Fix for break handling in sunsab.c for 2.6

Fix for break handling in sunsab.c for 2.6

[Jurij Smakov]

Hello,

Due to some bugs in break handling in the sunsab.c serial driver, sending 
a break over serial console (equivalent of Stop-A sent from keyboard) does 
not dump user to the PROM command line. AFAICT, in the 2.4 -> 2.6 
transition the checking for break has been moved from check_status() to 
receive_chars(). However, in the current version of the driver the BRK 
interrupt never reaches receive_chars(). Even if it would, for the break 
to take effect a non-zero count of characters must be present in the
buffer at the same time, which never happens. The patch below corrects the 
problem (I have tried to make it as unintrusive as possible).

--- build-sparc64.orig/drivers/serial/sunsab.c	2004-11-21 12:03:51.282281488 -0500
+++ build-sparc64/drivers/serial/sunsab.c	2004-11-21 11:49:04.000000000 -0500
@@ -143,6 +143,11 @@
  		writeb(SAB82532_CMDR_RMC, &up->regs->w.cmdr);
  	}

+	/* Count may be zero for BRK, so we check for it here */
+	if((stat->sreg.isr1 & SAB82532_ISR1_BRK) &&
+	   (up->port.line = up->port.cons->index))
+		saw_console_brk = 1;
+
  	for (i = 0; i < count; i++) {
  		unsigned char ch = buf[i];

@@ -172,8 +177,6 @@
  				stat->sreg.isr0 &= ~(SAB82532_ISR0_PERR |
  						     SAB82532_ISR0_FERR);
  				up->port.icount.brk++;
-				if (up->port.line = up->port.cons->index)
-					saw_console_brk = 1;
  				/*
  				 * We do the SysRQ and SAK checking
  				 * here because otherwise the break
@@ -325,8 +328,9 @@

  	tty = NULL;
  	if (status.stat) {
-		if (status.sreg.isr0 & (SAB82532_ISR0_TCD | SAB82532_ISR0_TIME |
-					SAB82532_ISR0_RFO | SAB82532_ISR0_RPF))
+		if ((status.sreg.isr0 & (SAB82532_ISR0_TCD | SAB82532_ISR0_TIME |
+					SAB82532_ISR0_RFO | SAB82532_ISR0_RPF)) ||
+		    (status.sreg.isr1 & SAB82532_ISR1_BRK))
  			tty = receive_chars(up, &status, regs);
  		if ((status.sreg.isr0 & SAB82532_ISR0_CDSC) ||
  		    (status.sreg.isr1 & SAB82532_ISR1_CSC))
@@ -352,8 +356,9 @@

  	tty = NULL;
  	if (status.stat) {
-		if (status.sreg.isr0 & (SAB82532_ISR0_TCD | SAB82532_ISR0_TIME |
-					SAB82532_ISR0_RFO | SAB82532_ISR0_RPF))
+		if ((status.sreg.isr0 & (SAB82532_ISR0_TCD | SAB82532_ISR0_TIME |
+					SAB82532_ISR0_RFO | SAB82532_ISR0_RPF)) ||
+		    (status.sreg.isr1 & SAB82532_ISR1_BRK))
  			tty = receive_chars(up, &status, regs);
  		if ((status.sreg.isr0 & SAB82532_ISR0_CDSC) ||
  		    (status.sreg.isr1 & (SAB82532_ISR1_BRK | SAB82532_ISR1_CSC)))

Best regards,

Jurij Smakov                                        jurij@wooyd.org
Key: http://www.wooyd.org/pgpkey/                   KeyID: C99E03CC

Re: Fix for break handling in sunsab.c for 2.6

[David S. Miller]

On Sun, 21 Nov 2004 13:43:05 -0500 (EST)
Jurij Smakov <jurij@wooyd.org> wrote:

> Due to some bugs in break handling in the sunsab.c serial driver, sending 
> a break over serial console (equivalent of Stop-A sent from keyboard) does 
> not dump user to the PROM command line. AFAICT, in the 2.4 -> 2.6 
> transition the checking for break has been moved from check_status() to 
> receive_chars(). However, in the current version of the driver the BRK 
> interrupt never reaches receive_chars(). Even if it would, for the break 
> to take effect a non-zero count of characters must be present in the
> buffer at the same time, which never happens. The patch below corrects the 
> problem (I have tried to make it as unintrusive as possible).

Patch applied, thanks a lot Jurij.