NetBSD Problem Report #52010

From www@NetBSD.org  Mon Feb 27 17:47:16 2017
Return-Path: <www@NetBSD.org>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
	(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits))
	(Client CN "mail.netbsd.org", Issuer "Postmaster NetBSD.org" (verified OK))
	by mollari.NetBSD.org (Postfix) with ESMTPS id DAA947A16C
	for <gnats-bugs@gnats.NetBSD.org>; Mon, 27 Feb 2017 17:47:16 +0000 (UTC)
Message-Id: <20170227174715.79A207A2A7@mollari.NetBSD.org>
Date: Mon, 27 Feb 2017 17:47:15 +0000 (UTC)
From: smesgr@gmail.com 
Reply-To: smesgr@gmail.com 
To: gnats-bugs@NetBSD.org
Subject: Regression: Gumstix Verdex is hanging in enabling cache + KASSERT ram_size
X-Send-Pr-Version: www-1.0

>Number:         52010
>Category:       port-evbarm
>Synopsis:       Regression: Gumstix Verdex is hanging in enabling cache + KASSERT ram_size
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    skrll
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Feb 27 17:50:00 +0000 2017
>Closed-Date:    Tue Mar 14 21:08:32 +0000 2017
>Last-Modified:  Tue Mar 14 21:08:32 +0000 2017
>Originator:     Stephan Meisinger
>Release:        7.0 and higher
>Organization:
>Environment:
>Description:
Gumstix Verdex (very likely every XSCALE Gumstix) is hanging after enabling cache code, because no cache clean address is set.

Additionally a KASSERT is triggered, which is afaik only for a __HAVE_MM_MD_DIRECT_MAPPED_PHYS setup.

Please see patch below. The regression was introduced with the switch to arm32_bootmem_init code

http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/arch/evbarm/gumstix/gumstix_machdep.c.diff?r1=1.48&r2=1.49&only_with_tag=MAIN&f=h
>How-To-Repeat:
boot GUMSTIX Verdex with NetBSD 7 or higher
>Fix:
Index: sys/arch/evbarm/gumstix/gumstix_machdep.c
===================================================================
RCS file: /cvsroot/src/sys/arch/evbarm/gumstix/gumstix_machdep.c,v
retrieving revision 1.57
diff -u -r1.57 gumstix_machdep.c
--- sys/arch/evbarm/gumstix/gumstix_machdep.c	20 Oct 2016 09:53:08 -0000	1.57
+++ sys/arch/evbarm/gumstix/gumstix_machdep.c	27 Feb 2017 17:31:28 -0000
@@ -456,6 +456,9 @@
 u_int
 initarm(void *arg)
 {
+#if defined(CPU_XSCALE)
+	extern vaddr_t xscale_cache_clean_addr;
+#endif  
 	extern char KERNEL_BASE_phys[];
 	extern uint32_t *u_boot_args[];
 	extern uint32_t ram_size;
@@ -594,7 +597,11 @@
 	bootconfig.dram[0].address = SDRAM_START;
 	bootconfig.dram[0].pages = ram_size / PAGE_SIZE;

+#if defined(CPU_XSCALE)
+	xscale_cache_clean_addr = 0xff000000U;
+#else
 	KASSERT(ram_size <= KERNEL_VM_BASE - KERNEL_BASE);
+#endif

 	arm32_bootmem_init(bootconfig.dram[0].address, ram_size,
 	    (uintptr_t) KERNEL_BASE_phys);

>Release-Note:

>Audit-Trail:

Responsible-Changed-From-To: port-evbarm-maintainer->skrll
Responsible-Changed-By: skrll@NetBSD.org
Responsible-Changed-When: Mon, 27 Feb 2017 21:25:32 +0000
Responsible-Changed-Why:
Take


From: Nick Hudson <skrll@netbsd.org>
To: gnats-bugs@NetBSD.org, port-evbarm-maintainer@netbsd.org,
 gnats-admin@netbsd.org, netbsd-bugs@netbsd.org
Cc: 
Subject: Re: port-evbarm/52010: Regression: Gumstix Verdex is hanging in
 enabling cache + KASSERT ram_size
Date: Tue, 28 Feb 2017 07:49:40 +0000

 This is a multi-part message in MIME format.
 --------------020700070203080902030601
 Content-Type: text/plain; charset=windows-1252; format=flowed
 Content-Transfer-Encoding: 7bit

 Does this patch work for you?

 Thanks,
 Nick

 --------------020700070203080902030601
 Content-Type: text/plain; charset=UTF-8;
  name="diff"
 Content-Transfer-Encoding: 7bit
 Content-Disposition: attachment;
  filename="diff"

 Index: sys/arch/evbarm/gumstix/gumstix_machdep.c
 ===================================================================
 RCS file: /cvsroot/src/sys/arch/evbarm/gumstix/gumstix_machdep.c,v
 retrieving revision 1.57
 diff -u -p -r1.57 gumstix_machdep.c
 --- sys/arch/evbarm/gumstix/gumstix_machdep.c	20 Oct 2016 09:53:08 -0000	1.57
 +++ sys/arch/evbarm/gumstix/gumstix_machdep.c	28 Feb 2017 07:48:38 -0000
 @@ -206,13 +206,16 @@
  #endif

  /*
 - * The range 0xc1000000 - 0xcfffffff is available for kernel VM space
 - * Core-logic registers and I/O mappings occupy 0xfd000000 - 0xffffffff
 + * The range 0xc1000000 - 0xfd000000 is available for kernel VM space
 + * Core-logic registers and I/O mappings occupy
 + *
 + *    0xfd000000 - 0xfd800000	on gumstix
 + *    0xc0000000 - 0xc0400000	on overo, duovero and pepper
   */
  #ifndef KERNEL_VM_BASE
 -#define	KERNEL_VM_BASE		0xc1000000
 +#define	KERNEL_VM_BASE		0xc8000000
  #endif
 -#define KERNEL_VM_SIZE		0x0f000000
 +#define KERNEL_VM_SIZE		0x35000000

  BootConfig bootconfig;		/* Boot config storage */
  static char bootargs[MAX_BOOT_STRING];
 @@ -487,6 +490,9 @@ initarm(void *arg)
  	 */

  #if defined(CPU_XSCALE)
 +	extern vaddr_t xscale_cache_clean_addr;
 +	xscale_cache_clean_addr = 0xff000000U;
 +
  	cpu_reset_address = NULL;
  #elif defined(OMAP2)
  	cpu_reset_address = omap_reset;

 --------------020700070203080902030601--

From: smesgr <smesgr@gmail.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: port-evbarm/52010
Date: Tue, 28 Feb 2017 19:31:45 +0100

 You patch also works for me, even through I can't test the other Gumstix 
 boards (e.g. boards with Cortex chips)

 Best Regards

 Stephan

From: "Nick Hudson" <skrll@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/52010 CVS commit: src/sys/arch/evbarm/gumstix
Date: Wed, 1 Mar 2017 16:44:54 +0000

 Module Name:	src
 Committed By:	skrll
 Date:		Wed Mar  1 16:44:54 UTC 2017

 Modified Files:
 	src/sys/arch/evbarm/gumstix: gumstix_machdep.c

 Log Message:
 Set xscale_cache_clean_addr appropriately and re-arrange default KVA
 layout to allow direct map for all boards.

 OVERO/DUOVERO/PEPPER aren't tested.

 PR/52010: Regression: Gumstix Verdex is hanging in enabling cache + KASSERT ram_size


 To generate a diff of this commit:
 cvs rdiff -u -r1.57 -r1.58 src/sys/arch/evbarm/gumstix/gumstix_machdep.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

State-Changed-From-To: open->pending-pullups
State-Changed-By: skrll@NetBSD.org
State-Changed-When: Wed, 01 Mar 2017 16:51:54 +0000
State-Changed-Why:
[pullup-7 #1366]


From: "Soren Jacobsen" <snj@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/52010 CVS commit: [netbsd-7] src/sys/arch
Date: Sat, 11 Mar 2017 07:40:21 +0000

 Module Name:	src
 Committed By:	snj
 Date:		Sat Mar 11 07:40:21 UTC 2017

 Modified Files:
 	src/sys/arch/arm/arm32 [netbsd-7]: pmap.c
 	src/sys/arch/arm/include/arm32 [netbsd-7]: pmap.h
 	src/sys/arch/evbarm/gumstix [netbsd-7]: gumstix_machdep.c

 Log Message:
 Pull up following revision(s) (requested by skrll in ticket #1366):
 	sys/arch/arm/include/arm32/pmap.h: 1.145
 	sys/arch/arm/arm32/pmap.c: 1.343, 1.344
 	sys/arch/evbarm/gumstix/gumstix_machdep.c: 1.58 via patch
 Fixup the compile time decisions around PMAP_{INCLUDE,NEEDS}_PTE_SYNC and
 fix the options for xscale boards which require the code in
 pmap_l2ptp_ctor marked as #ifndef PMAP_INCLUDE_PTE_SYNC.
 Fix the typo (pte -> opte) in this code block and consistently use opte
 elsewhere.
 PR/51990: Regression data_abort_handler: data_aborts fsr=0x406 far=0xbfffeff5  on copyout in init
 --
 fix unused.
 --
 Set xscale_cache_clean_addr appropriately and re-arrange default KVA
 layout to allow direct map for all boards.
 OVERO/DUOVERO/PEPPER aren't tested.
 PR/52010: Regression: Gumstix Verdex is hanging in enabling cache + KASSERT ram_size


 To generate a diff of this commit:
 cvs rdiff -u -r1.295.2.8 -r1.295.2.9 src/sys/arch/arm/arm32/pmap.c
 cvs rdiff -u -r1.135.2.2 -r1.135.2.3 src/sys/arch/arm/include/arm32/pmap.h
 cvs rdiff -u -r1.50 -r1.50.2.1 src/sys/arch/evbarm/gumstix/gumstix_machdep.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

State-Changed-From-To: pending-pullups->feedback
State-Changed-By: skrll@NetBSD.org
State-Changed-When: Sat, 11 Mar 2017 09:29:46 +0000
State-Changed-Why:
OK to close?


From: smesgr <smesgr@gmail.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: port-evbarm/52010
Date: Tue, 14 Mar 2017 19:26:26 +0100

 for me - yes thanks

State-Changed-From-To: feedback->closed
State-Changed-By: skrll@NetBSD.org
State-Changed-When: Tue, 14 Mar 2017 21:08:32 +0000
State-Changed-Why:
Confirmed fixed


>Unformatted:

Home	PR Database Search