public inbox for linux-arm-kernel@lists.infradead.org 
 help / color / mirror / Atom feed
From: gdavis@mvista•com (George G. Davis)
To: linux-arm-kernel@lists•infradead.org
Subject: [RFC/PATCH v4 6/7] ARM: ARM11 MPCore: DMA_CACHE_RWFO operations are not preempt safe
Date: Tue, 18 Oct 2011 19:26:02 -0400	[thread overview]
Message-ID: <20111018232602.GA237@mvista.com> (raw)
In-Reply-To: <alpine.LFD.2.02.1110181715120.17040@xanadu.home>

Hello,

On Tue, Oct 18, 2011 at 05:28:45PM -0400, Nicolas Pitre wrote:
> On Tue, 18 Oct 2011, gdavis at mvista.com wrote:
> 
> > @@ -241,6 +247,13 @@ v6_dma_inv_range:
> >  	blo	1b
> >  	mov	r0, #0
> >  	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer
> > +#if	defined(CONFIG_DMA_CACHE_RWFO) && defined(CONFIG_PREEMPT)
> > +	str	r3, [ip, #TI_PREEMPT]		@ restore preempt count
> > +	teq	r3, #0				@ preempt count == 0?
> > +	ldreq	r3, [ip, #TI_FLAGS]		@ load flags if yes
> > +	tst	r3, #_TIF_NEED_RESCHED		@ need resched?
> > +	bne	preempt_schedule		@ ret via preempt_schedule
> 
> This is buggy.  If the preempt count is _not_ zero, you end up not 
> loading the TI_FLAGS bits and testing _TIF_NEED_RESCHED against that 
> non-zero preempt count.

Sigh, yep.  FWIW, I had the following before submitting this but dropped
it due to brain failure rereading the code - convinced myself that the
movne was redundant:


diff --git a/arch/arm/mm/cache-v6.S b/arch/arm/mm/cache-v6.S
index af055dc..5c6ce38 100644
--- a/arch/arm/mm/cache-v6.S
+++ b/arch/arm/mm/cache-v6.S
@@ -285,6 +287,7 @@ v6_dma_inv_range:
 	str	r3, [ip, #TI_PREEMPT]		@ restore preempt count
 	teq	r3, #0				@ preempt count == 0?
 	ldreq	r3, [ip, #TI_FLAGS]		@ load flags if yes
+	movne	r3, #0				@ force flags to 0
 	tst	r3, #_TIF_NEED_RESCHED		@ need resched?
 	bne	preempt_schedule		@ ret via preempt_schedule
 #endif
@@ -321,6 +324,7 @@ v6_dma_clean_range:
 	str	r3, [ip, #TI_PREEMPT]		@ restore preempt count
 	teq	r3, #0				@ preempt count == 0?
 	ldreq	r3, [ip, #TI_FLAGS]		@ load flags if yes
+	movne	r3, #0				@ force flags to 0
 	tst	r3, #_TIF_NEED_RESCHED		@ need resched?
 	bne	preempt_schedule		@ ret via preempt_schedule
 #endif
@@ -362,6 +366,7 @@ ENTRY(v6_dma_flush_range)
 	str	r3, [ip, #TI_PREEMPT]		@ restore preempt count
 	teq	r3, #0				@ preempt count == 0?
 	ldreq	r3, [ip, #TI_FLAGS]		@ load flags if yes
+	movne	r3, #0				@ force flags to 0
 	tst	r3, #_TIF_NEED_RESCHED		@ need resched?
 	bne	preempt_schedule		@ ret via preempt_schedule
 #endif

In this case, I think preempt_{disable,enable} is the better option
(over interrupt disable/enable) due to variable DMA buffer sizes.

Thanks!

--
Regards,
George

> 
> 
> Nicolas

  reply	other threads:[~2011-10-18 23:26 UTC|newest]

Thread overview: 93+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-06-20 13:43 parallel load of modules on an ARM multicore EXTERNAL Waechtler Peter (Fa. TCP, CM-AI/PJ-CF31)
2011-06-21 15:50 ` Catalin Marinas
2011-06-23 14:39   ` AW: " EXTERNAL Waechtler Peter (Fa. TCP, CM-AI/PJ-CF31)
2011-06-23 14:52     ` Catalin Marinas
2011-06-23 15:12       ` Russell King - ARM Linux
2011-06-23 15:34         ` Catalin Marinas
2011-06-23 17:02           ` Catalin Marinas
2011-06-23 15:20       ` AW: " EXTERNAL Waechtler Peter (Fa. TCP, CM-AI/PJ-CF31)
2011-07-07  4:25   ` George G. Davis
2011-10-20  4:04   ` George G. Davis
2011-09-22  7:29 ` George G. Davis
2011-09-22  8:52   ` Catalin Marinas
2011-10-06  4:29     ` George G. Davis
2011-10-06  5:08       ` [RFC/PATCH 0/7] ARM: ARM11 MPCore preemption/task migration cache coherency fixups gdavis at mvista.com
2011-10-06  5:08         ` [RFC/PATCH 1/7] ARM: ARM11 MPCore: pgd_alloc is not preempt safe gdavis at mvista.com
2011-10-06 16:35           ` Russell King - ARM Linux
2011-10-06 19:38             ` George G. Davis
2011-10-06  5:08         ` [RFC/PATCH 2/7] ARM: ARM11 MPCore: pte_alloc_one{, _kernel} are " gdavis at mvista.com
2011-10-06  5:08         ` [RFC/PATCH 3/7] ARM: ARM11 MPCore: {clean, flush}_pmd_entry " gdavis at mvista.com
2011-10-06  5:08         ` [RFC/PATCH 4/7] ARM: ARM11 MPCore: clean_dcache_area is " gdavis at mvista.com
2011-10-06  5:08         ` [RFC/PATCH 5/7] ARM: Move get_thread_info macro definition to <asm/assembler.h> gdavis at mvista.com
2011-10-06  5:08         ` [RFC/PATCH 6/7] ARM: ARM11 MPCore: DMA_CACHE_RWFO operations are not preempt safe gdavis at mvista.com
2011-10-06 16:40           ` Russell King - ARM Linux
2011-10-06 19:41             ` George G. Davis
2011-10-06  5:08         ` [RFC/PATCH 7/7] ARM: ARM11 MPCore: cpu_v6_set_pte_ext is " gdavis at mvista.com
2011-10-06  7:46           ` Russell King - ARM Linux
2011-10-06 12:35             ` George G. Davis
2011-10-07  2:38         ` [RFC/PATCH v2 0/7] ARM11 MPCore: preemption/task migration cache coherency fixups gdavis at mvista.com
2011-10-07  2:38           ` [RFC/PATCH 1/7] ARM: ARM11 MPCore: pgd_alloc is not preempt safe gdavis at mvista.com
2011-10-07  2:38           ` [RFC/PATCH 2/7] ARM: ARM11 MPCore: pte_alloc_one{, _kernel} are " gdavis at mvista.com
2011-10-07  7:47             ` [RFC/PATCH 2/7] ARM: ARM11 MPCore: pte_alloc_one{,_kernel} " Russell King - ARM Linux
2011-10-07 15:31               ` [RFC/PATCH 2/7] ARM: ARM11 MPCore: pte_alloc_one{, _kernel} " George G. Davis
2011-10-07  2:38           ` [RFC/PATCH 3/7] ARM: ARM11 MPCore: {clean, flush}_pmd_entry " gdavis at mvista.com
2011-10-11  9:53             ` [RFC/PATCH 3/7] ARM: ARM11 MPCore: {clean,flush}_pmd_entry " Catalin Marinas
2011-10-12  2:34               ` [RFC/PATCH 3/7] ARM: ARM11 MPCore: {clean, flush}_pmd_entry " George G. Davis
2011-10-13 14:31                 ` [RFC/PATCH 3/7] ARM: ARM11 MPCore: {clean,flush}_pmd_entry " Russell King - ARM Linux
2011-10-14  1:34                   ` [RFC/PATCH 3/7] ARM: ARM11 MPCore: {clean, flush}_pmd_entry " George G. Davis
2011-10-07  2:38           ` [RFC/PATCH 4/7] ARM: ARM11 MPCore: clean_dcache_area is " gdavis at mvista.com
2011-10-07  2:38           ` [RFC/PATCH 5/7] ARM: Move get_thread_info macro definition to <asm/assembler.h> gdavis at mvista.com
2011-10-11  9:56             ` Catalin Marinas
2011-10-12  6:04               ` gdavis at mvista.com
2011-10-13 14:34                 ` Russell King - ARM Linux
2011-10-13 14:49                   ` Catalin Marinas
2011-10-13 14:53                     ` Russell King - ARM Linux
2011-10-14  1:46                       ` George G. Davis
2011-10-14  1:44                     ` George G. Davis
2011-10-14  1:42                   ` George G. Davis
2011-10-14  2:54                     ` Nicolas Pitre
2011-10-14 12:56                       ` George G. Davis
2011-10-07  2:38           ` [RFC/PATCH 6/7] ARM: ARM11 MPCore: DMA_CACHE_RWFO operations are not preempt safe gdavis at mvista.com
2011-10-07  2:38           ` [RFC/PATCH 7/7] ARM: ARM11 MPCore: cpu_v6_set_pte_ext is " gdavis at mvista.com
2011-10-07 16:26           ` [RFC/PATCH v3 0/7] ARM11 MPCore: preemption/task migration cache coherency fixups gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 1/7] ARM: ARM11 MPCore: pgd_alloc is not preempt safe gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 2/7] ARM: ARM11 MPCore: pte_alloc_one{, _kernel} are " gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 3/7] ARM: ARM11 MPCore: {clean, flush}_pmd_entry " gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 4/7] ARM: ARM11 MPCore: clean_dcache_area is " gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 5/7] ARM: Move get_thread_info macro definition to <asm/assembler.h> gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 6/7] ARM: ARM11 MPCore: DMA_CACHE_RWFO operations are not preempt safe gdavis at mvista.com
2011-10-07 16:26             ` [RFC/PATCH v3 7/7] ARM: ARM11 MPCore: cpu_v6_set_pte_ext is " gdavis at mvista.com
2011-10-18 13:47             ` [RFC/PATCH v4 0/7] ARM11 MPCore: preemption/task migration cache coherency fixups gdavis at mvista.com
2011-10-18 13:47               ` [RFC/PATCH v4 1/7] ARM: ARM11 MPCore: pgd_alloc is not preempt safe gdavis at mvista.com
2011-10-18 13:47               ` [RFC/PATCH v4 2/7] ARM: ARM11 MPCore: pte_alloc_one{, _kernel} are " gdavis at mvista.com
2011-10-18 13:47               ` [RFC/PATCH v4 3/7] ARM: ARM11 MPCore: {clean, flush}_pmd_entry " gdavis at mvista.com
2011-10-18 13:47               ` [RFC/PATCH v4 4/7] ARM: ARM11 MPCore: clean_dcache_area is " gdavis at mvista.com
2011-10-18 17:08                 ` Tony Lindgren
2011-10-18 17:30                   ` George G. Davis
2011-10-18 17:43                     ` Tony Lindgren
2011-10-18 18:13                       ` George G. Davis
2011-10-18 13:47               ` [RFC/PATCH v4 5/7] ARM: Move get_thread_info macro definition to <asm/assembler.h> gdavis at mvista.com
2011-10-18 13:47               ` [RFC/PATCH v4 6/7] ARM: ARM11 MPCore: DMA_CACHE_RWFO operations are not preempt safe gdavis at mvista.com
2011-10-18 21:28                 ` Nicolas Pitre
2011-10-18 23:26                   ` George G. Davis [this message]
2011-10-19  1:09                     ` Nicolas Pitre
2011-10-18 13:47               ` [RFC/PATCH v4 7/7] ARM: ARM11 MPCore: cpu_v6_set_pte_ext is " gdavis at mvista.com
2011-10-18 21:52                 ` Nicolas Pitre
2011-10-18 23:29                   ` George G. Davis
2012-06-12 20:40               ` [RFC/PATCH v5 0/7] ARM11 MPCore: preemption/task migration cache coherency fixups gdavis at mvista.com
2012-06-12 20:40                 ` [RFC/PATCH v5 1/7] ARM: ARM11 MPCore: Make pgd_alloc preempt safe gdavis at mvista.com
2012-06-12 20:40                 ` [RFC/PATCH v5 2/7] ARM: ARM11 MPCore: Make pte_alloc_one{, _kernel} " gdavis at mvista.com
2012-06-12 20:40                 ` [RFC/PATCH v5 3/7] ARM: ARM11 MPCore: Make {clean, flush}_pmd_entry " gdavis at mvista.com
2012-06-12 20:40                 ` [RFC/PATCH v5 4/7] ARM: Move get_thread_info macro definition to <asm/assembler.h> gdavis at mvista.com
2012-06-12 20:40                 ` [RFC/PATCH v5 5/7] ARM: ARM11 MPCore: cpu_v6_dcache_clean_area needs RFO gdavis at mvista.com
2012-06-13  9:32                   ` Catalin Marinas
2012-06-13  9:36                     ` Russell King - ARM Linux
2012-06-13  9:41                       ` Catalin Marinas
2012-06-13  9:45                         ` Russell King - ARM Linux
2012-06-13  9:54                           ` Catalin Marinas
2012-06-13 11:36                       ` George G. Davis
2012-06-13 11:21                     ` George G. Davis
2012-06-12 20:40                 ` [RFC/PATCH v5 6/7] ARM: ARM11 MPCore: Make DMA_CACHE_RWFO operations preempt safe gdavis at mvista.com
2012-06-12 20:40                 ` [RFC/PATCH v5 7/7] ARM: ARM11 MPCore: Make cpu_v6_set_pte_ext " gdavis at mvista.com
2012-06-13  9:34                   ` Catalin Marinas
2012-06-13 11:35                     ` George G. Davis

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=20111018232602.GA237@mvista.com \
    --to=gdavis@mvista$(echo .)com \
    --cc=linux-arm-kernel@lists$(echo .)infradead.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