This adds emulation support for the following integer instructions:
* Modulo Signed Word (modsw)
* Modulo Unsigned Word (moduw)
Signed-off-by: PrasannaKumar Muralidharan <prasannatsmkumar@gmail.com>
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds emulation support for the following integer instructions:
* Extend-Sign Word and Shift Left Immediate (extswsli[.])
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds emulation support for the following integer instructions:
* Count Trailing Zeros Word (cnttzw[.])
* Count Trailing Zeros Doubleword (cnttzd[.])
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds emulation support for the following integer instructions:
* Deliver A Random Number (darn)
As suggested by Michael, this uses a raw .long for specifying the
instruction word when using inline assembly to retain compatibility
with older binutils.
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds emulation support for the following integer instructions:
* Multiply-Add High Doubleword (maddhd)
* Multiply-Add High Doubleword Unsigned (maddhdu)
* Multiply-Add Low Doubleword (maddld)
As suggested by Michael, this uses a raw .long for specifying the
instruction word when using inline assembly to retain compatibility
with older binutils.
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Three conflicts, one of which, for marvell10g.c is non-trivial and
requires some follow-up from Heiner or someone else.
The issue is that Heiner converted the marvell10g driver over to
use the generic c45 code as much as possible.
However, in 'net' a bug fix appeared which makes sure that a new
local mask (MDIO_AN_10GBT_CTRL_ADV_NBT_MASK) with value 0x01e0
is cleared.
Signed-off-by: David S. Miller <davem@davemloft.net>
One fix for an oops when using SRIOV, introduced by the recent changes to
support compound IOMMU groups.
Thanks to:
Alexey Kardashevskiy.
-----BEGIN PGP SIGNATURE-----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=NBtz
-----END PGP SIGNATURE-----
Merge tag 'powerpc-5.0-6' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fix from Michael Ellerman:
"One fix for an oops when using SRIOV, introduced by the recent changes
to support compound IOMMU groups.
Thanks to Alexey Kardashevskiy"
* tag 'powerpc-5.0-6' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/powernv/sriov: Register IOMMU groups for VFs
Some stack pointers used to also be thread_info pointers
and were called tp. Now that they are only stack pointers,
rename them sp.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Now that current_thread_info is located at the beginning of 'current'
task struct, CURRENT_THREAD_INFO macro is not really needed any more.
This patch replaces it by loads of the value at PACA_THREAD_INFO(r13).
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Add PACA_THREAD_INFO rather than using PACACURRENT]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Now that thread_info is similar to task_struct, its address is in r2
so CURRENT_THREAD_INFO() macro is useless. This patch removes it.
This patch also moves the 'tovirt(r2, r2)' down just before the
reactivation of MMU translation, so that we keep the physical address
of 'current' in r2 until then. It avoids a few calls to tophys().
At the same time, as the 'cpu' field is not anymore in thread_info,
TI_CPU is renamed TASK_CPU by this patch.
It also allows to get rid of a couple of
'#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE' as ACCOUNT_CPU_USER_ENTRY()
and ACCOUNT_CPU_USER_EXIT() are empty when
CONFIG_VIRT_CPU_ACCOUNTING_NATIVE is not defined.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Fix a missed conversion of TI_CPU idle_6xx.S]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The table of pointers 'current_set' has been used for retrieving
the stack and current. They used to be thread_info pointers as
they were pointing to the stack and current was taken from the
'task' field of the thread_info.
Now, the pointers of 'current_set' table are now both pointers
to task_struct and pointers to thread_info.
As they are used to get current, and the stack pointer is
retrieved from current's stack field, this patch changes
their type to task_struct, and renames secondary_ti to
secondary_current.
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
thread_info is not anymore in the stack, so the entire stack
can now be used.
There is also no risk anymore of corrupting task_cpu(p) with a
stack overflow so the patch removes the test.
When doing this, an explicit test for NULL stack pointer is
needed in validate_sp() as it is not anymore implicitely covered
by the sizeof(thread_info) gap.
In the meantime, with the previous patch all pointers to the stacks
are not anymore pointers to thread_info so this patch changes them
to void*
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch activates CONFIG_THREAD_INFO_IN_TASK which
moves the thread_info into task_struct.
Moving thread_info into task_struct has the following advantages:
- It protects thread_info from corruption in the case of stack
overflows.
- Its address is harder to determine if stack addresses are leaked,
making a number of attacks more difficult.
This has the following consequences:
- thread_info is now located at the beginning of task_struct.
- The 'cpu' field is now in task_struct, and only exists when
CONFIG_SMP is active.
- thread_info doesn't have anymore the 'task' field.
This patch:
- Removes all recopy of thread_info struct when the stack changes.
- Changes the CURRENT_THREAD_INFO() macro to point to current.
- Selects CONFIG_THREAD_INFO_IN_TASK.
- Modifies raw_smp_processor_id() to get ->cpu from current without
including linux/sched.h to avoid circular inclusion and without
including asm/asm-offsets.h to avoid symbol names duplication
between ASM constants and C constants.
- Modifies klp_init_thread_info() to take a task_struct pointer
argument.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Add task_stack.h to livepatch.h to fix build fails]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Make sure CURRENT_THREAD_INFO() is used with r1 which is the virtual
address of the stack, in order to ease the switch to r2 when we enable
THREAD_INFO_IN_TASK, as we have no register having the phys address of
current.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Change current_pt_regs() to use task_stack_page() rather than
current_thread_info() so that it keeps working once we enable
THREAD_INFO_IN_TASK.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Split out of large patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When we enable THREAD_INFO_IN_TASK we will remove our definition of
current_thread_info(). Instead it will come from linux/thread_info.h
So switch processor.h to include the latter, so that it can continue
to find current_thread_info().
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Currently INIT_SP_LIMIT uses sizeof(init_thread_info), but that symbol
won't exist when we enable THREAD_INFO_IN_TASK. So just use the sizeof
the type which is the same value but will continue to work.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Rather than using the thread info use task_stack_page() to initialise
paca->kstack, that way it will work with THREAD_INFO_IN_TASK.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Update a few comments that talk about current_thread_info() in
preparation for THREAD_INFO_IN_TASK.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We have a few places that use current_thread_info()->task to access
current. This won't work with THREAD_INFO_IN_TASK so fix them now.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
A few places use CURRENT_THREAD_INFO, or the C version, to find the
stack. This will no longer work with THREAD_INFO_IN_TASK so change
them to find the stack in other ways.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The purpose of the pointer given to call_do_softirq() and
call_do_irq() is to point the new stack. Currently that's the same
thing as the thread_info, but won't be with THREAD_INFO_IN_TASK.
So change the parameter to void* and rename it 'sp'.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out of larger patch]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch renames THREAD_INFO to TASK_STACK, because it is in fact
the offset of the pointer to the stack in task_struct so this pointer
will not be impacted by the move of THREAD_INFO.
Also make it available on 64-bit, as we'll need it there when we
activate THREAD_INFO_IN_TASK.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Make available on 64-bit]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
[text copied from commit 9bbd4c56b0
("arm64: prep stack walkers for THREAD_INFO_IN_TASK")]
When CONFIG_THREAD_INFO_IN_TASK is selected, task stacks may be freed
before a task is destroyed. To account for this, the stacks are
refcounted, and when manipulating the stack of another task, it is
necessary to get/put the stack to ensure it isn't freed and/or re-used
while we do so.
This patch reworks the powerpc stack walking code to account for this.
When CONFIG_THREAD_INFO_IN_TASK is not selected these perform no
refcounting, and this should only be a structural change that does not
affect behaviour.
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Move try_get_task_stack() below tsk == NULL check in show_stack()]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When moving to CONFIG_THREAD_INFO_IN_TASK, the thread_info 'cpu' field
gets moved into task_struct and only defined when CONFIG_SMP is set.
This patch ensures that TI_CPU is only used when CONFIG_SMP is set and
that task_struct 'cpu' field is not used directly out of SMP code.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When activating CONFIG_THREAD_INFO_IN_TASK, linux/sched.h includes
asm/current.h. This generates a circular dependency. To avoid that,
asm/processor.h shall not be included in mmu-hash.h.
In order to do that, this patch moves into a new header called
asm/task_size_64/32.h all the TASK_SIZE related constants, which can
then be included in mmu-hash.h directly.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Split out all the TASK_SIZE constants not just 64-bit ones]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Since only the virtual address of allocated blocks is used,
lets use functions returning directly virtual address.
Those functions have the advantage of also zeroing the block.
Suggested-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
In __secondary_start() we load the thread_info of the idle task of the
secondary CPU from current_set[cpu], and then convert it into a stack
pointer before storing that back to paca->kstack.
As pointed out in commit f761622e59 ("powerpc: Initialise
paca->kstack before early_setup_secondary") it's important that we
initialise paca->kstack before calling the MMU setup code, in
particular slb_initialize(), because it will bolt the SLB entry for
the kstack into the SLB.
However we have already setup paca->kstack in cpu_idle_thread_init(),
since commit 3b5750644b ("[POWERPC] Bolt in SLB entry for kernel
stack on secondary cpus") (May 2008).
It's also in cpu_idle_thread_init() that we initialise current_set[cpu]
with the thread_info pointer, so there is no issue of the timing being
different between the two.
Therefore the initialisation of paca->kstack in __setup_secondary() is
completely redundant, so remove it.
This has the added benefit of removing code that runs in real mode,
and is therefore restricted by the RMO, and so opens the way for us to
enable THREAD_INFO_IN_TASK.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Currently in system_call_exit() we have an optimisation where we
disable MSR_RI (recoverable interrupt) and MSR_EE (external interrupt
enable) in a single mtmsrd instruction.
Unfortunately this will no longer work with THREAD_INFO_IN_TASK,
because then the load of TI_FLAGS might fault and faulting with MSR_RI
clear is treated as an unrecoverable exception which leads to a
panic().
So change the code to only clear MSR_EE prior to loading TI_FLAGS,
leaving the clear of MSR_RI until later. We have some latitude in
where do the clear of MSR_RI. A bit of experimentation has shown that
this location gives the least slow down.
This still causes a noticeable slow down in our null_syscall
performance. On a Power9 DD2.2:
Before After Delta Delta %
955 cycles 999 cycles -44 -4.6%
On the plus side this does simplify the code somewhat, because we
don't have to reenable MSR_RI on the restore_math() or
syscall_exit_work() paths which was necessitated previously by the
optimisation.
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
kcov provides kernel coverage data that's useful for fuzzing tools like
syzkaller.
Wire up kcov support on powerpc. Disable kcov instrumentation on the same
files where we currently disable gcov and UBSan instrumentation, plus some
additional exclusions which appear necessary to boot on book3e machines.
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Acked-by: Dmitry Vyukov <dvyukov@google.com>
Tested-by: Daniel Axtens <dja@axtens.net> # e6500
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
On 8xx, large pages (512kb or 8M) are used to map kernel linear
memory. Aligning to 8M reduces TLB misses as only 8M pages are used
in that case. We make 8M the default for data.
This patchs allows the user to do it via Kconfig.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch implements handling of STRICT_KERNEL_RWX with
large TLBs directly in the TLB miss handlers.
To do so, etext and sinittext are aligned on 512kB boundaries
and the miss handlers use 512kB pages instead of 8Mb pages for
addresses close to the boundaries.
It sets RO PP flags for addresses under sinittext.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Depending on the number of available BATs for mapping the different
kernel areas, it might be needed to increase the alignment of _etext
and/or of data areas.
This patchs allows the user to do it via Kconfig.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Today, STRICT_KERNEL_RWX is based on the use of regular pages
to map kernel pages.
On Book3s 32, it has three consequences:
- Using pages instead of BAT for mapping kernel linear memory severely
impacts performance.
- Exec protection is not effective because no-execute cannot be set at
page level (except on 603 which doesn't have hash tables)
- Write protection is not effective because PP bits do not provide RO
mode for kernel-only pages (except on 603 which handles it in software
via PAGE_DIRTY)
On the 603+, we have:
- Independent IBAT and DBAT allowing limitation of exec parts.
- NX bit can be set in segment registers to forbit execution on memory
mapped by pages.
- RO mode on DBATs even for kernel-only blocks.
On the 601, there is nothing much we can do other than warn the user
about it, because:
- BATs are common to instructions and data.
- BAT do not provide RO mode for kernel-only blocks.
- segment registers don't have the NX bit.
In order to use IBAT for exec protection, this patch:
- Aligns _etext to BAT block sizes (128kb)
- Set NX bit in kernel segment register (Except on vmalloc area when
CONFIG_MODULES is selected)
- Maps kernel text with IBATs.
In order to use DBAT for exec protection, this patch:
- Aligns RW DATA to BAT block sizes (4M)
- Maps kernel RO area with write prohibited DBATs
- Maps remaining memory with remaining DBATs
Here is what we get with this patch on a 832x when activating
STRICT_KERNEL_RWX:
Symbols:
c0000000 T _stext
c0680000 R __start_rodata
c0680000 R _etext
c0800000 T __init_begin
c0800000 T _sinittext
~# cat /sys/kernel/debug/block_address_translation
---[ Instruction Block Address Translation ]---
0: 0xc0000000-0xc03fffff 0x00000000 Kernel EXEC coherent
1: 0xc0400000-0xc05fffff 0x00400000 Kernel EXEC coherent
2: 0xc0600000-0xc067ffff 0x00600000 Kernel EXEC coherent
3: -
4: -
5: -
6: -
7: -
---[ Data Block Address Translation ]---
0: 0xc0000000-0xc07fffff 0x00000000 Kernel RO coherent
1: 0xc0800000-0xc0ffffff 0x00800000 Kernel RW coherent
2: 0xc1000000-0xc1ffffff 0x01000000 Kernel RW coherent
3: 0xc2000000-0xc3ffffff 0x02000000 Kernel RW coherent
4: 0xc4000000-0xc7ffffff 0x04000000 Kernel RW coherent
5: 0xc8000000-0xcfffffff 0x08000000 Kernel RW coherent
6: 0xd0000000-0xdfffffff 0x10000000 Kernel RW coherent
7: -
~# cat /sys/kernel/debug/segment_registers
---[ User Segments ]---
0x00000000-0x0fffffff Kern key 1 User key 1 VSID 0xa085d0
0x10000000-0x1fffffff Kern key 1 User key 1 VSID 0xa086e1
0x20000000-0x2fffffff Kern key 1 User key 1 VSID 0xa087f2
0x30000000-0x3fffffff Kern key 1 User key 1 VSID 0xa08903
0x40000000-0x4fffffff Kern key 1 User key 1 VSID 0xa08a14
0x50000000-0x5fffffff Kern key 1 User key 1 VSID 0xa08b25
0x60000000-0x6fffffff Kern key 1 User key 1 VSID 0xa08c36
0x70000000-0x7fffffff Kern key 1 User key 1 VSID 0xa08d47
0x80000000-0x8fffffff Kern key 1 User key 1 VSID 0xa08e58
0x90000000-0x9fffffff Kern key 1 User key 1 VSID 0xa08f69
0xa0000000-0xafffffff Kern key 1 User key 1 VSID 0xa0907a
0xb0000000-0xbfffffff Kern key 1 User key 1 VSID 0xa0918b
---[ Kernel Segments ]---
0xc0000000-0xcfffffff Kern key 0 User key 1 No Exec VSID 0x000ccc
0xd0000000-0xdfffffff Kern key 0 User key 1 No Exec VSID 0x000ddd
0xe0000000-0xefffffff Kern key 0 User key 1 No Exec VSID 0x000eee
0xf0000000-0xffffffff Kern key 0 User key 1 No Exec VSID 0x000fff
Aligning _etext to 128kb allows to map up to 32Mb text with 8 IBATs:
16Mb + 8Mb + 4Mb + 2Mb + 1Mb + 512kb + 256kb + 128kb (+ 128kb) = 32Mb
(A 9th IBAT is unneeded as 32Mb would need only a single 32Mb block)
Aligning data to 4M allows to map up to 512Mb data with 8 DBATs:
16Mb + 8Mb + 4Mb + 4Mb + 32Mb + 64Mb + 128Mb + 256Mb = 512Mb
Because some processors only have 4 BATs and because some targets need
DBATs for mapping other areas, the following patch will allow to
modify _etext and data alignment.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
setibat() and clearibat() allows to manipulate IBATs independently
of DBATs.
update_bats() allows to update bats after init. This is done
with MMU off.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
CONFIG_STRICT_KERNEL_RWX requires a special alignment
for DATA for some subarches. Today it is just defined
as an #ifdef in vmlinux.lds.S
In order to get more flexibility, this patch moves the
definition of this alignment in Kconfig
On some subarches, CONFIG_STRICT_KERNEL_RWX will
require a special alignment of _etext.
This patch also adds a configuration item for it in Kconfig
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch defined CONFIG_PPC_PAGE_SHIFT in order
to be able to use PAGE_SHIFT value inside Kconfig.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Add a helper to know whether STRICT_KERNEL_RWX is enabled.
This is based on rodata_enabled flag which is defined only
when CONFIG_STRICT_KERNEL_RWX is selected.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch add an helper which wraps 'mtsrin' instruction
to write into segment registers.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Do not set IBAT when setbat() is called without _PAGE_EXEC
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
wii_mmu_mapin_mem2() is not used anymore, remove it.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When CONFIG_BDI_SWITCH is set, the page tables have to be populated
allthough large TLBs are used, because the BDI switch knows nothing
about those large TLBs which are handled directly in TLB miss logic.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Now that mmu_mapin_ram() is able to handle other blocks
than the one starting at 0, the WII can use it for all
its blocks.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch reworks mmu_mapin_ram() to be more generic and map as much
blocks as possible. It now supports blocks not starting at address 0.
It scans DBATs array to find free ones instead of forcing the use of
BAT2 and BAT3.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
At the time being, mmu_mapin_ram() always maps RAM from the beginning.
But some platforms like the WII have to map a second block of RAM.
This patch adds to mmu_mapin_ram() the base address of the block.
At the moment, only base address 0 is supported.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
'nobats' kernel parameter or some options like CONFIG_DEBUG_PAGEALLOC
deny the use of BATS for mapping memory.
This patch makes sure that the specific wii RAM mapping function
takes it into account as well.
Fixes: de32400dd2 ("wii: use both mem1 and mem2 as ram")
Cc: stable@vger.kernel.org
Reviewed-by: Jonathan Neuschafer <j.neuschaefer@gmx.net>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
At the time being, initial MMU setup allows 24 Mbytes
of DATA and 8 Mbytes of code.
Some debug setup like CONFIG_KASAN generate huge
kernels with text size over the 8M limit and data over the
24 Mbytes limit.
Here is an 8xx kernel compiled with CONFIG_KASAN_INLINE for
one of my boards:
[root@po16846vm linux-powerpc]# size -x vmlinux
text data bss dec hex filename
0x111019c 0x41b0d4 0x490de0 26984528 19bc050 vmlinux
This patch maps up to 32 Mbytes code based on _einittext symbol
and allows 32 Mbytes of memory instead of 24.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch replaces most #ifdef mess by IS_ENABLED() in 8xx_mmu.c
This has the advantage of allowing syntax verification at compile
time regardless of selected options.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds test cases for the addc[.] instruction.
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds test cases for the add[.] instruction.
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This enhances the current selftest framework for validating
the in-kernel instruction emulation infrastructure by adding
support for compute type instructions i.e. integer ALU-based
instructions. Originally, this framework was limited to only
testing load and store instructions.
While most of the GPRs can be validated, support for SPRs is
limited to LR, CR and XER for now.
When writing the test cases, one must ensure that the Stack
Pointer (GPR1) or the Thread Pointer (GPR13) are not touched
by any means as these are vital non-volatile registers.
Signed-off-by: Sandipan Das <sandipan@linux.ibm.com>
[mpe: Use patch_site for the code patching]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
There are no major new features this time, just a collection of bug
fixes and improvements in various areas, including machine check
handling and context switching of protection-key-related registers.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJcb3lEAAoJEJ2a6ncsY3GflNwH/2ezxhHv7CRy18d2D3F+Kna+
YQs3V/pJfBRvVdV7ZLxnR03H/NmzAK3UOzRfqGodYUtbF+gUDqSuM27lAxMKrjBv
S87X5g/1ZdiQNnqYK7PIBn75Tx27vnw2kJAif8rXTfqbj8qLUsXcNhsziA16sJOA
azbD5PBp9mOVzTojawyriJ3H8LYqw+vinad0idvFrApFCuNmMxv56FR6H+IBadt7
1UJyx6AegQACdhxvy0CzmZjzzXw02z9zeFUa4lakm2sORc4fbbyyZ68CtkGURg7A
8rt2j9SGt649ExpjfG2Cz/UihMGIMXSAOrpqTZMfyd9UPzPgHeKx2FidnxASUBc=
=PIT8
-----END PGP SIGNATURE-----
Merge tag 'kvm-ppc-next-5.1-1' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into kvm-next
PPC KVM update for 5.1
There are no major new features this time, just a collection of bug
fixes and improvements in various areas, including machine check
handling and context switching of protection-key-related registers.
This patch moves the files related to page table dump in a
dedicated subdirectory.
The purpose is to clean a bit arch/powerpc/mm by regrouping
multiple files handling a dedicated function.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Shorten the file names while we're at it]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This merges in the "ppc-kvm" topic branch of the powerpc tree to get a
series of commits that touch both general arch/powerpc code and KVM
code. These commits will be merged both via the KVM tree and the
powerpc tree.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
When the hash MMU is active the AMR, IAMR and UAMOR are used for
pkeys. The AMR is directly writable by user space, and the UAMOR masks
those writes, meaning both registers are effectively user register
state. The IAMR is used to create an execute only key.
Also we must maintain the value of at least the AMR when running in
process context, so that any memory accesses done by the kernel on
behalf of the process are correctly controlled by the AMR.
Although we are correctly switching all registers when going into a
guest, on returning to the host we just write 0 into all regs, except
on Power9 where we restore the IAMR correctly.
This could be observed by a user process if it writes the AMR, then
runs a guest and we then return immediately to it without
rescheduling. Because we have written 0 to the AMR that would have the
effect of granting read/write permission to pages that the process was
trying to protect.
In addition, when using the Radix MMU, the AMR can prevent inadvertent
kernel access to userspace data, writing 0 to the AMR disables that
protection.
So save and restore AMR, IAMR and UAMOR.
Fixes: cf43d3b264 ("powerpc: Enable pkey subsystem")
Cc: stable@vger.kernel.org # v4.16+
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
The anon fd's ops releases the KVM reference in the release hook.
However we reference the KVM object after we create the fd so there is
small window when the release function can be called and
dereferenced the KVM object which potentially may free it.
It is not a problem at the moment as the file is created and KVM is
referenced under the KVM lock and the release function obtains the same
lock before dereferencing the KVM (although the lock is not held when
calling kvm_put_kvm()) but it is potentially fragile against future changes.
This references the KVM object before creating a file.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Currently trying to build without IOMMU support will fail:
(.text+0x1380): undefined reference to `kvmppc_h_get_tce'
(.text+0x1384): undefined reference to `kvmppc_rm_h_put_tce'
(.text+0x149c): undefined reference to `kvmppc_rm_h_stuff_tce'
(.text+0x14a0): undefined reference to `kvmppc_rm_h_put_tce_indirect'
This happens because turning off IOMMU support will prevent
book3s_64_vio_hv.c from being built because it is only built when
SPAPR_TCE_IOMMU is set, which depends on IOMMU support.
Fix it using ifdefs for the undefined references.
Fixes: 76d837a4c0 ("KVM: PPC: Book3S PR: Don't include SPAPR TCE code on non-pseries platforms")
Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
When using KASAN, there are parts of the shadow area where all
pages are mapped to the kasan_early_shadow_page. It is pointless
to dump one line for each of those pages (in the example below there
are 7168 entries pointing to the same physical page).
~# cat /sys/kernel/debug/kernel_page_tables
...
---[ kasan shadow mem start ]---
0xf7c00000-0xf8bfffff 0x06fac000 16M rw present dirty accessed
0xf8c00000-0xf8c03fff 0x00cd0000 16K r present dirty accessed
0xf8c04000-0xf8c07fff 0x00cd0000 16K r present dirty accessed
0xf8c08000-0xf8c0bfff 0x00cd0000 16K r present dirty accessed
0xf8c0c000-0xf8c0ffff 0x00cd0000 16K r present dirty accessed
0xf8c10000-0xf8c13fff 0x00cd0000 16K r present dirty accessed
... 7168 identical lines
0xffbfc000-0xffbfffff 0x00cd0000 16K r present dirty accessed
---[ kasan shadow mem end ]---
...
This patch modifies linux table dump to dump as a single line areas
where all addresses points to the same physical page. That physical
address is put inside [] to show that all virt pages points to the
same phys page.
~# cat /sys/kernel/debug/kernel_page_tables
...
---[ kasan shadow mem start ]---
0xf7c00000-0xf8bfffff 0x06fac000 16M rw present dirty accessed
0xf8c00000-0xffbfffff [0x00cd0000] 16K r present dirty accessed
---[ kasan shadow mem end ]---
...
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
40x/booke have another path to reach 3f from transfer_to_handler,
make sure it also calls ACCOUNT_CPU_USER_ENTRY() when
CONFIG_VIRT_CPU_ACCOUNTING_NATIVE is selected.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
For pages without _PAGE_USER, PP field is 00
For pages with _PAGE_USER, PP field is 10 for RW and 11 for RO.
This patch sets _PAGE_USER to 0x002 and _PAGE_RW to 0x001
is order to simplify TLB handling by reducing amount of shifts.
The location of _PAGE_PRESENT and _PAGE_HASHPTE doesn't matter
as they are only SW related flags.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
PAGE_ACCESSED is only needed for CONFIG_SWAP. When CONFIG_SWAP
is not set, just ignore it. If CONFIG_SWAP is set and PAGE_ACCESSED
is not, let's take a minor fault.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
PP bits take user access into account, so no need to check _PAGE_USER
here. A DSI or ISI will be generated if needed.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
PAGE_DIRTY corresponds to the C bit. If writing on
a page for which the C bit is not set, a DataStoreTLBMiss
is generated. No need to check it in DataLoadTLBMiss.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
_PAGE_RW and _PAGE_DIRTY do not matter for ITLB misses.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
ITLB miss on kernel pages only occur with CONFIG_MODULES and
CONFIG_DEBUG_PAGEALLOC.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Since commit c62ce9ef97 ("powerpc: remove remaining bits from
CONFIG_APUS"), tophys() has become a pure constant operation.
PAGE_OFFSET is known at compile time so the physical address
can be builtin directly.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Since commit c62ce9ef97 ("powerpc: remove remaining bits from
CONFIG_APUS"), tophys() has become a pure constant operation.
PAGE_OFFSET is known at compile time so the physical address
can be builtin directly.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Use SPRN_SPRG2 to store the current thread PGDIR and
avoid reading thread_struct.pgdir at every TLB miss.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When calling RTAS, the stack pointer is stored in SPRN_SPRG2
in order to be able to restore it in case of machine check in RTAS.
As machine check is not a perfomance critical path, this patch
frees SPRN_SPRG2 by using a field in thread struct instead.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
There is no reason to re-read each time the pointer at
location 0xf0 as it is fixed and known.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The 83xx has 8 SPRG registers and uses at least SPRG4
for DTLB handling LRU.
Fixes: 2319f12395 ("powerpc/mm: e300c2/c3/c4 TLB errata workaround")
Cc: stable@vger.kernel.org
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Looks like book3s/32 doesn't set RI on machine check, so
checking RI before calling die() will always be fatal
allthought this is not an issue in most cases.
Fixes: b96672dd84 ("powerpc: Machine check interrupt is a non-maskable interrupt")
Fixes: daf00ae71d ("powerpc/traps: restore recoverability of machine_check interrupts")
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Cc: stable@vger.kernel.org
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
MSR[RI] has already been cleared a few lines above.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When no machine description matches, display it clearly
before looping forever.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Commit 24be85a23d ("powerpc/powernv: Clear PECE1 in LPCR via stop-api
only on Hotplug", 2017-07-21) added two calls to opal_slw_set_reg()
inside pnv_cpu_offline(), with the aim of changing the LPCR value in
the SLW image to disable wakeups from the decrementer while a CPU is
offline. However, pnv_cpu_offline() gets called each time a secondary
CPU thread is woken up to participate in running a KVM guest, that is,
not just when a CPU is offlined.
Since opal_slw_set_reg() is a very slow operation (with observed
execution times around 20 milliseconds), this means that an offline
secondary CPU can often be busy doing the opal_slw_set_reg() call
when the primary CPU wants to grab all the secondary threads so that
it can run a KVM guest. This leads to messages like "KVM: couldn't
grab CPU n" being printed and guest execution failing.
There is no need to reprogram the SLW image on every KVM guest entry
and exit. So that we do it only when a CPU is really transitioning
between online and offline, this moves the calls to
pnv_program_cpu_hotplug_lpcr() into pnv_smp_cpu_kill_self().
Fixes: 24be85a23d ("powerpc/powernv: Clear PECE1 in LPCR via stop-api only on Hotplug")
Cc: stable@vger.kernel.org # v4.14+
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
In cpufeatures_process_feature(), if a provided CPU feature is unknown and
enable_unknown is false, we erroneously print that the feature is being
enabled and return true, even though no feature has been enabled, and
may also set feature bits based on the last entry in the match table.
Fix this so that we only set feature bits from the match table if we have
actually enabled a feature from that table, and when failing to enable an
unknown feature, always print the "not enabling" message and return false.
Coincidentally, some older gccs (<GCC 7), when invoked with
-fsanitize-coverage=trace-pc, cause a spurious uninitialised variable
warning in this function:
arch/powerpc/kernel/dt_cpu_ftrs.c: In function ‘cpufeatures_process_feature’:
arch/powerpc/kernel/dt_cpu_ftrs.c:686:7: warning: ‘m’ may be used uninitialized in this function [-Wmaybe-uninitialized]
if (m->cpu_ftr_bit_mask)
An upcoming patch will enable support for kcov, which requires this option.
This patch avoids the warning.
Fixes: 5a61ef74f2 ("powerpc/64s: Support new device tree binding for discovering CPU features")
Reported-by: Segher Boessenkool <segher@kernel.crashing.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
[ajd: add commit message]
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
The xmon debugger IPI handler waits in the callback function while
xmon is still active. This means they don't complete the IPI, and the
initiator always times out waiting for them.
Things manage to work after the timeout because there is some fallback
logic to keep NMI IPI state sane in case of the timeout, but this is a
bit ugly.
This patch changes NMI IPI back to half-asynchronous (i.e., wait for
everyone to call in, do not wait for IPI function to complete), but
the complexity is avoided by going one step further and allowing new
IPIs to be issued before the IPI functions to all complete.
If synchronization against that is required, it is left up to the
caller, but current callers don't require that. In fact with the
timeout handling, callers must be able to cope with this already.
Fixes: 5b73151fff ("powerpc: NMI IPI make NMI IPIs fully sychronous")
Cc: stable@vger.kernel.org # v4.19+
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The NMI IPI timeout logic is broken, if __smp_send_nmi_ipi() times out
on the first condition, delay_us will be zero which will send it into
the second spin loop with no timeout so it will spin forever.
Fixes: 5b73151fff ("powerpc: NMI IPI make NMI IPIs fully sychronous")
Cc: stable@vger.kernel.org # v4.19+
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
In commit 7820856a4f ("powerpc/mm/book3e/64: Remove unsupported
64Kpage size from 64bit booke") we dropped the 64K page size support
from the 64-bit nohash (Book3E) code.
But we didn't update the dependencies of the PPC_64K_PAGES option,
meaning a randconfig can still trigger this code and cause a build
breakage, eg:
arch/powerpc/include/asm/nohash/64/pgtable.h:14:2: error: #error "Page size not supported"
arch/powerpc/include/asm/nohash/mmu-book3e.h:275:2: error: #error Unsupported page size
So remove PPC_BOOK3E_64 from the dependencies. This also means we
don't need to worry about PPC_FSL_BOOK3E, because that was just trying
to prevent the PPC_BOOK3E_64=y && PPC_FSL_BOOK3E=y case.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We weren't using SYSCALL_DEFINE for sys_switch_endian(), which means
it wasn't able to be traced by CONFIG_FTRACE_SYSCALLS.
By using the macro we create the right metadata and the syscall is
visible. eg:
# cd /sys/kernel/debug/tracing
# echo 1 | tee events/syscalls/sys_*_switch_endian/enable
# ~/switch_endian_test
# cat trace
...
switch_endian_t-3604 [009] .... 315.175164: sys_switch_endian()
switch_endian_t-3604 [009] .... 315.175167: sys_switch_endian -> 0x5555aaaa5555aaaa
switch_endian_t-3604 [009] .... 315.175169: sys_switch_endian()
switch_endian_t-3604 [009] .... 315.175169: sys_switch_endian -> 0x5555aaaa5555aaaa
Fixes: 529d235a0e ("powerpc: Add a proper syscall for switching endianness")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
While the current kernel drivers/of/ code allows developers to be
sloppy and use a DTS status value of "ok", the current DTSpec 0.1
makes it clear that the proper spelling is "okay", so fix the small
number of PowerPC .dts files that do this.
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When updating page tables, we need to make sure we fill the page table
entry valid bits. We do this by or'ing in one of PGD/PUD/PMD_VAL_BITS.
The page table 'set' interfaces allow updating the raw value of page
table entries without setting the valid bits, so remove those
interfaces to avoid incorrect usage in future.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
[mpe: Reword commit message based on mailing list discussion]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Commit 8792468da5 "powerpc: Add the ability to save FPU without
giving it up" unexpectedly removed the MSR_FE0 and MSR_FE1 bits from
the bitmask used to update the MSR of the previous thread in
__giveup_fpu() causing a KVM-PR MacOS guest to lockup and panic the
host kernel.
Leaving FE0/1 enabled means unrelated processes might receive FPEs
when they're not expecting them and crash. In particular if this
happens to init the host will then panic.
eg (transcribed):
qemu-system-ppc[837]: unhandled signal 8 at 12cc9ce4 nip 12cc9ce4 lr 12cc9ca4 code 0
systemd[1]: unhandled signal 8 at 202f02e0 nip 202f02e0 lr 001003d4 code 0
Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
Reinstate these bits to the MSR bitmask to enable MacOS guests to run
under 32-bit KVM-PR once again without issue.
Fixes: 8792468da5 ("powerpc: Add the ability to save FPU without giving it up")
Cc: stable@vger.kernel.org # v4.6+
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The Processor Utilzation of Resource Registers (PURR) provide an
estimate of resources used by a cpu thread. Section 7.6 in Book III of
the ISA outlines how to calculate the percentage of shared resources
for threads using the ratio of the PURR delta and Timebase Register
delta for a sampled period.
This calculation is currently done erroneously by the lparstat tool
from the powerpc-utils package. This patch exports the current
timebase value after we sample the PURRs and exposes it to userspace
accounting tools via /proc/ppc64/lparcfg.
Signed-off-by: Tyrel Datwyler <tyreld@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The recent rework of PCI kconfig symbols exposed an existing bug in
the CURRITUCK kconfig logic.
It selects PPC4xx_PCI_EXPRESS which depends on PCI, but PCI is user
selectable and might be disabled, leading to a warning:
WARNING: unmet direct dependencies detected for PPC4xx_PCI_EXPRESS
Depends on [n]: PCI [=n] && 4xx [=y]
Selected by [y]:
- CURRITUCK [=y] && PPC_47x [=y]
Prior to commit eb01d42a77 ("PCI: consolidate PCI config entry in
drivers/pci") PCI was enabled by default for currituck_defconfig so we
didn't see the warning. The bad logic was still there, it just
required someone disabling PCI in their .config to hit it.
Fix it by forcing PCI on for CURRITUCK, which seems was always the
expectation anyway.
Fixes: eb01d42a77 ("PCI: consolidate PCI config entry in drivers/pci")
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch adds a debugfs interface to force scheduling a recovery event.
This can be used to recover a specific PE or schedule a "special" recovery
even that checks for errors at the PHB level.
To force a recovery of a normal PE, use:
echo '<#pe>:<#phb>' > /sys/kernel/debug/powerpc/eeh_force_recover
To force a scan for broken PHBs:
echo 'hwcheck' > /sys/kernel/debug/powerpc/eeh_force_recover
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Currently when we detect an error we automatically invoke the EEH recovery
handler. This can be annoying when debugging EEH problems, or when working
on EEH itself so this patch adds a debugfs knob that will prevent a
recovery event from being queued up when an issue is detected.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Add a helper to find the pci_controller structure based on the domain
number / phb id.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Reviewed-by: Sam Bobroff <sbobroff@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
To use this function at all #define DEBUG needs to be set in eeh_cache.c.
Considering that printing at pr_debug is probably not all that useful since
it adds the additional hurdle of requiring you to enable the debug print if
dynamic_debug is in use so this patch bumps it to pr_info.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Reviewed-by: Sam Bobroff <sbobroff@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Adds a debugfs file that can be read to view the contents of the EEH
address cache. This is pretty similar to the existing
eeh_addr_cache_print() function, but that function is intended to debug
issues inside of the kernel since it's #ifdef`ed out by default, and writes
into the kernel log.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Reviewed-by: Sam Bobroff <sbobroff@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The EEH address cache is used to map a physical MMIO address back to a PCI
device. It's useful to know when it's being manipulated, but currently this
requires recompiling with #define DEBUG set. This is pointless since we
have dynamic_debug nowdays, so remove the #ifdef guard and add a pr_debug()
for the remove case too.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Reviewed-by: Sam Bobroff <sbobroff@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
There's no need to the custom getter/setter functions so we should remove
them in favour of using the generic one. While we're here, change the type
of eeh_max_freeze to u32 and print the value in decimal rather than
hex because printing it in hex makes no sense.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Reviewed-by: Sam Bobroff <sbobroff@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Commit d4fde568a3 ("powerpc/64: Use optimized checksum routines on
little-endian") converted last powerpc user of GENERIC_CSUM.
This patch does a final cleanup dropping the Kconfig GENERIC_CSUM
option which is always 'n', and associated piece of code in
asm/checksum.h
Fixes: d4fde568a3 ("powerpc/64: Use optimized checksum routines on little-endian")
Reported-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The slbfee. instruction must have bit 24 of RB clear, failure to do
so can result in false negatives that result in incorrect assertions.
This is not obvious from the ISA v3.0B document, which only says:
The hardware ignores the contents of RB 36:38 40:63 -- p.1032
This patch fixes the bug and also clears all other bits from PPC bit
36-63, which is good practice when dealing with reserved or ignored
bits.
Fixes: e15a4fea4d ("powerpc/64s/hash: Add some SLB debugging tests")
Cc: stable@vger.kernel.org # v4.20+
Reported-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Tested-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch updates the kernel non-linear virtual map to 512TB when
we're built with 64K page size and are using the hash MMU. We allocate
one context for the vmalloc region and hence the max virtual area size
is limited by the context map size (512TB for 64K and 64TB for 4K page
size).
This patch fixes boot failures with large amounts of system RAM where
we need large vmalloc space to handle per cpu allocations.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Tested-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
GCC 8 warns about the logic in vr_get/set(), which with -Werror breaks
the build:
In function ‘user_regset_copyin’,
inlined from ‘vr_set’ at arch/powerpc/kernel/ptrace.c:628:9:
include/linux/regset.h:295:4: error: ‘memcpy’ offset [-527, -529] is
out of the bounds [0, 16] of object ‘vrsave’ with type ‘union
<anonymous>’ [-Werror=array-bounds]
arch/powerpc/kernel/ptrace.c: In function ‘vr_set’:
arch/powerpc/kernel/ptrace.c:623:5: note: ‘vrsave’ declared here
} vrsave;
This has been identified as a regression in GCC, see GCC bug 88273.
However we can avoid the warning and also simplify the logic and make
it more robust.
Currently we pass -1 as end_pos to user_regset_copyout(). This says
"copy up to the end of the regset".
The definition of the regset is:
[REGSET_VMX] = {
.core_note_type = NT_PPC_VMX, .n = 34,
.size = sizeof(vector128), .align = sizeof(vector128),
.active = vr_active, .get = vr_get, .set = vr_set
},
The end is calculated as (n * size), ie. 34 * sizeof(vector128).
In vr_get/set() we pass start_pos as 33 * sizeof(vector128), meaning
we can copy up to sizeof(vector128) into/out-of vrsave.
The on-stack vrsave is defined as:
union {
elf_vrreg_t reg;
u32 word;
} vrsave;
And elf_vrreg_t is:
typedef __vector128 elf_vrreg_t;
So there is no bug, but we rely on all those sizes lining up,
otherwise we would have a kernel stack exposure/overwrite on our
hands.
Rather than relying on that we can pass an explict end_pos based on
the sizeof(vrsave). The result should be exactly the same but it's
more obviously not over-reading/writing the stack and it avoids the
compiler warning.
Reported-by: Meelis Roos <mroos@linux.ee>
Reported-by: Mathieu Malaterre <malat@debian.org>
Cc: stable@vger.kernel.org
Tested-by: Mathieu Malaterre <malat@debian.org>
Tested-by: Meelis Roos <mroos@linux.ee>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The change_pte() notifier was designed to use as a quick path to
update secondary MMU PTEs on write permission changes or PFN changes.
For KVM, it could reduce the vm-exits when vcpu faults on the pages
that was touched up by KSM. It's not used to do cache invalidations,
for example, if we see the notifier will be called before the real PTE
update after all (please see set_pte_at_notify that set_pte_at was
called later).
All the necessary cache invalidation should all be done in
invalidate_range() already.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Alistair Popple <alistair@popple.id.au>
Reviewed-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds an "in_guest" parameter to machine_check_print_event_info()
so that we can avoid trying to translate guest NIP values into
symbolic form using the host kernel's symbol table.
Reviewed-by: Aravinda Prasad <aravinda@linux.vnet.ibm.com>
Reviewed-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This makes the handling of machine check interrupts that occur inside
a guest simpler and more robust, with less done in assembler code and
in real mode.
Now, when a machine check occurs inside a guest, we always get the
machine check event struct and put a copy in the vcpu struct for the
vcpu where the machine check occurred. We no longer call
machine_check_queue_event() from kvmppc_realmode_mc_power7(), because
on POWER8, when a vcpu is running on an offline secondary thread and
we call machine_check_queue_event(), that calls irq_work_queue(),
which doesn't work because the CPU is offline, but instead triggers
the WARN_ON(lazy_irq_pending()) in pnv_smp_cpu_kill_self() (which
fires again and again because nothing clears the condition).
All that machine_check_queue_event() actually does is to cause the
event to be printed to the console. For a machine check occurring in
the guest, we now print the event in kvmppc_handle_exit_hv()
instead.
The assembly code at label machine_check_realmode now just calls C
code and then continues exiting the guest. We no longer either
synthesize a machine check for the guest in assembly code or return
to the guest without a machine check.
The code in kvmppc_handle_exit_hv() is extended to handle the case
where the guest is not FWNMI-capable. In that case we now always
synthesize a machine check interrupt for the guest. Previously, if
the host thinks it has recovered the machine check fully, it would
return to the guest without any notification that the machine check
had occurred. If the machine check was caused by some action of the
guest (such as creating duplicate SLB entries), it is much better to
tell the guest that it has caused a problem. Therefore we now always
generate a machine check interrupt for guests that are not
FWNMI-capable.
Reviewed-by: Aravinda Prasad <aravinda@linux.vnet.ibm.com>
Reviewed-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
kvmhv_p9_guest_entry() implements a fast-path guest entry for Power9
when guest and host are both running with the Radix MMU.
Currently in that path we don't save the host AMR (Authority Mask
Register) value, and we always restore 0 on return to the host. That
is OK at the moment because the AMR is not used for storage keys with
the Radix MMU.
However we plan to start using the AMR on Radix to prevent the kernel
from reading/writing to userspace outside of copy_to/from_user(). In
order to make that work we need to save/restore the AMR value.
We only restore the value if it is different from the guest value,
which is already in the register when we exit to the host. This should
mean we rarely need to actually restore the value when running a
modern Linux as a guest, because it will be using the same value as
us.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Russell Currey <ruscur@russell.cc>
grow_halt_poll_ns() have a strange behaviour in case
(vcpu->halt_poll_ns != 0) &&
(vcpu->halt_poll_ns < halt_poll_ns_grow_start).
In this case, vcpu->halt_poll_ns will be multiplied by grow factor
(halt_poll_ns_grow) which will require several grow iteration in order
to reach a value bigger than halt_poll_ns_grow_start.
This means that growing vcpu->halt_poll_ns from value of 0 is slower
than growing it from a positive value less than halt_poll_ns_grow_start.
Which is misleading and inaccurate.
Fix issue by changing grow_halt_poll_ns() to set vcpu->halt_poll_ns
to halt_poll_ns_grow_start in any case that
(vcpu->halt_poll_ns < halt_poll_ns_grow_start).
Regardless if vcpu->halt_poll_ns is 0.
use READ_ONCE to get a consistent number for all cases.
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Liran Alon <liran.alon@oracle.com>
Signed-off-by: Nir Weiner <nir.weiner@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The hard-coded value 10000 in grow_halt_poll_ns() stands for the initial
start value when raising up vcpu->halt_poll_ns.
It actually sets the first timeout to the first polling session.
This value has significant effect on how tolerant we are to outliers.
On the standard case, higher value is better - we will spend more time
in the polling busyloop, handle events/interrupts faster and result
in better performance.
But on outliers it puts us in a busy loop that does nothing.
Even if the shrink factor is zero, we will still waste time on the first
iteration.
The optimal value changes between different workloads. It depends on
outliers rate and polling sessions length.
As this value has significant effect on the dynamic halt-polling
algorithm, it should be configurable and exposed.
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Liran Alon <liran.alon@oracle.com>
Signed-off-by: Nir Weiner <nir.weiner@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
grow_halt_poll_ns() have a strange behavior in case
(halt_poll_ns_grow == 0) && (vcpu->halt_poll_ns != 0).
In this case, vcpu->halt_pol_ns will be set to zero.
That results in shrinking instead of growing.
Fix issue by changing grow_halt_poll_ns() to not modify
vcpu->halt_poll_ns in case halt_poll_ns_grow is zero
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Liran Alon <liran.alon@oracle.com>
Signed-off-by: Nir Weiner <nir.weiner@oracle.com>
Suggested-by: Liran Alon <liran.alon@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
kvm_arch_memslots_updated() is at this point in time an x86-specific
hook for handling MMIO generation wraparound. x86 stashes 19 bits of
the memslots generation number in its MMIO sptes in order to avoid
full page fault walks for repeat faults on emulated MMIO addresses.
Because only 19 bits are used, wrapping the MMIO generation number is
possible, if unlikely. kvm_arch_memslots_updated() alerts x86 that
the generation has changed so that it can invalidate all MMIO sptes in
case the effective MMIO generation has wrapped so as to avoid using a
stale spte, e.g. a (very) old spte that was created with generation==0.
Given that the purpose of kvm_arch_memslots_updated() is to prevent
consuming stale entries, it needs to be called before the new generation
is propagated to memslots. Invalidating the MMIO sptes after updating
memslots means that there is a window where a vCPU could dereference
the new memslots generation, e.g. 0, and incorrectly reuse an old MMIO
spte that was created with (pre-wrap) generation==0.
Fixes: e59dbe09f8 ("KVM: Introduce kvm_arch_memslots_updated()")
Cc: <stable@vger.kernel.org>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
All new 32-bit architectures should have 64-bit userspace off_t type, but
existing architectures has 32-bit ones.
To enforce the rule, new config option is added to arch/Kconfig that defaults
ARCH_32BIT_OFF_T to be disabled for new 32-bit architectures. All existing
32-bit architectures enable it explicitly.
New option affects force_o_largefile() behaviour. Namely, if userspace
off_t is 64-bits long, we have no reason to reject user to open big files.
Note that even if architectures has only 64-bit off_t in the kernel
(arc, c6x, h8300, hexagon, nios2, openrisc, and unicore32),
a libc may use 32-bit off_t, and therefore want to limit the file size
to 4GB unless specified differently in the open flags.
Signed-off-by: Yury Norov <ynorov@caviumnetworks.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Yury Norov <ynorov@marvell.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
This adds an entry to the kvm_stats_debugfs directory which provides the
number of large (2M or 1G) pages which have been used to setup the guest
mappings, for radix guests.
Signed-off-by: Suraj Jitindar Singh <sjitindarsingh@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The SPAPR TCE KVM device references all hardware IOMMU tables assigned to
some IOMMU group to ensure that in-kernel KVM acceleration of H_PUT_TCE
can work. The tables are references when an IOMMU group gets registered
with the VFIO KVM device by the KVM_DEV_VFIO_GROUP_ADD ioctl;
KVM_DEV_VFIO_GROUP_DEL calls into the dereferencing code
in kvm_spapr_tce_release_iommu_group() which walks through the list of
LIOBNs, finds a matching IOMMU table and calls kref_put() when found.
However that code stops after the very first successful derefencing
leaving other tables referenced till the SPAPR TCE KVM device is destroyed
which normally happens on guest reboot or termination so if we do hotplug
and unplug in a loop, we are leaking IOMMU tables here.
This removes a premature return to let kvm_spapr_tce_release_iommu_group()
find and dereference all attached tables.
Fixes: 121f80ba68 ("KVM: PPC: VFIO: Add in-kernel acceleration for VFIO")
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Devices on the KVM_FAST_MMIO_BUS by definition have length zero and are
thus used for notification purposes rather than data transfer. For
example eventfd for virtio devices.
This means that when emulating mmio instructions which target devices on
this bus we can immediately handle them and return without needing to load
the instruction from guest memory.
For now we restrict this to stores as this is the only use case at
present.
For a normal guest the effect is negligible, however for a nested guest
we save on the order of 5us per access.
Signed-off-by: Suraj Jitindar Singh <sjitindarsingh@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Currently, the KVM code assumes that if the host kernel is using the
XIVE interrupt controller (the new interrupt controller that first
appeared in POWER9 systems), then the in-kernel XICS emulation will
use the XIVE hardware to deliver interrupts to the guest. However,
this only works when the host is running in hypervisor mode and has
full access to all of the XIVE functionality. It doesn't work in any
nested virtualization scenario, either with PR KVM or nested-HV KVM,
because the XICS-on-XIVE code calls directly into the native-XIVE
routines, which are not initialized and cannot function correctly
because they use OPAL calls, and OPAL is not available in a guest.
This means that using the in-kernel XICS emulation in a nested
hypervisor that is using XIVE as its interrupt controller will cause a
(nested) host kernel crash. To fix this, we change most of the places
where the current code calls xive_enabled() to select between the
XICS-on-XIVE emulation and the plain XICS emulation to call a new
function, xics_on_xive(), which returns false in a guest.
However, there is a further twist. The plain XICS emulation has some
functions which are used in real mode and access the underlying XICS
controller (the interrupt controller of the host) directly. In the
case of a nested hypervisor, this means doing XICS hypercalls
directly. When the nested host is using XIVE as its interrupt
controller, these hypercalls will fail. Therefore this also adds
checks in the places where the XICS emulation wants to access the
underlying interrupt controller directly, and if that is XIVE, makes
the code use the virtual mode fallback paths, which call generic
kernel infrastructure rather than doing direct XICS access.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The header search path -I. in kernel Makefiles is very suspicious;
it allows the compiler to search for headers in the top of $(srctree),
where obviously no header file exists.
Commit 46f43c6ee0 ("KVM: powerpc: convert marker probes to event
trace") first added these options, but they are completely useless.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Recent kernels, since commit e15a4fea4d ("powerpc/64s/hash: Add
some SLB debugging tests", 2018-10-03) use the slbfee. instruction,
which PR KVM currently does not have code to emulate. Consequently
recent kernels fail to boot under PR KVM. This adds emulation of
slbfee., enabling these kernels to boot successfully.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The compound IOMMU group rework moved iommu_register_group() together
in pnv_pci_ioda_setup_iommu_api() (which is a part of
ppc_md.pcibios_fixup). As the result, pnv_ioda_setup_bus_iommu_group()
does not create groups any more, it only adds devices to groups.
This works fine for boot time devices. However IOMMU groups for
SRIOV's VFs were added by pnv_ioda_setup_bus_iommu_group() so this got
broken: pnv_tce_iommu_bus_notifier() expects a group to be registered
for VF and it is not.
This adds missing group registration and adds a NULL pointer check
into the bus notifier so we won't crash if there is no group, although
it is not expected to happen now because of the change above.
Example oops seen prior to this patch:
$ echo 1 > /sys/bus/pci/devices/0000\:01\:00.0/sriov_numvfs
Unable to handle kernel paging request for data at address 0x00000030
Faulting instruction address: 0xc0000000004a6018
Oops: Kernel access of bad area, sig: 11 [#1]
LE SMP NR_CPUS=2048 NUMA PowerNV
CPU: 46 PID: 7006 Comm: bash Not tainted 4.15-ish
NIP: c0000000004a6018 LR: c0000000004a6014 CTR: 0000000000000000
REGS: c000008fc876b400 TRAP: 0300 Not tainted (4.15-ish)
MSR: 900000000280b033 <SF,HV,VEC,VSX,EE,FP,ME,IR,DR,RI,LE>
CFAR: c000000000d0be20 DAR: 0000000000000030 DSISR: 40000000 SOFTE: 1
...
NIP sysfs_do_create_link_sd.isra.0+0x68/0x150
LR sysfs_do_create_link_sd.isra.0+0x64/0x150
Call Trace:
pci_dev_type+0x0/0x30 (unreliable)
iommu_group_add_device+0x8c/0x600
iommu_add_device+0xe8/0x180
pnv_tce_iommu_bus_notifier+0xb0/0xf0
notifier_call_chain+0x9c/0x110
blocking_notifier_call_chain+0x64/0xa0
device_add+0x524/0x7d0
pci_device_add+0x248/0x450
pci_iov_add_virtfn+0x294/0x3e0
pci_enable_sriov+0x43c/0x580
mlx5_core_sriov_configure+0x15c/0x2f0 [mlx5_core]
sriov_numvfs_store+0x180/0x240
dev_attr_store+0x3c/0x60
sysfs_kf_write+0x64/0x90
kernfs_fop_write+0x1ac/0x240
__vfs_write+0x3c/0x70
vfs_write+0xd8/0x220
SyS_write+0x6c/0x110
system_call+0x58/0x6c
Fixes: 0bd971676e ("powerpc/powernv/npu: Add compound IOMMU groups")
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reported-by: Santwana Samantray <santwana.samantray@in.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
There is no need to provide anything but get_arch_dma_ops to
<linux/dma-mapping.h>. More the remaining declarations to <asm/iommu.h>
and drop all the includes.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
There is no good reason for this helper, just opencode it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Just fold the calculation into __phys_to_dma/__dma_to_phys as those are
the only places that should know about it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Now that we've switched all the powerpc nommu and swiotlb methods to
use the generic dma_direct_* calls we can remove these ops vectors
entirely and rely on the common direct mapping bypass that avoids
indirect function calls entirely. This also allows to remove a whole
lot of boilerplate code related to setting up these operations.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Switch the streaming DMA mapping and ownership transfer methods to the
functionally identical dma_direct_ versions. Factor the cache
maintainance helpers into the form expected by the common code for that.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The generic code allows a few nice things such as node local allocations
and dipping into the CMA area. The lookup of the right zone for a given
dma mask works a little different, but the results should be the same.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The only user left is powerpc, but even there the generic dma-direct
version works just as well, given that we guarantee that the swiotlb
buffer must always be addressable.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This function is largely identical to the generic version used
everywhere else. Replace it with the generic version.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This function is identical to the generic dma_direct_get_required_mask,
except that the generic version also takes the bus_dma_mask account,
which could lead to incorrect results in the powerpc version.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The coherent cache version of this function already is functionally
identicall to the default version, and by defining the
arch_dma_coherent_to_pfn hook the same is ture for the noncoherent
version as well.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Use the standard portable helper instead of the powerpc specific one,
which is about to go away.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Instead of letting the architecture supply all of dma_set_mask just
give it an additional hook selected by Kconfig.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We need to compare the last byte in the dma range and not the one after it
for the bus_dma_mask, just like we do for the regular dma_mask. Fix this
cleanly by merging the two comparisms into one.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The max_direct_dma_addr duplicates the bus_dma_mask field in struct
device. Use the generic field instead.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
pci_dma_dev_setup_swiotlb is only used by the fsl_pci code, and closely
related to it, so fsl_pci.c seems like a better place for it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This function is only used by the Cell iommu code, which can keep track
if it is using the iommu internally just as good.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
All iommu capable platforms now always use the iommu code with the
internal bypass, so there is not need for this magic anymore.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The ppc_md and pci_controller_ops methods are unused now and can be
removed. The dma_nommu implementation is generic to the generic one
except for using max_pfn instead of calling into the memblock API,
and all other dma_map_ops instances implement a method of their own.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Use the generic iommu bypass code instead of overriding set_dma_mask.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
These devices are not PCIe devices and do not have associated dma map
ops, so this is just dead code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This function is completely bogus - the fact that two PCIe devices come
from the same vendor has absolutely nothing to say about the DMA
capabilities and characteristics.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Use the generic iommu bypass code instead of overriding set_dma_mask.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
If dart_init failed we didn't have a chance to setup dma or controller
ops yet, so there is no point in resetting them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This gets rid of a lot of clumsy code and finally allows us to mark
dma_iommu_ops const.
Includes fixes from Michael Ellerman.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Configure the dma settings at device setup time, and stop playing games
with get_pci_dma_ops. This prepares for using the common dma_configure
code later on.
Includes fixes from Michael Ellerman.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Use the generic iommu bypass code instead of overriding set_dma_mask.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Call dma_get_required_mask_pSeriesLP directly instead of dma_iommu_ops
to simply the code a bit.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Add a new iommu_bypass flag to struct dev_archdata so that the dma_iommu
implementation can handle the direct mapping transparently instead of
switiching ops around. Setting of this flag is controlled by new
pci_controller_ops method.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
vio_dma_mapping_ops currently does a lot of indirect calls through
dma_iommu_ops, which not only make the code harder to follow but are
also expensive in the post-spectre world. Unwind the indirect calls
by calling the ppc_iommu_* or iommu_* APIs directly applicable, or
just use the dma_iommu_* methods directly where we can.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
