Hi,
On Mon, Jun 22, 2015 at 10:04:22AM +0300, Adrian Hunter wrote:
> On 22/06/15 06:41, Peng Fan wrote:
> > We should not call dma_free_coherent if host->adma_table is NULL,
> > otherwise may trigger panic.
> >
> >>From DMA-API.txt:
> > "
> > void
> > dma_free_coherent(struct device *dev, size_t size, void *cpu_addr,
> > dma_addr_t dma_handle)
> >
> > Free a region of consistent memory you previously allocated. dev,
> > size and dma_handle must all be the same as those passed into
> > dma_alloc_coherent(). cpu_addr must be the virtual address returned by
> > the dma_alloc_coherent().
> > "
> > So we should check cpu_addr, but not directly call dma_free_coherent.
> >
> > I met this when porting xen, log: "
> > sdhci: Secure Digital Host Controller Interface driver
> > sdhci: Copyright(c) Pierre Ossman
> > sdhci-pltfm: SDHCI platform and OF driver helper
> > Unable to handle kernel NULL pointer dereference at virtual address 00000000
> > pgd = 80004000
> > [00000000] *pgd=00000000
> > Internal error: Oops: 5 [#1] PREEMPT SMP ARM
> > Modules linked in:
> > CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3.14.38-02383-g5ccf32b-dirty #22
> > task: 84074000 ti: 84078000 task.ti: 84078000
> > PC is at bitmap_clear+0xc0/0xdc
> > LR is at bitmap_clear+0x54/0xdc
> > pc : [<8029deb8>] lr : [<8029de4c>] psr: 20000193
> > sp : 84079d80 ip : 00000001 fp : 00000000
> > r10: 00077fff r9 : 00000404 r8 : 00000001
> > r7 : 00000001 r6 : 00000001 r5 : 00000000 r4 : ffffffff
> > r3 : 00000001 r2 : 00000001 r1 : 20000193 r0 : 00000015
> > Flags: nzCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment kernel
> > Control: 10c53c7d Table: 8800406a DAC: 00000015
> > Process swapper/0 (pid: 1, stack limit = 0x84078238)
> > Stack: (0x84079d80 to 0x8407a000)
> > 9d80: 80000113 00000000 87efa000 81109918 00001000 800197f8 84128558
> > 00080008
> > 9da0: 84079dec 000000d0 84bfeac0 84126c10 84126c10 ffffffff 00000404
> > ffffffff
> > 9dc0: 00000000 00000402 00000000 00000000 84126c10 80310ba8 ffffffff
> > 00000000
> > 9de0: 00000000 00000524 84078000 00000000 00000000 ffffffff ffffffff
> > 84bfeac0
> > 9e00: 84bfe800 8000b407 07eb0000 8116e0f8 00000000 804ee81c ffffffff
> > ffffffff
> > 9e20: 00000000 84126c10 84c92010 84bfeac0 00000000 84126c10 84126c00
> > 84bfeac0
> > 9e40: 84078030 804f08e4 804f03d8 84126c10 fffffdfb 8115401c 8115401c
> > 00000000
> > 9e60: 0000010f 80362330 803622ec 84126c10 811c8098 00000000 8115401c
> > 80360b1c
> > 9e80: 84126c10 8115401c 84126c44 00000000 80de1888 80360d28 00000000
> > 8115401c
> > 9ea0: 80360c9c 8035f10c 8406965c 84123634 8115401c 84c8bf80 8112f3c8
> > 803602bc
> > 9ec0: 80d08314 8115401c 00000006 8115401c 00000006 8116e080 8116e080
> > 8036130c
> > 9ee0: 00000000 80e00f78 00000006 800088dc 8400f900 80c94fe0 840bd480
> > 80735184
> > 9f00: 00000000 8116e080 0000150c 8012d430 00000000 811105b0 60000113
> > 00000001
> > 9f20: 87ffc576 8075ca38 0000010f 8004b0f0 80d66884 00000006 87ffc583
> > 00000006
> > 9f40: 811105a0 80e00f78 00000006 8116e080 8116e080 80da150c 0000010f
> > 80df4154
> > 9f60: 80df4148 80da1c4c 00000006 00000006 80da150c 9355553c 84079f9c
> > 80731338
> > 9f80: 00000000 00000000 80727254 00000000 00000000 00000000 00000000
> > 00000000
> > 9fa0: 00000000 8072725c 00000000 8000ecf8 00000000 00000000 00000000
> > 00000000
> > 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000
> > 00000000
> > 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000 9355553c
> > 9355553c
> > [<8029deb8>] (bitmap_clear) from [<800197f8>]
> > (__arm_dma_free.isra.18+0xe4/0x228)
> > [<800197f8>] (__arm_dma_free.isra.18) from [<80310ba8>]
> > (xen_swiotlb_free_coherent+0xfc/0x140)
> > [<80310ba8>] (xen_swiotlb_free_coherent) from [<804ee81c>]
> > (sdhci_add_host+0xb34/0xe64)
> > [<804ee81c>] (sdhci_add_host) from [<804f08e4>]
> > (sdhci_esdhc_imx_probe+0x50c/0x808)
> > [<804f08e4>] (sdhci_esdhc_imx_probe) from [<80362330>]
> > (platform_drv_probe+0x44/0xa4)
> > [<80362330>] (platform_drv_probe) from [<80360b1c>]
> > (driver_probe_device+0x120/0x25c)
> > [<80360b1c>] (driver_probe_device) from [<80360d28>]
> > (__driver_attach+0x8c/0x90)
> > [<80360d28>] (__driver_attach) from [<8035f10c>]
> > (bus_for_each_dev+0x60/0x94)
> > [<8035f10c>] (bus_for_each_dev) from [<803602bc>]
> > (bus_add_driver+0x148/0x1f0)
> > [<803602bc>] (bus_add_driver) from [<8036130c>] (driver_register+0x78/0xf8)
> > [<8036130c>] (driver_register) from [<800088dc>]
> > (do_one_initcall+0xf8/0x144)
> > [<800088dc>] (do_one_initcall) from [<80da1c4c>]
> > (kernel_init_freeable+0x138/0x1d8)
> > [<80da1c4c>] (kernel_init_freeable) from [<8072725c>] (kernel_init+0x8/0xf0)
> > [<8072725c>] (kernel_init) from [<8000ecf8>] (ret_from_fork+0x14/0x3c)
> > Code: 10866003 1206601f 10633006 11e02312 (e5953000)
> > ---[ end trace f6f103bb73cc0503 ]---
> > note: swapper/0[1] exited with preempt_count 1
> > Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b"
> >
> > Because dma_alloc_coherent fail, it returns NULL, then
> > "if (!host->adma_table || !host->align_buffer)" will return true, then
> > dma_free_coherent trigger panic.
> >
> > After applying this patch, no panic and all work well, kernel log:
> > "
> > sdhci: Secure Digital Host Controller Interface driver
> > sdhci: Copyright(c) Pierre Ossman
> > sdhci-pltfm: SDHCI platform and OF driver helper
> > mmc0: Unable to allocate ADMA buffers. Falling back to standard DMA.
> > mmc0: no vqmmc regulator found
> > mmc0: SDHCI controller on 30b40000.usdhc [30b40000.usdhc] using DMA
> > "
> >
> > Signed-off-by: Peng Fan <van.free...@gmail.com>
>
> Acked-by: Adrian Hunter <adrian.hun...@intel.com>
>
> Looks like the problem started in v3.16 with:
>
> commit d1e49f77d7c7b75fdc022e1d46c1549bbc91c5b7
> Author: Russell King <rmk+ker...@arm.linux.org.uk>
> Date: Fri Apr 25 12:58:34 2014 +0100
>
> mmc: sdhci: convert ADMA descriptors to a coherent allocation
>
Will this patch be merged? Since more than 3 weeks passed, I did not see any
further comments about this patch.
>
>
> > ---
> > drivers/mmc/host/sdhci.c | 7 +++++--
> > 1 file changed, 5 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
> > index 706bb60..52e2327 100644
> > --- a/drivers/mmc/host/sdhci.c
> > +++ b/drivers/mmc/host/sdhci.c
> > @@ -2978,8 +2978,11 @@ int sdhci_add_host(struct sdhci_host *host)
> > GFP_KERNEL);
> > host->align_buffer = kmalloc(host->align_buffer_sz, GFP_KERNEL);
> > if (!host->adma_table || !host->align_buffer) {
> > - dma_free_coherent(mmc_dev(mmc), host->adma_table_sz,
> > - host->adma_table, host->adma_addr);
> > + if (host->adma_table)
> > + dma_free_coherent(mmc_dev(mmc),
> > + host->adma_table_sz,
> > + host->adma_table,
> > + host->adma_addr);
> > kfree(host->align_buffer);
> > pr_warn("%s: Unable to allocate ADMA buffers - falling
> > back to standard DMA\n",
> > mmc_hostname(mmc));
> >
>
Thanks,
Peng
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
