attached.
ron
On 11.12.2007 02:38, ron minnich wrote:
Add a nice pretty banner printer so we have standard banners.
Maybe make the banner a little bit shorter so bigger strings will fit without wrapping?
Make 'halt' just put out nulls forever instead of hlt(), to ensure that any fifos of any sort get flushed. This change also helps JTAG debuggers since it can be very hard to get out of a halt instruction, but very easy to get out of an IO loop. I have direct experience with this latter situation :-)
Signed-off-by: Ronald G. Minnich rminnich@gmail.com
Using banner() in halt() would flush the fifo as well. Can we use inb(0x80) in the loop instead? That way, we still have an IO loop, but that loop will not flood the receiver. Plus, it will work much more nicely with my soon to come printk buffer patches.
Regards, Carl-Daniel
On Dec 12, 2007 4:42 PM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
On 11.12.2007 02:38, ron minnich wrote:
Add a nice pretty banner printer so we have standard banners.
Maybe make the banner a little bit shorter so bigger strings will fit without wrapping?
Sure! what's a good length? 20 chars?
Make 'halt' just put out nulls forever instead of hlt(), to ensure that any fifos of any sort get flushed. This change also helps JTAG debuggers since it can be very hard to get out of a halt instruction, but very easy to get out of an IO loop. I have direct experience with this latter situation :-)
Signed-off-by: Ronald G. Minnich rminnich@gmail.com
Using banner() in halt() would flush the fifo as well. Can we use inb(0x80) in the loop instead? That way, we still have an IO loop, but that loop will not flood the receiver. Plus, it will work much more nicely with my soon to come printk buffer patches.
well, is the banner enough to flush all the fifo's? I thought they got pretty big nowadays. Anyone know?
Hey, carl-daniel, what's a good payload for testing my 'payload won't work' problem? Also, I wonder if it is because the payload runs above 0x100000?
ron
On Wed, Dec 12, 2007 at 04:46:32PM -0800, ron minnich wrote:
well, is the banner enough to flush all the fifo's? I thought they got pretty big nowadays. Anyone know?
16550 and compatible still only have 14/16/18/whatever bytes. EHCI debug port is unbuffered. .. are there more?
//Peter
On Dec 12, 2007 4:52 PM, Peter Stuge peter@stuge.se wrote:
On Wed, Dec 12, 2007 at 04:46:32PM -0800, ron minnich wrote:
well, is the banner enough to flush all the fifo's? I thought they got pretty big nowadays. Anyone know?
16550 and compatible still only have 14/16/18/whatever bytes. EHCI debug port is unbuffered.
then' I'll change die to do a banner and loop on in on port 80
ron
On 13.12.2007 01:46, ron minnich wrote:
Hey, carl-daniel, what's a good payload for testing my 'payload won't work' problem? Also, I wonder if it is because the payload runs above 0x100000?
Sorry, Ron, I have no idea. Maybe we should create a "hello world" payload. Memtest is probably too complicated.
Regards, Carl-Daniel
On Dec 12, 2007 8:53 PM, Carl-Daniel Hailfinger < c-d.hailfinger.devel.2006@gmx.net> wrote:
On 13.12.2007 01:46, ron minnich wrote:
Hey, carl-daniel, what's a good payload for testing my 'payload won't work' problem? Also, I wonder if it is because the payload runs above 0x100000?
Sorry, Ron, I have no idea. Maybe we should create a "hello world" payload. Memtest is probably too complicated.
Regards, Carl-Daniel
Would the "baremetal" in LBv1 be helpful? I saw that there was a "Hello, world" example payload, but I don't know if we can still use it.
-Corey
-
Carl-Daniel Hailfinger -
Corey Osgood -
Peter Stuge -
ron minnich