[parisc-linux] sticon/sercon patch for new bootfloppies...

[parisc-linux] sticon/sercon patch for new bootfloppies...

[Helge Deller]

[-- Attachment #1: Type: text/plain, Size: 428 bytes --]

Hi,

the attached patch enables only those consoles, which are given on the linux 
kernel command line.
Examples: "console=tty0" activates STIcon only, "console=ttyS0" activates 
serial console only and if you use both on the commandline you will get the 
system messages on both consoles.

Known bugs: 
- STIcon will not show any messages which happened before the STI
	initialisation

Please test and comment
Greetings,
Helge

[-- Attachment #2: patch --]
[-- Type: text/x-c, Size: 2069 bytes --]

Index: arch/parisc/kernel/pdc_cons.c
===================================================================
RCS file: /home/cvs/parisc/linux/arch/parisc/kernel/pdc_cons.c,v
retrieving revision 1.25
diff -u -r1.25 pdc_cons.c
--- pdc_cons.c	2001/04/06 05:10:54	1.25
+++ pdc_cons.c	2001/06/13 06:08:26
@@ -11,6 +11,10 @@
 #include <asm/system.h>
 #include <asm/pdc.h>	/* for iodc_call() proto and friends */
 
+
+/* Enable to use this simple PDC console (for debugging purposes only!): */
+#undef EARLY_BOOTUP_DEBUG
+
 static void pdc_console_write(struct console *co, const char *s, unsigned count)
 {
 	while(count--)
@@ -46,15 +50,18 @@
 	if (PAGE0->mem_cons.cl_class == CL_DUPLEX)
 		memcpy(&PAGE0->mem_kbd, &PAGE0->mem_cons, sizeof(PAGE0->mem_cons));
 
+#ifdef EARLY_BOOTUP_DEBUG
 	pdc_console_write(0, "PDC Console Initialized\n", 24);
 	/* register the pdc console */
 	register_console(&pdc_cons);
+#endif /* EARLY_BOOTUP_DEBUG */
 }
 
 
 /* Unregister the pdc console with the printk console layer */
 void pdc_console_die(void)
 {
+#ifdef EARLY_BOOTUP_DEBUG
 	extern unsigned long log_size; /* kernel/printk.c */
 
 	printk("Switching from PDC console\n");
@@ -74,6 +81,7 @@
 #endif
 
 	unregister_console(&pdc_cons);
+#endif /* EARLY_BOOTUP_DEBUG */
 }
 
 
Index: arch/parisc/kernel/setup.c
===================================================================
RCS file: /home/cvs/parisc/linux/arch/parisc/kernel/setup.c,v
retrieving revision 1.81
diff -u -r1.81 setup.c
--- setup.c	2001/04/06 05:10:54	1.81
+++ setup.c	2001/06/13 06:08:27
@@ -425,14 +425,10 @@
 			boot_cpu_data.cpu_hz / 1000000,
 			boot_cpu_data.cpu_hz % 1000000	);
 
-#ifdef CONFIG_SERIAL_CONSOLE
-	/* nothing */
-#elif CONFIG_VT
-#if   defined(CONFIG_STI_CONSOLE)
-	conswitchp = &dummy_con;	/* we use take_over_console() later ! */
-#elif defined(CONFIG_DUMMY_CONSOLE)
-	conswitchp = &dummy_con;
-#endif
+#ifdef CONFIG_VT
+# if defined(CONFIG_STI_CONSOLE) || defined(CONFIG_DUMMY_CONSOLE)
+        conswitchp = &dummy_con;        /* we use take_over_console() later ! */
+# endif
 #endif
 
 }

Re: [parisc-linux] sticon/sercon patch for new bootfloppies...

[Richard Hirst]

On Wed, Jun 13, 2001 at 02:24:49AM +0200, Helge Deller wrote:
> Hi,
> 
> the attached patch enables only those consoles, which are given on the linux 
> kernel command line.
> Examples: "console=tty0" activates STIcon only, "console=ttyS0" activates 
> serial console only and if you use both on the commandline you will get the 
> system messages on both consoles.
> 
> Known bugs: 
> - STIcon will not show any messages which happened before the STI
> 	initialisation


Just tried a kernel with serial console, sti console, and sti fb all
compiled in, with palo/unstable.  Seems to do the right thing on my B180 :)

The kernel knows it cannot support b180 fb, so uses sti console automatically
but this kernel should use sti fb on those machines that we support.

I left the b180 alone until the "screen blanking" cut in.  It just turns
off the cursor, btw, doesn't blank the screen.

When I came back to it the cursor had gone back to top-left, and all
chars are printed as blobs of garbage.  System is still happy apart from
that.

I think the cursor going to top-left on a screen blank is a result of
your last lot of bug fixes, but not certain of that.

The other thing for people to note is that if you specify both
console=ttyS0 and console=tty0, then the order is important.  The
real console is the second one you specify, the other just gets
a copy of boot messages (all kernel printk's, I guess).

Richard