linux-toradex.git/arch/parisc/include, branch v4.7-rc2

parisc: Move die_if_kernel() prototype into traps.h header

2016-06-05T06:49:01+00:00

Signed-off-by: Helge Deller

parisc: Whitespace cleanups in unistd.h

2016-05-25T13:40:49+00:00

Clean up whitespaces and mark unused syscalls as such.

Signed-off-by: Helge Deller

parisc: Fix typos in eisa_eeprom.h

2016-05-22T19:57:40+00:00

Signed-off-by: Andrea Gelmini 
Signed-off-by: Helge Deller

parisc: Fix typo in ldcw.h

2016-05-22T19:56:37+00:00

Signed-off-by: Andrea Gelmini 
Signed-off-by: Helge Deller

parisc: Fix typo in pdc.h

2016-05-22T19:55:33+00:00

Signed-off-by: Andrea Gelmini 
Signed-off-by: Helge Deller

parisc: Update futex.h to match generic implementation

2016-05-22T19:48:54+00:00

The attached patch updates the parisc version of futex.h to match the
current generic implementation except for the spinlock code.

Signed-off-by: John David Anglin 
Signed-off-by: Helge Deller

parisc: Add native high-resolution sched_clock() implementation

2016-05-22T19:39:25+00:00

Add a native implementation for the sched_clock() function which utilizes the
processor-internal cycle counter (Control Register 16) as high-resolution time
source.

With this patch we now get much more fine-grained resolutions in various
in-kernel time measurements (e.g. when viewing the function tracing logs), and
probably a more accurate scheduling on SMP systems.

There are a few specific implementation details in this patch:

1. On a 32bit kernel we emulate the higher 32bits of the required 64-bit
resolution of sched_clock() by increasing a per-cpu counter at every
wrap-around of the 32bit cycle counter.

2. In a SMP system, the cycle counters of the various CPUs are not syncronized
(similiar to the TSC in a x86_64 system). To cope with this we define
HAVE_UNSTABLE_SCHED_CLOCK and let the upper layers do the adjustment work.

3. Since we need HAVE_UNSTABLE_SCHED_CLOCK, we need to provide a cmpxchg64()
function even on a 32-bit kernel.

4. A 64-bit SMP kernel which is started on a UP system will mark the
sched_clock() implementation as "stable", which means that we don't expect any
jumps in the returned counter. This is true because we then run only on one
CPU.

Signed-off-by: Helge Deller

parisc: Add ARCH_TRACEHOOK and regset support

2016-05-22T19:39:13+00:00

By adding TRACEHOOK support we now get a clean user interface to access
registers via PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and
PTRACE_SETFPREGS.

The user-visible regset struct user_regs_struct and user_fp_struct are
modelled similiar to x86 and can be accessed via PTRACE_GETREGSET.

Signed-off-by: Helge Deller

parisc: Add 64bit get_user() and put_user() for 32bit kernel

2016-05-22T19:39:04+00:00

Allow accessing 64-bit values in userspace from a 32-bit kernel.
The access is not atomic.

Signed-off-by: Helge Deller

parisc: Simplify and speed up get_user() and put_user()

2016-05-22T19:38:56+00:00

This patch simplifies the code for get_user() and put_user() a lot.

Instead of accessing kernel memory (%sr0) and userspace memory (%sr3)
hard-coded in the assembler instruction, we now preload %sr2 with either
%sr0 (for accessing KERNEL_DS) or with sr3 (to access USER_DS) and
use %sr2 in the load directly.

The generated code avoids a branch and speeds up execution by generating
less assembler instructions.

Signed-off-by: Helge Deller 
Tested-by: Rolf Eike Beer