linux-stable.git/drivers/crypto, branch v4.10.6

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

2017-03-22T11:57:05+00:00

commit 28b62b1458685d8f68f67d9b2d511bf8fa32b746 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)
    [] (unwind_backtrace) from [] (show_stack+0x10/0x14)
    [] (show_stack) from [] (dump_stack+0x78/0x8c)
    [] (dump_stack) from [] (do_raw_spin_lock+0x11c/0x120)
    [] (do_raw_spin_lock) from [] (_raw_spin_lock_irqsave+0x20/0x28)
    [] (_raw_spin_lock_irqsave) from [] (s5p_aes_crypt+0x2c/0xb4)
    [] (s5p_aes_crypt) from [] (do_encrypt+0x78/0xb0 [lrw])
    [] (do_encrypt [lrw]) from [] (encrypt_done+0x24/0x54 [lrw])
    [] (encrypt_done [lrw]) from [] (s5p_aes_complete+0x60/0xcc)
    [] (s5p_aes_complete) from [] (s5p_aes_interrupt+0x134/0x1a0)
    [] (s5p_aes_interrupt) from [] (irq_thread_fn+0x1c/0x54)
    [] (irq_thread_fn) from [] (irq_thread+0x12c/0x1e0)
    [] (irq_thread) from [] (kthread+0x108/0x138)
    [] (kthread) from [] (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: Nathan Royce 
Signed-off-by: Krzysztof Kozlowski 
Signed-off-by: Herbert Xu 
Signed-off-by: Greg Kroah-Hartman

crypto: vmx - Use skcipher for xts fallback

2017-03-12T05:44:15+00:00

commit 5839f555fa576be57371686265206398d9ea1480 upstream.

Signed-off-by: Paulo Flabiano Smorigo 
Signed-off-by: Herbert Xu 
Signed-off-by: Greg Kroah-Hartman

crypto: vmx - Use skcipher for cbc fallback

2017-03-12T05:44:15+00:00

commit c96d0a1c47abd5c4fa544dcedb5fac4d020ac58b upstream.

Signed-off-by: Paulo Flabiano Smorigo 
Signed-off-by: Herbert Xu 
Signed-off-by: Greg Kroah-Hartman

Merge branch 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6

2017-02-06T22:16:23+00:00

Pull crypto fixes from Herbert Xu:

 - use-after-free in algif_aead

 - modular aesni regression when pcbc is modular but absent

 - bug causing IO page faults in ccp

 - double list add in ccp

 - NULL pointer dereference in qat (two patches)

 - panic in chcr

 - NULL pointer dereference in chcr

 - out-of-bound access in chcr

* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: chcr - Fix key length for RFC4106
  crypto: algif_aead - Fix kernel panic on list_del
  crypto: aesni - Fix failure when pcbc module is absent
  crypto: ccp - Fix double add when creating new DMA command
  crypto: ccp - Fix DMA operations when IOMMU is enabled
  crypto: chcr - Check device is allocated before use
  crypto: chcr - Fix panic on dma_unmap_sg
  crypto: qat - zero esram only for DH85x devices
  crypto: qat - fix bar discovery for c62x

crypto: chcr - Fix key length for RFC4106

2017-02-03T10:09:31+00:00

Check keylen before copying salt to avoid wrap around of Integer.

Signed-off-by: Harsh Jain 
Signed-off-by: Herbert Xu

crypto: ccp - Fix double add when creating new DMA command

2017-02-03T09:45:46+00:00

Eliminate a double-add by creating a new list to manage
command descriptors when created; move the descriptor to
the pending list when the command is submitted.

Cc: 
Signed-off-by: Gary R Hook 
Signed-off-by: Herbert Xu

crypto: ccp - Fix DMA operations when IOMMU is enabled

2017-02-03T09:45:45+00:00

An I/O page fault occurs when the IOMMU is enabled on a
system that supports the v5 CCP.  DMA operations use a
Request ID value that does not match what is expected by
the IOMMU, resulting in the I/O page fault.  Setting the
Request ID value to 0 corrects this issue.

Cc: 
Signed-off-by: Gary R Hook 
Signed-off-by: Herbert Xu

crypto: chcr - Check device is allocated before use

2017-02-03T09:45:44+00:00

Ensure dev is allocated for crypto uld context before using the device
for crypto operations.

Cc: 
Signed-off-by: Atul Gupta 
Signed-off-by: Herbert Xu

crypto: chcr - Fix panic on dma_unmap_sg

2017-02-03T09:45:44+00:00

Save DMA mapped sg list addresses to request context buffer.

Signed-off-by: Atul Gupta 
Signed-off-by: Herbert Xu

crypto: qat - zero esram only for DH85x devices

2017-02-02T13:54:53+00:00

Zero embedded ram in DH85x devices. This is not
needed for newer generations as it is done by HW.

Cc: 
Signed-off-by: Giovanni Cabiddu 
Signed-off-by: Herbert Xu