The poll() changes were not well thought out, and completely
unexplained. They also caused a huge performance regression, because
"->poll()" was no longer a trivial file operation that just called down
to the underlying file operations, but instead did at least two indirect
calls.
Indirect calls are sadly slow now with the Spectre mitigation, but the
performance problem could at least be largely mitigated by changing the
"->get_poll_head()" operation to just have a per-file-descriptor pointer
to the poll head instead. That gets rid of one of the new indirections.
But that doesn't fix the new complexity that is completely unwarranted
for the regular case. The (undocumented) reason for the poll() changes
was some alleged AIO poll race fixing, but we don't make the common case
slower and more complex for some uncommon special case, so this all
really needs way more explanations and most likely a fundamental
redesign.
[ This revert is a revert of about 30 different commits, not reverted
individually because that would just be unnecessarily messy - Linus ]
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull crypto fixes from Herbert Xu:
- Fix use after free in chtls
- Fix RBP breakage in sha3
- Fix use after free in hwrng_unregister
- Fix overread in morus640
- Move sleep out of kernel_neon in arm64/aes-blk
* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
hwrng: core - Always drop the RNG in hwrng_unregister()
crypto: morus640 - Fix out-of-bounds access
crypto: don't optimize keccakf()
crypto: arm64/aes-blk - fix and move skcipher_walk_done out of kernel_neon_begin, _end
crypto: chtls - use after free in chtls_pt_recvmsg()
enable_best_rng() is used in hwrng_unregister() to switch away from the
currently active RNG, if that is the one currently being removed.
However enable_best_rng() might fail, if the next RNG's init routine
fails. In that case enable_best_rng() will return an error code and
the currently active RNG will remain active.
After unregistering this might lead to crashes due to use-after-free.
Fix this by dropping the currently active RNG, if enable_best_rng()
failed. This will result in no RNG to be active, if the next-best
one failed to initialize.
This problem was introduced by 142a27f0a7
Fixes: 142a27f0a7 ("hwrng: core - Reset user selected rng by...")
Reported-by: Wirz <spam@lukas-wirz.de>
Tested-by: Wirz <spam@lukas-wirz.de>
Signed-off-by: Michael Büsch <m@bues.ch>
Cc: stable@vger.kernel.org
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Pull sparc updates from David Miller:
- a FPE signal fix that was also merged upstream
- privileged ADI driver from Tom Hromatka
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
sparc: fix compat siginfo ABI regression
selftests: sparc64: char: Selftest for privileged ADI driver
char: sparc64: Add privileged ADI driver
Pull TPM updates from James Morris:
"From Jarkko:
This purely a bug fix release.
The only major change is to move the event log code to its own
subdirectory because there starts to be so much of it"
* 'next-tpm' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
tpm: fix race condition in tpm_common_write()
tpm: reduce polling time to usecs for even finer granularity
tpm: replace kmalloc() + memcpy() with kmemdup()
tpm: replace kmalloc() + memcpy() with kmemdup()
tpm: fix use after free in tpm2_load_context()
tpm: reduce poll sleep time in tpm_transmit()
tpm_tis: verify locality released before returning from release_locality
tpm: tpm_crb: relinquish locality on error path.
tpm/st33zp24: Fix spelling mistake in macro ST33ZP24_TISREGISTER_UKNOWN
tpm: Move eventlog declarations to its own header
tpm: Move shared eventlog functions to common.c
tpm: Move eventlog files to a subdirectory
tpm: Add explicit endianness cast
tpm: st33zp24: remove redundant null check on chip
tpm: move the delay_msec increment after sleep in tpm_transmit()
Pull IPMI updates from Corey Minyard:
"It's been a busy release for the IPMI driver. Some notable changes:
- A user was running into timeout issues doing maintenance commands
over the IPMB network behind an IPMI controller.
Extend the maintenance mode concept to messages over IPMB and allow
the timeouts to be tuned.
- Lots of cleanup, style fixing, some bugfixes, and such.
- At least one user was having trouble with the way the IPMI driver
would lock the i2c driver module it used.
The IPMI driver was not designed for hotplug. However, hotplug is a
reality now, so the IPMI driver was modified to support hotplug.
- The proc interface code is now completely removed. Long live sysfs!"
* tag 'for-linus-4.18' of git://github.com/cminyard/linux-ipmi: (35 commits)
ipmi: Properly release srcu locks on error conditions
ipmi: NPCM7xx KCS BMC: enable interrupt to the host
ipmi:bt: Set the timeout before doing a capabilities check
ipmi: Remove the proc interface
ipmi_ssif: Fix uninitialized variable issue
ipmi: add an NPCM7xx KCS BMC driver
ipmi_si: Clean up shutdown a bit
ipmi_si: Rename intf_num to si_num
ipmi: Remove smi->intf checks
ipmi_ssif: Get rid of unused intf_num
ipmi: Get rid of ipmi_user_t and ipmi_smi_t in include files
ipmi: ipmi_unregister_smi() cannot fail, have it return void
ipmi_devintf: Add an error return on invalid ioctls
ipmi: Remove usecount function from interfaces
ipmi_ssif: Remove usecount handling
ipmi: Remove condition on interface shutdown
ipmi_ssif: Convert over to a shutdown handler
ipmi_si: Convert over to a shutdown handler
ipmi: Rework locking and shutdown for hot remove
ipmi: Fix some counter issues
...
Pull char/misc driver updates from Greg KH:
"Here is the "big" char and misc driver patches for 4.18-rc1.
It's not a lot of stuff here, but there are some highlights:
- coreboot driver updates
- soundwire driver updates
- android binder updates
- fpga big sync, mostly documentation
- lots of minor driver updates
All of these have been in linux-next for a while with no reported
issues"
* tag 'char-misc-4.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc: (81 commits)
vmw_balloon: fixing double free when batching mode is off
MAINTAINERS: Add driver-api/fpga path
fpga: clarify that unregister functions also free
documentation: fpga: move fpga-region.txt to driver-api
documentation: fpga: add bridge document to driver-api
documentation: fpga: move fpga-mgr.txt to driver-api
Documentation: fpga: move fpga overview to driver-api
fpga: region: kernel-doc fixes
fpga: bridge: kernel-doc fixes
fpga: mgr: kernel-doc fixes
fpga: use SPDX
fpga: region: change api, add fpga_region_create/free
fpga: bridge: change api, don't use drvdata
fpga: manager: change api, don't use drvdata
fpga: region: don't use drvdata in common fpga code
Drivers: hv: vmbus: Removed an unnecessary cast from void *
ver_linux: Drop redundant calls to system() to test if file is readable
ver_linux: Move stderr redirection from function parameter to function body
misc: IBM Virtual Management Channel Driver (VMC)
rpmsg: Correct support for MODULE_DEVICE_TABLE()
...
SPARC M7 and newer processors utilize ADI to version and
protect memory. This driver is capable of reading/writing
ADI/MCD versions from privileged user space processes.
Addresses in the adi file are mapped linearly to physical
memory at a ratio of 1:adi_blksz. Thus, a read (or write)
of offset K in the file operates upon the ADI version at
physical address K * adi_blksz. The version information
is encoded as one version per byte. Intended consumers
are makedumpfile and crash.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Reviewed-by: Khalid Aziz <khalid.aziz@oracle.com>
Reviewed-by: Shannon Nelson <shannon.nelson@oracle.com>
Reviewed-by: Anthony Yznaga <anthony.yznaga@oracle.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull aio updates from Al Viro:
"Majority of AIO stuff this cycle. aio-fsync and aio-poll, mostly.
The only thing I'm holding back for a day or so is Adam's aio ioprio -
his last-minute fixup is trivial (missing stub in !CONFIG_BLOCK case),
but let it sit in -next for decency sake..."
* 'work.aio-1' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (46 commits)
aio: sanitize the limit checking in io_submit(2)
aio: fold do_io_submit() into callers
aio: shift copyin of iocb into io_submit_one()
aio_read_events_ring(): make a bit more readable
aio: all callers of aio_{read,write,fsync,poll} treat 0 and -EIOCBQUEUED the same way
aio: take list removal to (some) callers of aio_complete()
aio: add missing break for the IOCB_CMD_FDSYNC case
random: convert to ->poll_mask
timerfd: convert to ->poll_mask
eventfd: switch to ->poll_mask
pipe: convert to ->poll_mask
crypto: af_alg: convert to ->poll_mask
net/rxrpc: convert to ->poll_mask
net/iucv: convert to ->poll_mask
net/phonet: convert to ->poll_mask
net/nfc: convert to ->poll_mask
net/caif: convert to ->poll_mask
net/bluetooth: convert to ->poll_mask
net/sctp: convert to ->poll_mask
net/tipc: convert to ->poll_mask
...
There is a race condition in tpm_common_write function allowing
two threads on the same /dev/tpm<N>, or two different applications
on the same /dev/tpmrm<N> to overwrite each other commands/responses.
Fixed this by taking the priv->buffer_mutex early in the function.
Also converted the priv->data_pending from atomic to a regular size_t
type. There is no need for it to be atomic since it is only touched
under the protection of the priv->buffer_mutex.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
The big change is that random_read_wait and random_write_wait are merged
into a single waitqueue that uses keyed wakeups. Because wait_event_*
doesn't know about that this will lead to occassional spurious wakeups
in _random_read and add_hwgenerator_randomness, but wait_event_* is
designed to handle these and were are not in a a hot path there.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Original kcs_bmc_npcm7xx.c was missing enabling to send interrupt to the
host on writes to output buffer.
This patch fixes it by setting the bits that enables the generation of
IRQn events by hardware control based on the status of the OBF flag.
Signed-off-by: Avi Fishman <AviFishman70@gmail.com>
Reviewed-by: Haiyue Wang <haiyue.wang@linux.intel.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
There was one place where the timeout value for an operation was
not being set, if a capabilities request was done from idle. Move
the timeout value setting to before where that change might be
requested.
IMHO the cause here is the invisible returns in the macros. Maybe
that's a job for later, though.
Reported-by: Nordmark Claes <Claes.Nordmark@tieto.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Cc: stable@vger.kernel.org
The TPM burstcount and status commands are supposed to return very
quickly [2][3]. This patch further reduces the TPM poll sleep time to usecs
in get_burstcount() and wait_for_tpm_stat() by calling usleep_range()
directly.
After this change, performance on a system[1] with a TPM 1.2 with an 8 byte
burstcount for 1000 extends improved from ~10.7 sec to ~7 sec.
[1] All tests are performed on an x86 based, locked down, single purpose
closed system. It has Infineon TPM 1.2 using LPC Bus.
[2] From the TCG Specification "TCG PC Client Specific TPM Interface
Specification (TIS), Family 1.2":
"NOTE : It takes roughly 330 ns per byte transfer on LPC. 256 bytes would
take 84 us, which is a long time to stall the CPU. Chipsets may not be
designed to post this much data to LPC; therefore, the CPU itself is
stalled for much of this time. Sending 1 kB would take 350 μs. Therefore,
even if the TPM_STS_x.burstCount field is a high value, software SHOULD
be interruptible during this period."
[3] From the TCG Specification 2.0, "TCG PC Client Platform TPM Profile
(PTP) Specification":
"It takes roughly 330 ns per byte transfer on LPC. 256 bytes would take
84 us. Chipsets may not be designed to post this much data to LPC;
therefore, the CPU itself is stalled for much of this time. Sending 1 kB
would take 350 us. Therefore, even if the TPM_STS_x.burstCount field is a
high value, software should be interruptible during this period. For SPI,
assuming 20MHz clock and 64-byte transfers, it would take about 120 usec
to move 256B of data. Sending 1kB would take about 500 usec. If the
transactions are done using 4 bytes at a time, then it would take about
1 msec. to transfer 1kB of data."
Signed-off-by: Nayna Jain <nayna@linux.vnet.ibm.com>
Reviewed-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Acked-by: Jay Freyensee <why2jjj.linux@gmail.com>
Tested-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Just set up the show callback in the tty_operations, and use
proc_create_single_data to create the file without additional
boilerplace code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Variants of proc_create{,_data} that directly take a seq_file show
callback and drastically reduces the boilerplate code in the callers.
All trivial callers converted over.
Signed-off-by: Christoph Hellwig <hch@lst.de>
