Re: [ANNOUNCE] 3.0.1-rt8 - Regression to 3.0-rt7

["Tim Sander" <tim.sander@hbm.com>]

Am Donnerstag, 11. August 2011, 14:41:38 schrieb Peter Zijlstra:
> On Thu, 2011-08-11 at 14:02 +0200, Peter Zijlstra wrote:
> > On Thu, 2011-08-11 at 13:30 +0200, Tim Sander wrote:
> > > The error does not occur  when configured with
> > > CONFIG_PREEMPT_RT_FULL=y
> > > but with
> > > CONFIG_PREEMPT_RTB=y.
> > > The same error also happens with
> > > CONFIG_PREEMPT_LL=y. (A log is appended after the other log just
> > > search for CONFIG_PREEMPT_LL
> > > for starting point).
> > 
> > Does the below (on top of -rt8) cure things for you? Shouldn't make a
> > difference for PREEMPT_LL though, so maybe there's something else busted
> > as well. But from what I can tell PREEMPT_BASE should be fine with such
> > things disabling preemption..
Yes it works for CONFIG_PREEMPT_RTB=y and as you say not for 
CONFIG_PREEMPT_LL=y.

However with the following patches on top of rt8:

diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h
index 44b3751..21ce185 100644
--- a/include/linux/uaccess.h
+++ b/include/linux/uaccess.h
@@ -22,7 +22,7 @@ static inline void raw_pagefault_enable(void)
        preempt_check_resched();
 }
 
-#ifndef CONFIG_PREEMPT_RT_FULL
+#ifndef CONFIG_PREEMPT_RT_BASE
 static inline void pagefault_disable(void)
 {
        raw_pagefault_disable();
diff --git a/mm/filemap.c b/mm/filemap.c
index 71e2a1f..af63de3 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2040,7 +2040,7 @@ size_t iov_iter_copy_from_user_atomic(struct page *page,
        char *kaddr;
        size_t copied;
 
-       BUG_ON(!current->pagefault_disabled);
+       BUG_ON(!(in_atomic() || current->pagefault_disabled));
        kaddr = kmap_atomic(page, KM_USER0);
        if (likely(i->nr_segs == 1)) {
                int left;
diff --git a/mm/memory.c b/mm/memory.c
index cf71bd9..1b5a116 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3382,7 +3382,7 @@ unlock:
        return 0;
 }
 
-#ifdef CONFIG_PREEMPT_RT_FULL
+#ifdef CONFIG_PREEMPT_RT_BASE
 void pagefault_disable(void)
 {
        migrate_disable();

i get the following bug:

kernel BUG at kernel/posix-cpu-timers.c:1315!
Unable to handle kernel NULL pointer dereference at virtual address 00000000
pgd = c0004000
[00000000] *pgd=00000000
Internal error: Oops: 817 [#1] PREEMPT
Modules linked in: i2c_imx(+) netx netx_reset dspload fpgaload mxc_nand(+) nand nand_ids nand_ecc
CPU: 0    Not tainted  (3.0.0-pmx-rt8-00009-g39305ab #167)
PC is at __bug+0x24/0x30
LR is at walk_stackframe+0x24/0x40
pc : [<c002cbf0>]    lr : [<c002cbb0>]    psr: 60000113
sp : c782ff40  ip : c782fe28  fp : c782ff4c
r10: 00000001  r9 : 00000000  r8 : c78663c0
r7 : c030463c  r6 : c782e000  r5 : 00000000  r4 : c78663c0
r3 : 00000000  r2 : 00000001  r1 : c782fe58  r0 : 00000034
Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
Control: 00c5387d  Table: 87b70008  DAC: 00000017
Process posixcputmr/0 (pid: 6, stack limit = 0xc782e268)
Stack: (0xc782ff40 to 0xc7830000)
ff40: c782ff8c c782ff50 c005a5c0 c002cbd8 00000000 c782ff54 c782ff54 00000000
ff60: 00000000 c78663c0 00000000 c782e000 c030463c c78663c0 00000000 00000001
ff80: c782ffbc c782ff90 c005a88c c005a594 00000000 c7817f28 00000000 c782ffcc
ffa0: c005a800 00000000 00000000 00000000 c782fff4 c782ffc0 c00570f0 c005a80c
ffc0: c7817f28 00000000 00000000 00000000 c782ffd0 c782ffd0 c7817f28 c0057064
ffe0: c0040868 00000013 00000000 c782fff8 c0040868 c0057070 cccdcd6c ffdec890
Backtrace: 
[<c002cbcc>] (__bug+0x0/0x30) from [<c005a5c0>] (__run_posix_cpu_timers+0x38/0x278)
[<c005a588>] (__run_posix_cpu_timers+0x0/0x278) from [<c005a88c>] (posix_cpu_timers_thread+0x8c/0x124)
[<c005a800>] (posix_cpu_timers_thread+0x0/0x124) from [<c00570f0>] (kthread+0x8c/0x94)
[<c0057064>] (kthread+0x0/0x94) from [<c0040868>] (do_exit+0x0/0x2d8)
 r7:00000013 r6:c0040868 r5:c0057064 r4:c7817f28
Code: e59f0010 e1a01003 eb086a88 e3a03000 (e5833000) 
---[ end trace d632afeaa11559d4 ]---

Otherwise the system boots and works

Best regards
Tim

Hottinger Baldwin Messtechnik GmbH, Im Tiefen See 45, 64293 Darmstadt, Germany | www.hbm.com 

Registered as GmbH (German limited liability corporation) in the commercial register at the local court of Darmstadt, HRB 1147  
Company domiciled in Darmstadt | CEO: Andreas Huellhorst | Chairman of the board: James Charles Webster

Als Gesellschaft mit beschraenkter Haftung eingetragen im Handelsregister des Amtsgerichts Darmstadt unter HRB 1147 
Sitz der Gesellschaft: Darmstadt | Geschaeftsfuehrung: Andreas Huellhorst | Aufsichtsratsvorsitzender: James Charles Webster

The information in this email is confidential. It is intended solely for the addressee. If you are not the intended recipient, please let me know and delete this email.

Die in dieser E-Mail enthaltene Information ist vertraulich und lediglich für den Empfaenger bestimmt. Sollten Sie nicht der eigentliche Empfaenger sein, informieren Sie mich bitte kurz und loeschen diese E-Mail.