystems.de wrote:
Can't we use 2f8 as base address here and put console on COM2?
Reordering com ports is kind of evil. It will also change what is ttyS0 in Linux..
actually it won't. All I'm doing is making the one available com port that is connected BE COM1 -- base address, irq, everything. Linux won't know. Linux will use ttyS0 and as far as it is concerned, nothing is out of the ordinary. It all looks like COM2. COM1 is not even connected to anything on this platform.
Ah ok, that makes perfect sense.