kernel/linux.git/arch/s390/lib, branch v4.13-rc6

s390/uaccess: use sane length for __strncpy_from_user()

2017-05-09T08:43:24+00:00

The average string that is copied from user space to kernel space is rather short. E.g. booting a system involves about 50.000 strncpy_from_user() calls where the NULL terminated string has an average size of 27 bytes. By default our s390 specific strncpy_from_user() implementation however copies up to 4096 bytes, which is a waste of cpu cycles and cache lines. Therefore reduce the default length to L1_CACHE_BYTES (256 bytes), which also reduces the average execution time of strncpy_from_user() by 30-40%. Alternatively we could have switched to the generic strncpy_from_user() implementation, however it turned out that that variant would be slower than the now optimized s390 variant. Reported-by: Al Viro Reported-by: Linus Torvalds Signed-off-by: Heiko Carstens Reviewed-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky

s390/uprobes: fix compile for !KPROBES

2017-05-03T07:08:57+00:00

Fix the following compile error(s) if CONFIG_KPROBES is disabled: arch/s390/kernel/uprobes.c:79:14: error: implicit declaration of function 'probe_get_fixup_type' arch/s390/kernel/uprobes.c:87:14: error: 'FIXUP_PSW_NORMAL' undeclared (first use in this function) Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky

Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux

2017-05-02T16:50:09+00:00

Pull s390 updates from Martin Schwidefsky: - three merges for KVM/s390 with changes for vfio-ccw and cpacf. The patches are included in the KVM tree as well, let git sort it out. - add the new 'trng' random number generator - provide the secure key verification API for the pkey interface - introduce the z13 cpu counters to perf - add a new system call to set up the guarded storage facility - simplify TASK_SIZE and arch_get_unmapped_area - export the raw STSI data related to CPU topology to user space - ... and the usual churn of bug-fixes and cleanups. * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (74 commits) s390/crypt: use the correct module alias for paes_s390. s390/cpacf: Introduce kma instruction s390/cpacf: query instructions use unique parameters for compatibility with KMA s390/trng: Introduce s390 TRNG device driver. s390/crypto: Provide s390 specific arch random functionality. s390/crypto: Add new subfunctions to the cpacf PRNO function. s390/crypto: Renaming PPNO to PRNO. s390/pageattr: avoid unnecessary page table splitting s390/mm: simplify arch_get_unmapped_area[_topdown] s390/mm: make TASK_SIZE independent from the number of page table levels s390/gs: add regset for the guarded storage broadcast control block s390/kvm: Add use_cmma field to mm_context_t s390/kvm: Add PGSTE manipulation functions vfio: ccw: improve error handling for vfio_ccw_mdev_remove vfio: ccw: remove unnecessary NULL checks of a pointer s390/spinlock: remove compare and delay instruction s390/spinlock: use atomic primitives for spinlocks s390/cpumf: simplify detection of guest samples s390/pci: remove forward declaration s390/pci: increase the PCI_NR_FUNCTIONS default ...

s390/spinlock: remove compare and delay instruction

2017-04-12T06:43:33+00:00

The CAD instruction never worked quite as expected for the spinlock code. It has been disabled by default with git commit 61b0b01686d48220, if the "cad" kernel parameter is specified it is enabled for both user space and the spinlock code. Leave the option to enable the instruction for user space but remove it from the spinlock code. Signed-off-by: Martin Schwidefsky

s390/spinlock: use atomic primitives for spinlocks

2017-04-12T06:43:33+00:00

Add a couple more __atomic_xxx function to atomic_ops.h and use them to replace the compare-and-swap inlines in the spinlock code. This changes the type of the lock value from unsigned int to int. Signed-off-by: Martin Schwidefsky

s390: get rid of zeroing, switch to RAW_COPY_USER

2017-03-30T14:47:28+00:00

[folded a fix from Martin] Signed-off-by: Al Viro

s390: replace ACCESS_ONCE with READ_ONCE

2017-02-17T06:40:46+00:00

Remove the last places of ACCESS_ONCE in s390 code. Signed-off-by: Christian Borntraeger Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky

s390: Audit and remove any remaining unnecessary uses of module.h

2017-02-17T06:40:41+00:00

Historically a lot of these existed because we did not have a distinction between what was modular code and what was providing support to modules via EXPORT_SYMBOL and friends. That changed when we forked out support for the latter into the export.h file. This means we should be able to reduce the usage of module.h in code that is obj-y Makefile or bool Kconfig. The advantage in doing so is that module.h itself sources about 15 other headers; adding significantly to what we feed cpp, and it can obscure what headers we are effectively using. Since module.h was the source for init.h (for __init) and for export.h (for EXPORT_SYMBOL) we consider each change instance for the presence of either and replace as needed. An instance where module_param was used without moduleparam.h was also fixed, as well as implicit use of ptrace.h and string.h headers. Signed-off-by: Paul Gortmaker Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky

s390/lib: improve memmove, memset and memcpy

2017-01-16T06:27:51+00:00

Improve the memmove implementation to save one instruction and use better label names. Also use better label names for the memset and memcpy implementations so everything looks consistent. Suggested-by: Jens Remus Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky

s390/lib: add missing memory barriers to string inline assemblies

2016-12-14T15:33:41+00:00

We have a couple of inline assemblies like memchr() and strlen() that read from memory, but tell the compiler only they need the addresses of the strings they access. This allows the compiler to omit the initialization of such strings and therefore generate broken code. Add the missing memory barrier to all string related inline assemblies to fix this potential issue. It looks like the compiler currently does not generate broken code due to these bugs. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky