public inbox for linuxppc-dev@ozlabs.org 
 help / color / mirror / Atom feed
From: Olof Johansson <olof@lixom•net>
To: linuxppc-dev@ozlabs•org
Subject: [patch 24/35] Work around errata 4628
Date: Thu, 05 Jul 2007 12:02:57 -0500	[thread overview]
Message-ID: <20070705170239.982028000@lixom.net> (raw)
In-Reply-To: 20070705170233.258351000@lixom.net

Errata 4628 causes lockups during table walks that are recovered from
when a ptesync is received. So, issue a ptesync every timer interrupt,
and for CPUs that have not yet spun up, at regular intervals. This seems
sufficient to avoid hitting it.

Index: 2.6.21/arch/powerpc/kernel/head_64.S
===================================================================
--- 2.6.21.orig/arch/powerpc/kernel/head_64.S
+++ 2.6.21/arch/powerpc/kernel/head_64.S
@@ -123,10 +123,22 @@ _GLOBAL(__secondary_hold)
 	std	r24,__secondary_hold_acknowledge@l(0)
 	sync
 
+#ifdef CONFIG_SMP
 	/* All secondary cpus wait here until told to start. */
 100:	ld	r4,__secondary_hold_spinloop@l(0)
 	cmpdi	0,r4,1
 	bne	100b
+#else
+	lis	r3, 0x40
+	mtctr	r3
+100:	bdnz	.
+	ptesync
+	ptesync
+	lis	r3, 0x40
+	mtctr	r3
+	b	100b
+	
+#endif
 
 #if defined(CONFIG_SMP) || defined(CONFIG_KEXEC)
 	LOAD_REG_IMMEDIATE(r4, .generic_secondary_smp_init)
@@ -1582,16 +1594,32 @@ _GLOBAL(generic_secondary_smp_init)
 2:	mtspr	SPRN_SPRG3,r13		/* Save vaddr of paca in SPRG3	 */
 	/* From now on, r24 is expected to be logical cpuid */
 	mr	r24,r5
+
+	lis	r23, 0x40
+	mtctr	r23
+
 3:	HMT_LOW
 	lbz	r23,PACAPROCSTART(r13)	/* Test if this processor should */
-					/* start.			 */
-	sync
 
 #ifndef CONFIG_SMP
 	b	3b			/* Never go on non-SMP		 */
 #else
 	cmpwi	0,r23,0
-	beq	3b			/* Loop until told to go	 */
+	bne	5f			/* Loop until told to go	 */
+					/* start.			 */
+	isync
+
+4:	bdnz	3b
+	ptesync
+	isync
+	sync
+	ptesync
+	lis	r23, 0x40
+	mtctr	r23
+	b	3b
+
+5:
+
 
 	/* See if we need to call a cpu state restore handler */
 	LOAD_REG_IMMEDIATE(r23, cur_cpu_spec)
Index: 2.6.21/arch/powerpc/kernel/time.c
===================================================================
--- 2.6.21.orig/arch/powerpc/kernel/time.c
+++ 2.6.21/arch/powerpc/kernel/time.c
@@ -619,6 +619,8 @@ void timer_interrupt(struct pt_regs * re
 	unsigned long ticks;
 	u64 tb_next_jiffy;
 
+	asm("ptesync");
+
 #ifdef CONFIG_PPC32
 	if (atomic_read(&ppc_n_lost_interrupts) != 0)
 		do_IRQ(regs);

--

  parent reply	other threads:[~2007-07-05 17:03 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-07-05 17:02 [patch 00/35] PA Semi patch set Olof Johansson
2007-07-05 17:02 ` [patch 01/35] pasemi: rename platform Olof Johansson
2007-07-08 23:52   ` Stephen Rothwell
2007-07-09  0:18     ` Olof Johansson
2007-07-09  7:23       ` Stephen Rothwell
2007-07-09 17:19         ` Olof Johansson
2007-07-09 17:21           ` Olof Johansson
2007-07-05 17:02 ` [patch 02/35] PA Semi EDAC driver Olof Johansson
2007-07-05 17:02 ` [patch 03/35] Change powerpc64 ioaddr_t to u_int Olof Johansson
2007-07-05 17:02 ` [patch 04/35] pasemi_mac: Fix TX interrupt threshold Olof Johansson
2007-07-05 17:02 ` [patch 05/35] pasemi_mac: Clean TX ring in poll Olof Johansson
2007-07-05 17:02 ` [patch 06/35] pasemi_mac: Abstract out register access Olof Johansson
2007-07-05 17:02 ` [patch 07/35] pasemi_mac: stop using the pci config space accessors for register read/writes Olof Johansson
2007-07-26 21:25   ` Marian Balakowicz
2007-07-28  8:35     ` Olof Johansson
2007-08-01  9:27       ` Marian Balakowicz
2007-07-05 17:02 ` [patch 08/35] pasemi_mac: Enable L2 caching of packet headers Olof Johansson
2007-07-05 17:02 ` [patch 09/35] pasemi_mac: Simplify memcpy for short receives Olof Johansson
2007-07-05 17:02 ` [patch 10/35] pasemi_mac: Minor performance tweaks Olof Johansson
2007-07-05 17:02 ` [patch 11/35] pasemi_mac: Reduce locking when cleaning TX ring Olof Johansson
2007-07-05 17:02 ` [patch 12/35] pasemi_mac: Enable LLTX Olof Johansson
2007-07-05 17:02 ` [patch 13/35] Cleanup marvell phy driver init Olof Johansson
2007-07-05 17:02 ` [patch 14/35] Add 88E1112 PHY ID to the marvell driver Olof Johansson
2007-07-05 17:02 ` [patch 15/35] Export HID registers via sysfs Olof Johansson
2007-07-05 17:02 ` [patch 16/35] Electra ide platform glue Olof Johansson
2007-07-05 17:02 ` [patch 17/35] CF driver for PA Semi Electra Olof Johansson
2007-07-05 17:02 ` [patch 18/35] Spread IRQs among cpus by default Olof Johansson
2007-07-05 17:02 ` [patch 19/35] Improve machine check output Olof Johansson
2007-07-05 17:02 ` [patch 20/35] Remove idle_spin function pointer Olof Johansson
2007-07-05 17:02 ` [patch 21/35] Use MSR_PMM to disable profiling of the idle loop Olof Johansson
2007-07-05 17:02 ` [patch 22/35] Work around errata 4111 Olof Johansson
2007-07-05 17:02 ` [patch 23/35] Work around UART erratas Olof Johansson
2007-07-05 17:02 ` Olof Johansson [this message]
2007-07-05 17:02 ` [patch 25/35] Work around errata 4025 Olof Johansson
2007-07-05 17:02 ` [patch 26/35] Work around errata 4713 Olof Johansson
2007-07-05 17:03 ` [patch 27/35] Work around errata 4290 Olof Johansson
2007-07-05 18:50   ` Gabriel Paubert
2007-07-05 19:26     ` Olof Johansson
2007-07-05 17:03 ` [patch 28/35] Work around errata 4712 Olof Johansson
2007-07-05 17:03 ` [patch 29/35] Work around errata 4910 Olof Johansson
2007-07-05 17:03 ` [patch 30/35] Disable PURR on pa6t Olof Johansson
2007-07-05 17:03 ` [patch 31/35] Work around errata 4161 Olof Johansson
2007-07-05 17:03 ` [patch 32/35] Dont reset openpic on init Olof Johansson
2007-07-05 17:03 ` [patch 33/35] Work around errata 5667 Olof Johansson
2007-07-05 17:03 ` [patch 34/35] Work around errata 4505 Olof Johansson
2007-07-05 17:03 ` [patch 35/35] Work around errata 5652 Olof Johansson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20070705170239.982028000@lixom.net \
    --to=olof@lixom$(echo .)net \
    --cc=linuxppc-dev@ozlabs$(echo .)org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox