Commit 7814c9bd authored by Krzysztof Kozlowski's avatar Krzysztof Kozlowski Committed by Greg Kroah-Hartman

crypto: s5p-sss - Fix spinlock recursion on LRW(AES)

commit 28b62b14 upstream.

Running TCRYPT with LRW compiled causes spinlock recursion:

    testing speed of async lrw(aes) (lrw(ecb-aes-s5p)) encryption
    tcrypt: test 0 (256 bit key, 16 byte blocks): 19007 operations in 1 seconds (304112 bytes)
    tcrypt: test 1 (256 bit key, 64 byte blocks): 15753 operations in 1 seconds (1008192 bytes)
    tcrypt: test 2 (256 bit key, 256 byte blocks): 14293 operations in 1 seconds (3659008 bytes)
    tcrypt: test 3 (256 bit key, 1024 byte blocks): 11906 operations in 1 seconds (12191744 bytes)
    tcrypt: test 4 (256 bit key, 8192 byte blocks):
    BUG: spinlock recursion on CPU#1, irq/84-10830000/89
     lock: 0xeea99a68, .magic: dead4ead, .owner: irq/84-10830000/89, .owner_cpu: 1
    CPU: 1 PID: 89 Comm: irq/84-10830000 Not tainted 4.11.0-rc1-00001-g897ca6d0800d #559
    Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
    [<c010e1ec>] (unwind_backtrace) from [<c010ae1c>] (show_stack+0x10/0x14)
    [<c010ae1c>] (show_stack) from [<c03449c0>] (dump_stack+0x78/0x8c)
    [<c03449c0>] (dump_stack) from [<c015de68>] (do_raw_spin_lock+0x11c/0x120)
    [<c015de68>] (do_raw_spin_lock) from [<c0720110>] (_raw_spin_lock_irqsave+0x20/0x28)
    [<c0720110>] (_raw_spin_lock_irqsave) from [<c0572ca0>] (s5p_aes_crypt+0x2c/0xb4)
    [<c0572ca0>] (s5p_aes_crypt) from [<bf1d8aa4>] (do_encrypt+0x78/0xb0 [lrw])
    [<bf1d8aa4>] (do_encrypt [lrw]) from [<bf1d8b00>] (encrypt_done+0x24/0x54 [lrw])
    [<bf1d8b00>] (encrypt_done [lrw]) from [<c05732a0>] (s5p_aes_complete+0x60/0xcc)
    [<c05732a0>] (s5p_aes_complete) from [<c0573440>] (s5p_aes_interrupt+0x134/0x1a0)
    [<c0573440>] (s5p_aes_interrupt) from [<c01667c4>] (irq_thread_fn+0x1c/0x54)
    [<c01667c4>] (irq_thread_fn) from [<c0166a98>] (irq_thread+0x12c/0x1e0)
    [<c0166a98>] (irq_thread) from [<c0136a28>] (kthread+0x108/0x138)
    [<c0136a28>] (kthread) from [<c0107778>] (ret_from_fork+0x14/0x3c)

Interrupt handling routine was calling req->base.complete() under
spinlock.  In most cases this wasn't fatal but when combined with some
of the cipher modes (like LRW) this caused recursion - starting the new
encryption (s5p_aes_crypt()) while still holding the spinlock from
previous round (s5p_aes_complete()).

Beside that, the s5p_aes_interrupt() error handling path could execute
two completions in case of error for RX and TX blocks.

Rewrite the interrupt handling routine and the completion by:

1. Splitting the operations on scatterlist copies from
   s5p_aes_complete() into separate s5p_sg_done(). This still should be
   done under lock.
   The s5p_aes_complete() now only calls req->base.complete() and it has
   to be called outside of lock.

2. Moving the s5p_aes_complete() out of spinlock critical sections.
   In interrupt service routine s5p_aes_interrupts(), it appeared in few
   places, including error paths inside other functions called from ISR.
   This code was not so obvious to read so simplify it by putting the
   s5p_aes_complete() only within ISR level.
Reported-by: default avatarNathan Royce <nroycea+kernel@gmail.com>
Signed-off-by: default avatarKrzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: default avatarHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
parent 4310604e
...@@ -270,7 +270,7 @@ static void s5p_sg_copy_buf(void *buf, struct scatterlist *sg, ...@@ -270,7 +270,7 @@ static void s5p_sg_copy_buf(void *buf, struct scatterlist *sg,
scatterwalk_done(&walk, out, 0); scatterwalk_done(&walk, out, 0);
} }
static void s5p_aes_complete(struct s5p_aes_dev *dev, int err) static void s5p_sg_done(struct s5p_aes_dev *dev)
{ {
if (dev->sg_dst_cpy) { if (dev->sg_dst_cpy) {
dev_dbg(dev->dev, dev_dbg(dev->dev,
...@@ -281,8 +281,11 @@ static void s5p_aes_complete(struct s5p_aes_dev *dev, int err) ...@@ -281,8 +281,11 @@ static void s5p_aes_complete(struct s5p_aes_dev *dev, int err)
} }
s5p_free_sg_cpy(dev, &dev->sg_src_cpy); s5p_free_sg_cpy(dev, &dev->sg_src_cpy);
s5p_free_sg_cpy(dev, &dev->sg_dst_cpy); s5p_free_sg_cpy(dev, &dev->sg_dst_cpy);
}
/* holding a lock outside */ /* Calls the completion. Cannot be called with dev->lock hold. */
static void s5p_aes_complete(struct s5p_aes_dev *dev, int err)
{
dev->req->base.complete(&dev->req->base, err); dev->req->base.complete(&dev->req->base, err);
dev->busy = false; dev->busy = false;
} }
...@@ -368,51 +371,44 @@ static int s5p_set_indata(struct s5p_aes_dev *dev, struct scatterlist *sg) ...@@ -368,51 +371,44 @@ static int s5p_set_indata(struct s5p_aes_dev *dev, struct scatterlist *sg)
} }
/* /*
* Returns true if new transmitting (output) data is ready and its * Returns -ERRNO on error (mapping of new data failed).
* address+length have to be written to device (by calling * On success returns:
* s5p_set_dma_outdata()). False otherwise. * - 0 if there is no more data,
* - 1 if new transmitting (output) data is ready and its address+length
* have to be written to device (by calling s5p_set_dma_outdata()).
*/ */
static bool s5p_aes_tx(struct s5p_aes_dev *dev) static int s5p_aes_tx(struct s5p_aes_dev *dev)
{ {
int err = 0; int ret = 0;
bool ret = false;
s5p_unset_outdata(dev); s5p_unset_outdata(dev);
if (!sg_is_last(dev->sg_dst)) { if (!sg_is_last(dev->sg_dst)) {
err = s5p_set_outdata(dev, sg_next(dev->sg_dst)); ret = s5p_set_outdata(dev, sg_next(dev->sg_dst));
if (err) if (!ret)
s5p_aes_complete(dev, err); ret = 1;
else
ret = true;
} else {
s5p_aes_complete(dev, err);
dev->busy = true;
tasklet_schedule(&dev->tasklet);
} }
return ret; return ret;
} }
/* /*
* Returns true if new receiving (input) data is ready and its * Returns -ERRNO on error (mapping of new data failed).
* address+length have to be written to device (by calling * On success returns:
* s5p_set_dma_indata()). False otherwise. * - 0 if there is no more data,
* - 1 if new receiving (input) data is ready and its address+length
* have to be written to device (by calling s5p_set_dma_indata()).
*/ */
static bool s5p_aes_rx(struct s5p_aes_dev *dev) static int s5p_aes_rx(struct s5p_aes_dev *dev/*, bool *set_dma*/)
{ {
int err; int ret = 0;
bool ret = false;
s5p_unset_indata(dev); s5p_unset_indata(dev);
if (!sg_is_last(dev->sg_src)) { if (!sg_is_last(dev->sg_src)) {
err = s5p_set_indata(dev, sg_next(dev->sg_src)); ret = s5p_set_indata(dev, sg_next(dev->sg_src));
if (err) if (!ret)
s5p_aes_complete(dev, err); ret = 1;
else
ret = true;
} }
return ret; return ret;
...@@ -422,33 +418,73 @@ static irqreturn_t s5p_aes_interrupt(int irq, void *dev_id) ...@@ -422,33 +418,73 @@ static irqreturn_t s5p_aes_interrupt(int irq, void *dev_id)
{ {
struct platform_device *pdev = dev_id; struct platform_device *pdev = dev_id;
struct s5p_aes_dev *dev = platform_get_drvdata(pdev); struct s5p_aes_dev *dev = platform_get_drvdata(pdev);
bool set_dma_tx = false; int err_dma_tx = 0;
bool set_dma_rx = false; int err_dma_rx = 0;
bool tx_end = false;
unsigned long flags; unsigned long flags;
uint32_t status; uint32_t status;
int err;
spin_lock_irqsave(&dev->lock, flags); spin_lock_irqsave(&dev->lock, flags);
/*
* Handle rx or tx interrupt. If there is still data (scatterlist did not
* reach end), then map next scatterlist entry.
* In case of such mapping error, s5p_aes_complete() should be called.
*
* If there is no more data in tx scatter list, call s5p_aes_complete()
* and schedule new tasklet.
*/
status = SSS_READ(dev, FCINTSTAT); status = SSS_READ(dev, FCINTSTAT);
if (status & SSS_FCINTSTAT_BRDMAINT) if (status & SSS_FCINTSTAT_BRDMAINT)
set_dma_rx = s5p_aes_rx(dev); err_dma_rx = s5p_aes_rx(dev);
if (status & SSS_FCINTSTAT_BTDMAINT)
set_dma_tx = s5p_aes_tx(dev); if (status & SSS_FCINTSTAT_BTDMAINT) {
if (sg_is_last(dev->sg_dst))
tx_end = true;
err_dma_tx = s5p_aes_tx(dev);
}
SSS_WRITE(dev, FCINTPEND, status); SSS_WRITE(dev, FCINTPEND, status);
/* if (err_dma_rx < 0) {
* Writing length of DMA block (either receiving or transmitting) err = err_dma_rx;
* will start the operation immediately, so this should be done goto error;
* at the end (even after clearing pending interrupts to not miss the }
* interrupt). if (err_dma_tx < 0) {
*/ err = err_dma_tx;
if (set_dma_tx) goto error;
s5p_set_dma_outdata(dev, dev->sg_dst); }
if (set_dma_rx)
s5p_set_dma_indata(dev, dev->sg_src); if (tx_end) {
s5p_sg_done(dev);
spin_unlock_irqrestore(&dev->lock, flags);
s5p_aes_complete(dev, 0);
dev->busy = true;
tasklet_schedule(&dev->tasklet);
} else {
/*
* Writing length of DMA block (either receiving or
* transmitting) will start the operation immediately, so this
* should be done at the end (even after clearing pending
* interrupts to not miss the interrupt).
*/
if (err_dma_tx == 1)
s5p_set_dma_outdata(dev, dev->sg_dst);
if (err_dma_rx == 1)
s5p_set_dma_indata(dev, dev->sg_src);
spin_unlock_irqrestore(&dev->lock, flags);
}
return IRQ_HANDLED;
error:
s5p_sg_done(dev);
spin_unlock_irqrestore(&dev->lock, flags); spin_unlock_irqrestore(&dev->lock, flags);
s5p_aes_complete(dev, err);
return IRQ_HANDLED; return IRQ_HANDLED;
} }
...@@ -597,8 +633,9 @@ static void s5p_aes_crypt_start(struct s5p_aes_dev *dev, unsigned long mode) ...@@ -597,8 +633,9 @@ static void s5p_aes_crypt_start(struct s5p_aes_dev *dev, unsigned long mode)
s5p_unset_indata(dev); s5p_unset_indata(dev);
indata_error: indata_error:
s5p_aes_complete(dev, err); s5p_sg_done(dev);
spin_unlock_irqrestore(&dev->lock, flags); spin_unlock_irqrestore(&dev->lock, flags);
s5p_aes_complete(dev, err);
} }
static void s5p_tasklet_cb(unsigned long data) static void s5p_tasklet_cb(unsigned long data)
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment