From: Manuel Lauss <manuel.lauss@googlemail.com>
To: Linux-MIPS <linux-mips@linux-mips.org>
Cc: Manuel Lauss <manuel.lauss@gmail.com>
Subject: [PATCH v2] MIPS: Alchemy: fix dbdma ring destruction memory debugcheck.
Date: Tue, 26 Jan 2010 20:39:33 +0100 [thread overview]
Message-ID: <1264534773-24909-1-git-send-email-manuel.lauss@gmail.com> (raw)
DBDMA descriptors need to be located at 32-byte aligned addresses;
however kmalloc in conjunction with the SLAB allocator and
CONFIG_DEBUG_SLUB enabled doesn't deliver any. The dbdma code works
around that by allocating a larger area and realigning the start
address within it.
When freeing a channel however this adjustment is not taken into
account which results in an oops:
Kernel bug detected[#1]:
[...]
Call Trace:
[<80186010>] cache_free_debugcheck+0x284/0x318
[<801869d8>] kfree+0xe8/0x2a0
[<8010b31c>] au1xxx_dbdma_chan_free+0x2c/0x7c
[<80388dc8>] au1x_pcm_dbdma_free+0x34/0x4c
[<80388fa8>] au1xpsc_pcm_close+0x28/0x38
[<80383cb8>] soc_codec_close+0x14c/0x1cc
[<8036dbb4>] snd_pcm_release_substream+0x60/0xac
[<8036dc40>] snd_pcm_release+0x40/0xa0
[<8018c7a8>] __fput+0x11c/0x228
[<80188f60>] filp_close+0x7c/0x98
[<80189018>] sys_close+0x9c/0xe4
[<801022a0>] stack_done+0x20/0x3c
Fix this by recording the address delivered by kmalloc() and using
it as parameter to kfree().
This fix is only necessary with the SLAB allocator and CONFIG_DEBUG_SLAB
enabled; non-debug SLAB, SLUB do return nicely aligned addresses,
debug-enabled SLUB currently panics early in the boot process.
Signed-off-by: Manuel Lauss <manuel.lauss@gmail.com>
---
v2: slightly improved the patch rationale, thanks to David Daney.
arch/mips/alchemy/common/dbdma.c | 7 +++++--
arch/mips/include/asm/mach-au1x00/au1xxx_dbdma.h | 1 +
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/arch/mips/alchemy/common/dbdma.c b/arch/mips/alchemy/common/dbdma.c
index 40071bd..3b2ccc0 100644
--- a/arch/mips/alchemy/common/dbdma.c
+++ b/arch/mips/alchemy/common/dbdma.c
@@ -411,8 +411,11 @@ u32 au1xxx_dbdma_ring_alloc(u32 chanid, int entries)
if (desc_base == 0)
return 0;
+ ctp->cdb_membase = desc_base;
desc_base = ALIGN_ADDR(desc_base, sizeof(au1x_ddma_desc_t));
- }
+ } else
+ ctp->cdb_membase = desc_base;
+
dp = (au1x_ddma_desc_t *)desc_base;
/* Keep track of the base descriptor. */
@@ -829,7 +832,7 @@ void au1xxx_dbdma_chan_free(u32 chanid)
au1xxx_dbdma_stop(chanid);
- kfree((void *)ctp->chan_desc_base);
+ kfree((void *)ctp->cdb_membase);
stp->dev_flags &= ~DEV_FLAGS_INUSE;
dtp->dev_flags &= ~DEV_FLAGS_INUSE;
diff --git a/arch/mips/include/asm/mach-au1x00/au1xxx_dbdma.h b/arch/mips/include/asm/mach-au1x00/au1xxx_dbdma.h
index c098b45..8c6b110 100644
--- a/arch/mips/include/asm/mach-au1x00/au1xxx_dbdma.h
+++ b/arch/mips/include/asm/mach-au1x00/au1xxx_dbdma.h
@@ -305,6 +305,7 @@ typedef struct dbdma_chan_config {
dbdev_tab_t *chan_dest;
au1x_dma_chan_t *chan_ptr;
au1x_ddma_desc_t *chan_desc_base;
+ u32 cdb_membase; /* kmalloc base of above */
au1x_ddma_desc_t *get_ptr, *put_ptr, *cur_ptr;
void *chan_callparam;
void (*chan_callback)(int, void *);
--
1.6.6
next reply other threads:[~2010-01-26 19:39 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-01-26 19:39 Manuel Lauss [this message]
2010-01-27 22:18 ` [PATCH v2] MIPS: Alchemy: fix dbdma ring destruction memory debugcheck Ralf Baechle
2010-01-28 8:23 ` Manuel Lauss
2010-01-28 8:23 ` Manuel Lauss
2010-01-28 11:22 ` Ralf Baechle
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=1264534773-24909-1-git-send-email-manuel.lauss@gmail.com \
--to=manuel.lauss@googlemail.com \
--cc=linux-mips@linux-mips.org \
--cc=manuel.lauss@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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;
as well as URLs for NNTP newsgroup(s).