On Mon, Sep 28, 2020 at 09:46:22AM +0200, Michael Walle wrote:
> Am 2020-09-28 09:29, schrieb Krzysztof Kozlowski:
> > On Mon, 28 Sep 2020 at 01:28, Vladimir Oltean <olte...@gmail.com> wrote:
> > >
> > > On Mon, Sep 28, 2020 at 12:43:36AM +0200, Michael Walle wrote:
> > > > Since commit 530b5affc675 ("spi: fsl-dspi: fix use-after-free in remove
> > > > path") this driver causes a kernel oops:
> > > >
> > > > [ 1.891065] Unable to handle kernel NULL pointer dereference at
> > > > virtual address 0000000000000080
> > > > [ 1.899889] Mem abort info:
> > > > [ 1.902692] ESR = 0x96000004
> > > > [ 1.905754] EC = 0x25: DABT (current EL), IL = 32 bits
> > > > [ 1.911089] SET = 0, FnV = 0
> > > > [ 1.914156] EA = 0, S1PTW = 0
> > > > [ 1.917303] Data abort info:
> > > > [ 1.920193] ISV = 0, ISS = 0x00000004
> > > > [ 1.924044] CM = 0, WnR = 0
> > > > [ 1.927022] [0000000000000080] user address but active_mm is swapper
> > > > [ 1.933403] Internal error: Oops: 96000004 [#1] PREEMPT SMP
> > > > [ 1.938995] Modules linked in:
> > > > [ 1.942060] CPU: 0 PID: 1 Comm: swapper/0 Not tainted
> > > > 5.9.0-rc6-next-20200925-00026-gae556cc74e28-dirty #94
> > > > [ 1.951838] Hardware name: Kontron SMARC-sAL28 (Single PHY) on SMARC
> > > > Eval 2.0 carrier (DT)
> > > > [ 1.960135] pstate: 40000005 (nZcv daif -PAN -UAO -TCO BTYPE=--)
> > > > [ 1.966168] pc : dspi_setup+0xc8/0x2e0
> > > > [ 1.969926] lr : dspi_setup+0xbc/0x2e0
> > > > [ 1.973684] sp : ffff80001139b930
> > > > [ 1.977005] x29: ffff80001139b930 x28: ffff00207a5d2000
> > > > [ 1.982338] x27: 0000000000000006 x26: ffff00207a44d410
> > > > [ 1.987669] x25: ffff002079c08100 x24: ffff00207a5d2400
> > > > [ 1.993000] x23: ffff00207a5d2600 x22: ffff800011169948
> > > > [ 1.998332] x21: ffff800010cbcd20 x20: ffff00207a58a800
> > > > [ 2.003663] x19: ffff00207a76b700 x18: 0000000000000010
> > > > [ 2.008994] x17: 0000000000000001 x16: 0000000000000019
> > > > [ 2.014326] x15: ffffffffffffffff x14: 0720072007200720
> > > > [ 2.019657] x13: 0720072007200720 x12: ffff8000111fc5e0
> > > > [ 2.024989] x11: 0000000000000003 x10: ffff8000111e45a0
> > > > [ 2.030320] x9 : 0000000000000000 x8 : ffff00207a76b780
> > > > [ 2.035651] x7 : 0000000000000000 x6 : 000000000000003f
> > > > [ 2.040982] x5 : 0000000000000040 x4 : ffff80001139b918
> > > > [ 2.046313] x3 : 0000000000000001 x2 : 64b62cc917af5100
> > > > [ 2.051643] x1 : 0000000000000000 x0 : 0000000000000000
> > > > [ 2.056973] Call trace:
> > > > [ 2.059425] dspi_setup+0xc8/0x2e0
> > > > [ 2.062837] spi_setup+0xcc/0x248
> > > > [ 2.066160] spi_add_device+0xb4/0x198
> > > > [ 2.069918] of_register_spi_device+0x250/0x370
> > > > [ 2.074462] spi_register_controller+0x4f4/0x770
> > > > [ 2.079094] dspi_probe+0x5bc/0x7b0
> > > > [ 2.082594] platform_drv_probe+0x5c/0xb0
> > > > [ 2.086615] really_probe+0xec/0x3c0
> > > > [ 2.090200] driver_probe_device+0x60/0xc0
> > > > [ 2.094308] device_driver_attach+0x7c/0x88
> > > > [ 2.098503] __driver_attach+0x60/0xe8
> > > > [ 2.102263] bus_for_each_dev+0x7c/0xd0
> > > > [ 2.106109] driver_attach+0x2c/0x38
> > > > [ 2.109692] bus_add_driver+0x194/0x1f8
> > > > [ 2.113538] driver_register+0x6c/0x128
> > > > [ 2.117385] __platform_driver_register+0x50/0x60
> > > > [ 2.122105] fsl_dspi_driver_init+0x24/0x30
> > > > [ 2.126302] do_one_initcall+0x54/0x2d0
> > > > [ 2.130149] kernel_init_freeable+0x1ec/0x258
> > > > [ 2.134520] kernel_init+0x1c/0x120
> > > > [ 2.138018] ret_from_fork+0x10/0x34
> > > > [ 2.141606] Code: 97e0b11d aa0003f3 b4000680 f94006e0 (f9404000)
> > > > [ 2.147723] ---[ end trace 26cf63e6cbba33a8 ]---
> > > > [ 2.152374] Kernel panic - not syncing: Attempted to kill init!
> > > > exitcode=0x0000000b
> > > > [ 2.160061] SMP: stopping secondary CPUs
> > > > [ 2.163999] Kernel Offset: disabled
> > > > [ 2.167496] CPU features: 0x0040022,20006008
> > > > [ 2.171777] Memory Limit: none
> > > > [ 2.174840] ---[ end Kernel panic - not syncing: Attempted to kill
> > > > init! exitcode=0x0000000b ]---
> > > >
> > > > This is because since this commit, the allocation of the drivers private
> > > > data is done explicitly and in this case spi_alloc_master() won't set
> > > > the
> > > > correct pointer.
> > > >
> > > > Fixes: 530b5affc675 ("spi: fsl-dspi: fix use-after-free in remove path")
> > > > Signed-off-by: Michael Walle <mich...@walle.cc>
> > > > ---
> > >
> > > Sascha, how did you test commit 530b5affc675?
> >
> > Hi,
> >
> > I just hit it on my Vybrid systems as well. It fails on every boot, so
> > I have doubts that it was actually tested. The fix was posted on 23rd
> > and applied within a few hours... also no time for anyone else to test
> > it.
>
> Mhh, given the benefit of the doubt, I could imagine that the allocs align
> up in a way, that the pointer is valid afterwards, no?
Or Sasha used generic evalkit board, where SPI by default might not have
any devices attached.
> > >
> > > > drivers/spi/spi-fsl-dspi.c | 2 ++
> > > > 1 file changed, 2 insertions(+)
> > > >
> > > > diff --git a/drivers/spi/spi-fsl-dspi.c b/drivers/spi/spi-fsl-dspi.c
> > > > index a939618f5e47..dd80be987bf9 100644
> > > > --- a/drivers/spi/spi-fsl-dspi.c
> > > > +++ b/drivers/spi/spi-fsl-dspi.c
> > > > @@ -1236,6 +1236,8 @@ static int dspi_probe(struct platform_device
> > > > *pdev)
> > > > if (!ctlr)
> > > > return -ENOMEM;
> > > >
> > > > + spi_controller_set_devdata(ctlr, dspi);
> >
> > Michael,
> >
> > How about moving here platform_set_drvdata(pdev, dspi) from the end of
> > the probe to keep them close to each other?
>
> Given that this patch has a fixes tag, I'd rather keep the changes to a
> minimum to avoid future conflicts.
>
> OTOH I don't know if its better if the "move platform_set_drvdata()" is in a
> seperate patch, conflict-wise.
Just looking at these two patches there are no worries:
1. The original patch does not target stable,
2. The fix is only for this cycle.
However it is likely that autosel will pick up these patches so your
questions are good. If I understand code correctly, these two calls
(spi_controller_set_devdata and platform_set_drvdata) can be reordered
freely between allocation and spi_register_controller(). Therefore
moving these calls around should not affect backporting to stable.
I would propose to keep these calls together, either at beginning or
just before spi_register_controller().
Best regards,
Krzysztof
>
> Any suggestions?
>
> -michael
