Commit Graph

76 Commits

Author SHA1 Message Date
Max Filippov
9d9043f6a8 xtensa: clean up system_call/xtensa_rt_sigreturn interaction
system_call assembly code always pushes pointer to struct pt_regs as the
last additional parameter for all system calls. The only user of this
feature is xtensa_rt_sigreturn.
Avoid this special case. Define xtensa_rt_sigreturn as accepting no
argiments. Use current_pt_regs to get pointer to struct pt_regs in
xtensa_rt_sigreturn. Don't pass additional parameter from system_call
code.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-11-29 19:37:12 -08:00
Max Filippov
02ce94c229 xtensa: fix system_call interaction with ptrace
Don't overwrite return value if system call was cancelled at entry by
ptrace. Return status code from do_syscall_trace_enter so that
pt_regs::syscall doesn't need to be changed to skip syscall.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-11-29 15:47:54 -08:00
Max Filippov
ba9c1d6599 xtensa: rearrange syscall tracing
system_call saves and restores syscall number across system call to make
clone and execv entry and exit tracing match. This complicates things
when syscall code may be changed by ptrace.
Preserve syscall code in copy_thread and start_thread directly instead of
doing tricks in system_call.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-11-29 14:39:12 -08:00
Valentin Schneider
8b5d7e5242 xtensa: entry: Remove unneeded need_resched() loop
Since the enabling and disabling of IRQs within preempt_schedule_irq()
is contained in a need_resched() loop, we don't need the outer arch
code loop.

Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Message-Id: <20190923143620.29334-10-valentin.schneider@arm.com>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-11-26 11:33:39 -08:00
Max Filippov
09f8a6db20 xtensa: add support for call0 ABI in userspace
Provide a Kconfig choice to select whether only the default ABI, only
call0 ABI or both are supported. The default for XEA2 is windowed, but
it may change for XEA3. Call0 only runs userspace with PS.WOE disabled.
Supporting both windowed and call0 ABIs is tricky, as there's no
indication in the ELF binaries which ABI they use. So it is done by
probing: each process is started with PS.WOE disabled, but the handler
of an illegal instruction exception taken with PS.WOE retries faulting
instruction after enabling PS.WOE. It must happen before any signal is
delivered to the process, otherwise it may be delivered incorrectly.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-09-01 13:11:57 -07:00
Max Filippov
9e1e41c447 xtensa: clean up PS_WOE_BIT usage
PS_WOE_BIT is mainly used to generate PS.WOE mask in the code. Introduce
PS_WOE_MASK macro and use it instead.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-09-01 13:11:57 -07:00
Max Filippov
d6d5f19e21 xtensa: abstract 'entry' and 'retw' in assembly code
Provide abi_entry, abi_entry_default, abi_ret and abi_ret_default macros
that allocate aligned stack frame in windowed and call0 ABIs.
Provide XTENSA_SPILL_STACK_RESERVE macro that specifies required stack
frame size when register spilling is involved.
Replace all uses of 'entry' and 'retw' with the above macros.
This makes most of the xtensa assembly code ready for XEA3 and call0 ABI.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-07-08 10:04:48 -07:00
Max Filippov
2201f31f2c xtensa: use actual syscall number in do_syscall_trace_leave
Syscall may alter pt_regs structure passed to it, resulting in a
mismatch between syscall entry end syscall exit entries in the ftrace.
Temporary restore syscall field of the pt_regs for the duration of
do_syscall_trace_leave.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2019-04-03 23:18:36 -07:00
Max Filippov
3aee3e25de xtensa: call do_syscall_trace_{enter,leave} selectively
Check whether calls to do_syscall_trace_{enter,leave} are necessary in
the system_call function. Define _TIF_WORK_MASK to a bitmask of flags
that reuire the calls. Fix comment.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2018-12-17 13:48:21 -08:00
Max Filippov
6a986984b6 xtensa: use NO_SYSCALL instead of -1
For the sake of clarity define macro NO_SYSCALL and use it for
setting/checking struct pt_regs::syscall field.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2018-12-17 13:48:21 -08:00
Max Filippov
8ed3a5426b xtensa: drop fast_syscall_kernel
There must be no xtensa-specific syscalls from the kernel code: register
spilling uses call+entry sequence and atomics have proper function
implementations. Drop fast_syscall_xtensa.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2018-12-03 01:39:22 -08:00
Firoz Khan
a86067f587 xtensa: add __NR_syscalls along with __NR_syscall_count
__NR_syscall_count macro holds the number of system call
exist in xtensa architecture. We have to change the value
of __NR_syscall_count, if we add or delete a system call.

One of the patch in this patch series has a script which
will generate a uapi header based on syscall.tbl file.
The syscall.tbl file contains the total number of system
calls information. So we have two option to update __NR-
_syscall_count value.

1. Update __NR_syscall_count in asm/unistd.h manually by
   counting the no.of system calls. No need to update __NR-
   _syscall_count until we either add a new system call or
   delete existing system call.

2. We can keep this feature it above mentioned script,
   that will count the number of syscalls and keep it in
   a generated file. In this case we don't need to expli-
   citly update __NR_syscall_count in asm/unistd.h file.

The 2nd option will be the recommended one. For that, I
added the __NR_syscalls macro in uapi/asm/unistd.h.
The macro __NR_syscalls also added for making the name
convention same across all architecture. While __NR_syscalls
isn't strictly part of the uapi, having it as part of the
generated header to simplifies the implementation. We also
need to enclose this macro with #ifdef __KERNEL__ to avoid
side effects.

Signed-off-by: Firoz Khan <firoz.khan@linaro.org>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
[Max: Drop __NR_syscall_count completely, use __NR_syscalls instead]
2018-12-02 23:45:21 -08:00
Linus Torvalds
050e9baa9d Kbuild: rename CC_STACKPROTECTOR[_STRONG] config variables
The changes to automatically test for working stack protector compiler
support in the Kconfig files removed the special STACKPROTECTOR_AUTO
option that picked the strongest stack protector that the compiler
supported.

That was all a nice cleanup - it makes no sense to have the AUTO case
now that the Kconfig phase can just determine the compiler support
directly.

HOWEVER.

It also meant that doing "make oldconfig" would now _disable_ the strong
stackprotector if you had AUTO enabled, because in a legacy config file,
the sane stack protector configuration would look like

  CONFIG_HAVE_CC_STACKPROTECTOR=y
  # CONFIG_CC_STACKPROTECTOR_NONE is not set
  # CONFIG_CC_STACKPROTECTOR_REGULAR is not set
  # CONFIG_CC_STACKPROTECTOR_STRONG is not set
  CONFIG_CC_STACKPROTECTOR_AUTO=y

and when you ran this through "make oldconfig" with the Kbuild changes,
it would ask you about the regular CONFIG_CC_STACKPROTECTOR (that had
been renamed from CONFIG_CC_STACKPROTECTOR_REGULAR to just
CONFIG_CC_STACKPROTECTOR), but it would think that the STRONG version
used to be disabled (because it was really enabled by AUTO), and would
disable it in the new config, resulting in:

  CONFIG_HAVE_CC_STACKPROTECTOR=y
  CONFIG_CC_HAS_STACKPROTECTOR_NONE=y
  CONFIG_CC_STACKPROTECTOR=y
  # CONFIG_CC_STACKPROTECTOR_STRONG is not set
  CONFIG_CC_HAS_SANE_STACKPROTECTOR=y

That's dangerously subtle - people could suddenly find themselves with
the weaker stack protector setup without even realizing.

The solution here is to just rename not just the old RECULAR stack
protector option, but also the strong one.  This does that by just
removing the CC_ prefix entirely for the user choices, because it really
is not about the compiler support (the compiler support now instead
automatially impacts _visibility_ of the options to users).

This results in "make oldconfig" actually asking the user for their
choice, so that we don't have any silent subtle security model changes.
The end result would generally look like this:

  CONFIG_HAVE_CC_STACKPROTECTOR=y
  CONFIG_CC_HAS_STACKPROTECTOR_NONE=y
  CONFIG_STACKPROTECTOR=y
  CONFIG_STACKPROTECTOR_STRONG=y
  CONFIG_CC_HAS_SANE_STACKPROTECTOR=y

where the "CC_" versions really are about internal compiler
infrastructure, not the user selections.

Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-06-14 12:21:18 +09:00
Max Filippov
40d1a07b33 xtensa: enable stack protector
The implementation is adopted from the ARM arch. GCC 7.3, 8 or newer is
required for building the xtensa kernel with SSP.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2017-12-16 22:37:07 -08:00
Max Filippov
0013aceb30 xtensa: clean up fixups in assembly code
Remove duplicate definitions of EX() and similar TRY/CATCH and SRC/DST
macros from assembly sources and put single definition into asm/asmmacro.h

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2017-12-10 14:48:53 -08:00
Max Filippov
2da03d4114 xtensa: use call instead of callx in assembly code
Now that xtensa assembly sources are compiled with -mlongcalls let the
assembler and linker relax call instructions into l32r + callx where
needed. This change makes the code cleaner and potentially a bit faster.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2017-12-10 14:48:52 -08:00
Max Filippov
f8f02ca73c xtensa: build kernel with text-section-literals
vmlinux.lds.S doesn't do anything special with literals, so instead of
keeping them separate put them into the corresponding text sections.
Drop explicit .literal sections from the vmlinux.lds.S, use standard
section macros. Mark literal pool locations in the assembly sources.
Unfortunately assembler doesn't put literals into .init sections and
external libgcc may still have .literal sections, so sed transformation
to the linker script is still needed.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2017-12-10 14:48:51 -08:00
Max Filippov
f984409a00 xtensa: use generic tracehooks
Use tracehook_report_syscall_{entry,exit} instead of a local copy of it
in do_syscall_trace. Allow tracehook to cancel syscall by returning
invalid syscall number to the system_call function.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2017-05-01 07:00:10 -07:00
Linus Torvalds
4cdf8dbe2d Merge branch 'work.uaccess2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull uaccess.h prepwork from Al Viro:
 "Preparations to tree-wide switch to use of linux/uaccess.h (which,
  obviously, will allow to start unifying stuff for real). The last step
  there, ie

    PATT='^[[:blank:]]*#[[:blank:]]*include[[:blank:]]*<asm/uaccess.h>'
    sed -i -e "s!$PATT!#include <linux/uaccess.h>!" \
            `git grep -l "$PATT"|grep -v ^include/linux/uaccess.h`

  is not taken here - I would prefer to do it once just before or just
  after -rc1.  However, everything should be ready for it"

* 'work.uaccess2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  remove a stray reference to asm/uaccess.h in docs
  sparc64: separate extable_64.h, switch elf_64.h to it
  score: separate extable.h, switch module.h to it
  mips: separate extable.h, switch module.h to it
  x86: separate extable.h, switch sections.h to it
  remove stray include of asm/uaccess.h from cacheflush.h
  mn10300: remove a bogus processor.h->uaccess.h include
  xtensa: split uaccess.h into C and asm sides
  bonding: quit messing with IOCTL
  kill __kernel_ds_p off
  mn10300: finish verify_area() off
  frv: move HAVE_ARCH_UNMAPPED_AREA to pgtable.h
  exceptions: detritus removal
2016-10-11 23:38:39 -07:00
Al Viro
76580237d1 xtensa: split uaccess.h into C and asm sides
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-09-27 21:15:21 -04:00
Max Filippov
a9f2fc628e xtensa: cleanup MMU setup and kernel layout macros
Make kernel load address explicit, independent of the selected MMU
configuration and configurable from Kconfig. Do not restrict it to the
first 512MB of the physical address space.

Cleanup kernel memory layout macros:

- rename VECBASE_RESET_VADDR to VECBASE_VADDR, XC_VADDR to VECTOR_VADDR;
- drop VIRTUAL_MEMORY_ADDRESS and LOAD_MEMORY_ADDRESS;
- introduce PHYS_OFFSET and use it in __va and __pa definitions;
- synchronize MMU/noMMU vectors, drop unused NMI vector;
- replace hardcoded vectors offset of 0x3000 with Kconfig symbol.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2016-07-24 06:33:58 +03:00
Max Filippov
c91e02bd97 xtensa: support hardware breakpoints/watchpoints
Use perf framework to manage hardware instruction and data breakpoints.
Add two new ptrace calls: PTRACE_GETHBPREGS and PTRACE_SETHBPREGS to
query and set instruction and data breakpoints.
Address bit 0 choose instruction (0) or data (1) break register, bits
31..1 are the register number.
Both calls transfer two 32-bit words: address (0) and control (1).
Instruction breakpoint contorl word is 0 to clear breakpoint, 1 to set.
Data breakpoint control word bit 31 is 'trigger on store', bit 30 is
'trigger on load, bits 29..0 are length. Length 0 is used to clear a
breakpoint. To set a breakpoint length must be a power of 2 in the range
1..64 and the address must be length-aligned.

Introduce new thread_info flag: TIF_DB_DISABLED. Set it if debug
exception is raised by the kernel code accessing watched userspace
address and disable corresponding data breakpoint. On exit to userspace
check that flag and, if set, restore all data breakpoints.

Handle debug exceptions raised with PS.EXCM set. This may happen when
window overflow/underflow handler or fast exception handler hits data
breakpoint, in which case save and disable all data breakpoints,
single-step faulting instruction and restore data breakpoints.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2016-03-11 08:53:32 +00:00
Max Filippov
6ec7026ac0 xtensa: use context structure for debug exceptions
With implementation of data breakpoints debug exceptions raised when
PS.EXCM is set need to be handled, e.g. window overflow code can write
to watched userspace address. Currently debug exception handler uses
EXCSAVE and DEPC SRs to save temporary registers, but DEPC may not be
available when PS.EXCM is set and more space will be needed to save
additional state.
Reorganize debug context: create per-CPU structure debug_table instance
and store its address in the EXCSAVE<debug level> instead of
debug_exception function address. Expand this structure when more save
space is needed.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2016-03-11 08:53:32 +00:00
Max Filippov
5029615e25 xtensa: fixes for configs without loop option
Build-time fixes:
- make lbeg/lend/lcount save/restore conditional on kernel entry;
- don't clear lcount in platform_restart functions unconditionally.

Run-time fixes:
- use correct end of range register in __endla paired with __loopt, not
  the unused temporary register. This fixes .bss zero-initialization.
  Update comments in asmmacro.h;
- don't clobber a10 in the usercopy that leads to access to unmapped
  memory.

Cc: <stable@vger.kernel.org>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-11-02 18:02:47 +03:00
Max Filippov
38fef73c21 xtensa: implement fake NMI
In case perf IRQ is the highest of the medium-level IRQs, and is alone
on its level, it may be treated as NMI:
- LOCKLEVEL is defined to be one level less than EXCM level,
- IRQ masking never lowers current IRQ level,
- new fake exception cause code, EXCCAUSE_MAPPED_NMI is assigned to that
  IRQ; new second level exception handler, do_nmi, assigned to it
  handles it as NMI,
- atomic operations in configurations without s32c1i still need to mask
  all interrupts.

Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-08-17 07:33:39 +03:00
Max Filippov
98e298329e xtensa: don't touch EXC_TABLE_FIXUP in _switch_to
There's no way _switch_to can produce double exceptions now, don't
enter/leave EXC_TABLE_FIXUP critical section.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-08-17 07:33:37 +03:00
Max Filippov
77d6273e79 xtensa: fix kernel register spilling
call12 can't be safely used as the first call in the inline function,
because the compiler does not extend the stack frame of the bounding
function accordingly, which may result in corruption of local variables.

If a call needs to be done, do call8 first followed by call12.

For pure assembly code in _switch_to increase stack frame size of the
bounding function.

Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-08-17 07:33:35 +03:00
Max Filippov
7d5f6a9a24 xtensa: reorganize irq flags tracing
entry.s only disables IRQs on hardware IRQ, move trace_hardirqs_off call
into do_interrupt. Check actual intlevel that will be restored on return
from exception handler to decide if trace_hardirqs_on should be called.
Annotate IRQ on/off points in the TIF_* handling loop on return from
exception handler.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-08-17 07:33:34 +03:00
Max Filippov
b6569439f1 xtensa: keep exception/interrupt stack continuous
Restore original a0 in the kernel exception stack frame. This way it
looks like the frame that got interrupt/exception did alloca (copy a0 and
a1 spilled under old stack to the new location as well) to save registers
and then did a call to handler.
The point where interrupt/exception was taken is not in the stack chain,
only in pt_regs (call4 from that address can be simulated to keep it in
the stack trace).

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-08-17 07:32:48 +03:00
Max Filippov
4229fb12a0 xtensa: fix threadptr reload on return to userspace
Userspace return code may skip restoring THREADPTR register if there are
no registers that need to be zeroed. This leads to spurious failures in
libc NPTL tests.

Always restore THREADPTR on return to userspace.

Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2015-07-15 01:48:43 +03:00
Max Filippov
9184289c97 xtensa: deprecate fast_xtensa and fast_spill_registers syscalls
These syscalls are not used by userspace tools for some time now, and
they have issues when called with invalid arguments. It's not worth
changing signal delivery mechanism as we don't expect any new users for
these syscalls. Let's keep them for backwards compatibility under #ifdef,
disabled by default.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2014-08-14 11:59:33 +04:00
Max Filippov
d1b6ba82a5 xtensa: fix a6 and a7 handling in fast_syscall_xtensa
Remove restoring a6 on some return paths and instead modify and restore
it in a single place, using symbolic name.
Correctly restore a7 from PT_AREG7 in case of illegal a6 value.

Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2014-08-14 11:59:31 +04:00
Max Filippov
7128039fe2 xtensa: fix TLBTEMP_BASE_2 region handling in fast_second_level_miss
Current definition of TLBTEMP_BASE_2 is always 32K above the
TLBTEMP_BASE_1, whereas fast_second_level_miss handler for the TLBTEMP
region analyzes virtual address bit (PAGE_SHIFT + DCACHE_ALIAS_ORDER)
to determine TLBTEMP region where the fault happened. The size of the
TLBTEMP region is also checked incorrectly: not 64K, but twice data
cache way size (whicht may as well be less than the instruction cache
way size).

Fix TLBTEMP_BASE_2 to be TLBTEMP_BASE_1 + data cache way size.
Provide TLBTEMP_SIZE that is a greater of doubled data cache way size or
the instruction cache way size, and use it to determine if the second
level TLB miss occured in the TLBTEMP region.

Practical occurence of page faults in the TLBTEMP area is extremely
rare, this code can be tested by deletion of all w[di]tlb instructions
in the tlbtemp_mapping region.

Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2014-08-14 11:59:19 +04:00
Max Filippov
5224712374 xtensa: fix access to THREAD_RA/THREAD_SP/THREAD_DS
With SMP and a lot of debug options enabled task_struct::thread gets out
of reach of s32i/l32i instructions with base pointing at task_struct,
breaking build with the following messages:

  arch/xtensa/kernel/entry.S: Assembler messages:
  arch/xtensa/kernel/entry.S:1002: Error: operand 3 of 'l32i.n' has invalid value '1048'
  arch/xtensa/kernel/entry.S:1831: Error: operand 3 of 's32i.n' has invalid value '1040'
  arch/xtensa/kernel/entry.S:1832: Error: operand 3 of 's32i.n' has invalid value '1044'

Change base to point to task_struct::thread in such cases.
Don't use a10 in _switch_to to save/restore prev pointer as a2 is not
clobbered.

Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2014-08-14 11:59:16 +04:00
Chris Zankel
6b5a1f74e5 xtensa: fix fast_syscall_spill_registers
The original implementation could clobber registers under certain conditions.

The Xtensa processor architecture uses windowed registers and the original
implementation was using a4 as a temporary register, which under certain
conditions could be register a0 of the oldest window frame, and didn't always
restore the content correctly.

By moving the _spill_registers routine inside the fast system call, it frees
up one more register (the return address is not required anymore) for the
spill routine.

Signed-off-by: Chris Zankel <chris@zankel.net>
2014-01-28 22:09:51 -08:00
Max Filippov
3251f1e27a xtensa: save current register frame in fast_syscall_spill_registers_fixup
We need it saved because it contains a3 where we track which register
windows we still need to spill, and fixup handler may call C exception
handlers. Also fix comments.

Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2014-01-25 23:24:38 +04:00
Max Filippov
e2fd1374c7 xtensa: introduce spill_registers_kernel macro
Most in-kernel users want registers spilled on the kernel stack and
don't require PS.EXCM to be set. That means that they don't need fixup
routine and could reuse regular window overflow mechanism for that,
which makes spill routine very simple.

Cc: stable@vger.kernel.org
Suggested-by: Chris Zankel <chris@zankel.net>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
2014-01-25 23:20:09 +04:00
Max Filippov
244066f4be xtensa: fix fast_syscall_spill_registers_fixup
fast_syscall_spill_registers_fixup was not correctly updated by the
'keep a3 and excsave1 on entry to exception handlers' patch: it doesn't
preserve a3 that it gets on entry, breaking _spill_registers in case of
page fault on stack during register spilling, leading to unhandled
exception in kernel mode.

Preserve a3 by saving it in the original _spill_registers stack frame's
a3 during exception handling and restoring it afterwards.

Also fix comments and function bounds annotations.

Reported-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Tested-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-10-15 13:39:06 -07:00
Max Filippov
fff96d69f2 xtensa: new fast_alloca handler
Instead of emulating movsp instruction in the kernel use window
underflow handler to load missing register window and retry failed
movsp.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-09-06 09:48:12 -07:00
Max Filippov
99d5040ebc xtensa: keep a3 and excsave1 on entry to exception handlers
Based on the SMP patch by Joe Taylor and subsequent fixes.
Preserve exception table pointer (normally stored in excsave1 SR) as it
cannot be easily restored in SMP environment.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-09-06 09:47:41 -07:00
Max Filippov
16c5becf39 xtensa: enable kernel preemption
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-09-06 09:46:49 -07:00
Max Filippov
aea8e7c80a xtensa: check thread flags atomically on return from user exception
Check pending signals and rescheduling thread flags with interrupts
disabled, and don't enable them if no flags are set. Call
trace_hardirqs_on after thread flags handling, so that rescheduling is
done and hardirqs tracking flag is updated in the correct task context.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-09-06 09:46:35 -07:00
Max Filippov
a99e07ee5e xtensa: check TLB sanity on return to userspace
- check that user TLB mappings correspond to the current page table;
- check that TLB mapping VPN is in the kernel/user address range
  in accordance with its ASID.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-07-08 01:18:56 -07:00
Max Filippov
51fc41a906 xtensa: fix fast_store_prohibited _PAGE_WRITABLE_BIT test
Before _PAGE_WRITABLE_BIT test fast_store_prohibited must make sure that
PTE is present. Otherwise 'writable' bit is undefined and may be reused
in the 'file offset' or 'swap type' PTE fields.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-05-20 11:48:23 -07:00
Max Filippov
895666a992 xtensa: disable IRQs while IRQ handler is running
IRQ handlers are expected to run with IRQs disabled.
See e.g. http://lwn.net/Articles/380931/ for a longer story.

This was overlooked in the commit
  2d1c645 xtensa: dispatch medium-priority interrupts
Revert to old behavior and simplify interrupt entry and exit code.
Interrupt handler still honours IRQ priority.

do_notify_resume/schedule must be called with interrupts enabled, enable
interrupts if we return from user exception.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-05-09 01:07:11 -07:00
Max Filippov
c92931b200 xtensa: add irq flags trace support
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-05-09 01:07:11 -07:00
Chris Zankel
c50842df47 xtensa: add support for TLS
The Xtensa architecture provides a global register called THREADPTR
for the purpose of Thread Local Storage (TLS) support. This allows us
to use a fairly simple implementation, keeping the thread pointer in
the regset and simply saving and restoring it upon entering/exiting
the from user space.

Signed-off-by: Chris Zankel <chris@zankel.net>
2013-02-23 19:35:57 -08:00
dann
e6ffe17ec4 xtensa: add support for oprofile
Support call graph profiling.
Keep upper two bits of PC unchanged through backtrace rather than take
them from sp (a1). The stack pointer is usually in the same GB (same
upper 2 bits) as PC, but technically doesn't always have to be (and
might not in the future, when taking full advantage of MMU v3).

Signed-off-by: Dan Nicolaescu <dann@xtensa-linux.org>
Signed-off-by: Pete Delaney <piet@tensilica.com>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-02-23 19:22:54 -08:00
Marc Gauthier
2d1c645cc5 xtensa: dispatch medium-priority interrupts
Add support for dispatching medium-priority interrupts, that is,
interrupts of priority levels 2 to EXCM_LEVEL. IRQ handling may be
preempted by higher priority IRQ.

Signed-off-by: Marc Gauthier <marc@tensilica.com>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
2013-02-23 19:12:52 -08:00
Chris Zankel
c4c4594b00 xtensa: clean up files to make them code-style compliant
Remove heading and trailing spaces, trim trailing lines, and wrap lines
that are longer than 80 characters.

Signed-off-by: Chris Zankel <chris@zankel.net>
2012-12-18 21:10:25 -08:00