This is in preparation to adding support for kuap with hash translation.
In preparation for that rename/move kuap related functions to
non radix names. Also move the feature bit closer to MMU_FTR_KUEP.
MMU_FTR_KUEP is renamed to MMU_FTR_BOOK3S_KUEP to indicate the feature
is only relevant to BOOK3S_64
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201127044424.40686-8-aneesh.kumar@linux.ibm.com
When offlining a CPU, powerpc/64s does not flush TLBs, rather it just
leaves the CPU set in mm_cpumasks, so it continues to receive TLBIEs
to manage its TLBs.
However the exit_flush_lazy_tlbs() function expects that after
returning, all CPUs (except self) have flushed TLBs for that mm, in
which case TLBIEL can be used for this flush. This breaks for offline
CPUs because they don't get the IPI to flush their TLB. This can lead
to stale translations.
Fix this by clearing the CPU from mm_cpumasks, then flushing all TLBs
before going offline.
These offlined CPU bits stuck in the cpumask also prevents the cpumask
from being trimmed back to local mode, which means continual broadcast
IPIs or TLBIEs are needed for TLB flushing. This patch prevents that
situation too.
A cast of many were involved in working this out, but in particular
Milton, Aneesh, Paul made key discoveries.
Fixes: 0cef77c779 ("powerpc/64s/radix: flush remote CPUs out of single-threaded mm_cpumask")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Debugged-by: Milton Miller <miltonm@us.ibm.com>
Debugged-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Debugged-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201126102530.691335-5-npiggin@gmail.com
Using DECLARE_STATIC_KEY_FALSE needs linux/jump_table.h.
Otherwise the build fails with eg:
arch/powerpc/include/asm/book3s/64/kup-radix.h:66:1: warning: data definition has no type or storage class
66 | DECLARE_STATIC_KEY_FALSE(uaccess_flush_key);
Fixes: 9a32a7e78b ("powerpc/64s: flush L1D after user accesses")
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
[mpe: Massage change log]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201123184016.693fe464@canb.auug.org.au
In kup.h we currently include kup-radix.h for all 64-bit builds, which
includes Book3S and Book3E. The latter doesn't make sense, Book3E
never uses the Radix MMU.
This has worked up until now, but almost by accident, and the recent
uaccess flush changes introduced a build breakage on Book3E because of
the bad structure of the code.
So disentangle things so that we only use kup-radix.h for Book3S. This
requires some more stubs in kup.h and fixing an include in
syscall_64.c.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
IBM Power9 processors can speculatively operate on data in the L1 cache
before it has been completely validated, via a way-prediction mechanism. It
is not possible for an attacker to determine the contents of impermissible
memory using this method, since these systems implement a combination of
hardware and software security measures to prevent scenarios where
protected data could be leaked.
However these measures don't address the scenario where an attacker induces
the operating system to speculatively execute instructions using data that
the attacker controls. This can be used for example to speculatively bypass
"kernel user access prevention" techniques, as discovered by Anthony
Steinhauser of Google's Safeside Project. This is not an attack by itself,
but there is a possibility it could be used in conjunction with
side-channels or other weaknesses in the privileged code to construct an
attack.
This issue can be mitigated by flushing the L1 cache between privilege
boundaries of concern. This patch flushes the L1 cache after user accesses.
This is part of the fix for CVE-2020-4788.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
powerpc used to set the PTE specific flags in set_pte_at(). That is
different from other architectures. To be consistent with other
architectures powerpc updated pfn_pte() to set _PAGE_PTE in commit
379c926d63 ("powerpc/mm: move setting pte specific flags to
pfn_pte")
That commit didn't do the same for pfn_pmd() because we expect
pmd_mkhuge() to do that. But as per Linus that is a bad rule:
The rule that you must use "pmd_mkhuge()" seems _completely_ wrong.
The only valid use to ever make a pmd out of a pfn is to make a
huge-page.
Hence update pfn_pmd() to set _PAGE_PTE.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201022091115.39568-1-aneesh.kumar@linux.ibm.com
Pull powerpc updates from Michael Ellerman:
- A series from Nick adding ARCH_WANT_IRQS_OFF_ACTIVATE_MM & selecting
it for powerpc, as well as a related fix for sparc.
- Remove support for PowerPC 601.
- Some fixes for watchpoints & addition of a new ptrace flag for
detecting ISA v3.1 (Power10) watchpoint features.
- A fix for kernels using 4K pages and the hash MMU on bare metal
Power9 systems with > 16TB of RAM, or RAM on the 2nd node.
- A basic idle driver for shallow stop states on Power10.
- Tweaks to our sched domains code to better inform the scheduler about
the hardware topology on Power9/10, where two SMT4 cores can be
presented by firmware as an SMT8 core.
- A series doing further reworks & cleanups of our EEH code.
- Addition of a filter for RTAS (firmware) calls done via sys_rtas(),
to prevent root from overwriting kernel memory.
- Other smaller features, fixes & cleanups.
Thanks to: Alexey Kardashevskiy, Andrew Donnellan, Aneesh Kumar K.V,
Athira Rajeev, Biwen Li, Cameron Berkenpas, Cédric Le Goater, Christophe
Leroy, Christoph Hellwig, Colin Ian King, Daniel Axtens, David Dai, Finn
Thain, Frederic Barrat, Gautham R. Shenoy, Greg Kurz, Gustavo Romero,
Ira Weiny, Jason Yan, Joel Stanley, Jordan Niethe, Kajol Jain, Konrad
Rzeszutek Wilk, Laurent Dufour, Leonardo Bras, Liu Shixin, Luca
Ceresoli, Madhavan Srinivasan, Mahesh Salgaonkar, Nathan Lynch, Nicholas
Mc Guire, Nicholas Piggin, Nick Desaulniers, Oliver O'Halloran, Pedro
Miraglia Franco de Carvalho, Pratik Rajesh Sampat, Qian Cai, Qinglang
Miao, Ravi Bangoria, Russell Currey, Satheesh Rajendran, Scott Cheloha,
Segher Boessenkool, Srikar Dronamraju, Stan Johnson, Stephen Kitt,
Stephen Rothwell, Thiago Jung Bauermann, Tyrel Datwyler, Vaibhav Jain,
Vaidyanathan Srinivasan, Vasant Hegde, Wang Wensheng, Wolfram Sang, Yang
Yingliang, zhengbin.
* tag 'powerpc-5.10-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (228 commits)
Revert "powerpc/pci: unmap legacy INTx interrupts when a PHB is removed"
selftests/powerpc: Fix eeh-basic.sh exit codes
cpufreq: powernv: Fix frame-size-overflow in powernv_cpufreq_reboot_notifier
powerpc/time: Make get_tb() common to PPC32 and PPC64
powerpc/time: Make get_tbl() common to PPC32 and PPC64
powerpc/time: Remove get_tbu()
powerpc/time: Avoid using get_tbl() and get_tbu() internally
powerpc/time: Make mftb() common to PPC32 and PPC64
powerpc/time: Rename mftbl() to mftb()
powerpc/32s: Remove #ifdef CONFIG_PPC_BOOK3S_32 in head_book3s_32.S
powerpc/32s: Rename head_32.S to head_book3s_32.S
powerpc/32s: Setup the early hash table at all time.
powerpc/time: Remove ifdef in get_dec() and set_dec()
powerpc: Remove get_tb_or_rtc()
powerpc: Remove __USE_RTC()
powerpc: Tidy up a bit after removal of PowerPC 601.
powerpc: Remove support for PowerPC 601
powerpc: Remove PowerPC 601
powerpc: Drop SYNC_601() ISYNC_601() and SYNC()
powerpc: Remove CONFIG_PPC601_SYNC_FIX
...
powerpc used to set the pte specific flags in set_pte_at(). This is
different from other architectures. To be consistent with other
architecture update pfn_pte to set _PAGE_PTE on ppc64. Also, drop now
unused pte_mkpte.
We add a VM_WARN_ON() to catch the usage of calling set_pte_at() without
setting _PAGE_PTE bit. We will remove that after a few releases.
With respect to huge pmd entries, pmd_mkhuge() takes care of adding the
_PAGE_PTE bit.
[akpm@linux-foundation.org: whitespace fix, per Christophe]
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lkml.kernel.org/r/20200902114222.181353-3-aneesh.kumar@linux.ibm.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
MAX_PHYSMEM #define is used along with sparsemem to determine the SECTION_SHIFT
value. Powerpc also uses the same value to limit the max memory enabled on the
system. With 4K PAGE_SIZE and hash translation mode, we want to limit the max
memory enabled to 64TB due to page table size restrictions. However, with
radix translation, we don't have these restrictions. Hence split the radix
and hash MA_PHYSMEM limit and use different limit for each of them.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200608070904.387440-4-aneesh.kumar@linux.ibm.com
With commit: 0034d395f8 ("powerpc/mm/hash64: Map all the kernel
regions in the same 0xc range"), we now split the 64TB address range
into 4 contexts each of 16TB. That implies we can do only 16TB linear
mapping.
On some systems, eg. Power9, memory attached to nodes > 0 will appear
above 16TB in the linear mapping. This resulted in kernel crash when
we boot such systems in hash translation mode with 4K PAGE_SIZE.
This patch updates the kernel mapping such that we now start supporting upto
61TB of memory with 4K. The kernel mapping now looks like below 4K PAGE_SIZE
and hash translation.
vmalloc start = 0xc0003d0000000000
IO start = 0xc0003e0000000000
vmemmap start = 0xc0003f0000000000
Our MAX_PHYSMEM_BITS for 4K is still 64TB even though we can only map 61TB.
We prevent bolt mapping anything outside 61TB range by checking against
H_VMALLOC_START.
Fixes: 0034d395f8 ("powerpc/mm/hash64: Map all the kernel regions in the same 0xc range")
Reported-by: Cameron Berkenpas <cam@neo-zeon.de>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200608070904.387440-3-aneesh.kumar@linux.ibm.com
If the hypervisor doesn't support hugepages, the kernel ends up allocating a large
number of page table pages. The early page table allocation was wrongly
setting the max memblock limit to ppc64_rma_size with radix translation
which resulted in boot failure as shown below.
Kernel panic - not syncing:
early_alloc_pgtable: Failed to allocate 16777216 bytes align=0x1000000 nid=-1 from=0x0000000000000000 max_addr=0xffffffffffffffff
CPU: 0 PID: 0 Comm: swapper Not tainted 5.8.0-24.9-default+ #2
Call Trace:
[c0000000016f3d00] [c0000000007c6470] dump_stack+0xc4/0x114 (unreliable)
[c0000000016f3d40] [c00000000014c78c] panic+0x164/0x418
[c0000000016f3dd0] [c000000000098890] early_alloc_pgtable+0xe0/0xec
[c0000000016f3e60] [c0000000010a5440] radix__early_init_mmu+0x360/0x4b4
[c0000000016f3ef0] [c000000001099bac] early_init_mmu+0x1c/0x3c
[c0000000016f3f10] [c00000000109a320] early_setup+0x134/0x170
This was because the kernel was checking for the radix feature before we enable the
feature via mmu_features. This resulted in the kernel using hash restrictions on
radix.
Rework the early init code such that the kernel boot with memblock restrictions
as imposed by hash. At that point, the kernel still hasn't finalized the
translation the kernel will end up using.
We have three different ways of detecting radix.
1. dt_cpu_ftrs_scan -> used only in case of PowerNV
2. ibm,pa-features -> Used when we don't use cpu_dt_ftr_scan
3. CAS -> Where we negotiate with hypervisor about the supported translation.
We look at 1 or 2 early in the boot and after that, we look at the CAS vector to
finalize the translation the kernel will use. We also support a kernel command
line option (disable_radix) to switch to hash.
Update the memblock limit after mmu_early_init_devtree() if the kernel is going
to use radix translation. This forces some of the memblock allocations we do before
mmu_early_init_devtree() to be within the RMA limit.
Fixes: 2bfd65e45e ("powerpc/mm/radix: Add radix callbacks for early init routines")
Reported-by: Shirisha Ganta <shiganta@in.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Reviewed-by: Hari Bathini <hbathini@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200828100852.426575-1-aneesh.kumar@linux.ibm.com
ISA v3.1 does not support the SAO storage control attribute required to
implement PROT_SAO. PROT_SAO was used by specialised system software
(Lx86) that has been discontinued for about 7 years, and is not thought
to be used elsewhere, so removal should not cause problems.
We rather remove it than keep support for older processors, because
live migrating guest partitions to newer processors may not be possible
if SAO is in use (or worse allowed with silent races).
- PROT_SAO stays in the uapi header so code using it would still build.
- arch_validate_prot() is removed, the generic version rejects PROT_SAO
so applications would get a failure at mmap() time.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Drop KVM change for the time being]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200703011958.1166620-3-npiggin@gmail.com
UAMOR values are not application-specific. The kernel initializes
its value based on different reserved keys. Remove the thread-specific
UAMOR value and don't switch the UAMOR on context switch.
Move UAMOR initialization to key initialization code and remove
thread_struct.uamor because it is not used anymore.
Before commit: 4a4a5e5d2a ("powerpc/pkeys: key allocation/deallocation must not change pkey registers")
we used to update uamor based on key allocation and free.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200709032946.881753-20-aneesh.kumar@linux.ibm.com
As we kexec across kernels that use AMR/IAMR for different purposes
we need to ensure that new kernels get kexec'd with a reset value
of AMR/IAMR. For ex: the new kernel can use key 0 for kernel mapping and the old
AMR value prevents access to key 0.
This patch also removes reset if IAMR and AMOR in kexec_sequence. Reset of AMOR
is not needed and the IAMR reset is partial (it doesn't do the reset
on secondary cpus) and is redundant with this patch.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200709032946.881753-19-aneesh.kumar@linux.ibm.com
To enable memory unplug without splitting kernel page table
mapping, we force the max mapping size to the LMB size. LMB
size is the unit in which hypervisor will do memory add/remove
operation.
Pseries systems supports max LMB size of 256MB. Hence on pseries,
we now end up mapping memory with 2M page size instead of 1G. To improve
that we want hypervisor to hint the kernel about the hotplug
memory range. That was added that as part of
commit b6eca183e2 ("powerpc/kernel: Enables memory
hot-remove after reboot on pseries guests")
But PowerVM doesn't provide that hint yet. Once we get PowerVM
updated, we can then force the 2M mapping only to hot-pluggable
memory region using memblock_is_hotpluggable(). Till then
let's depend on LMB size for finding the mapping page size
for linear range.
With this change KVM guest will also be doing linear mapping with
2M page size.
The actual TLB benefit of mapping guest page table entries with
hugepage size can only be materialized if the partition scoped
entries are also using the same or higher page size. A guest using
1G hugetlbfs backing guest memory can have a performance impact with
the above change.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
[mpe: Fold in fix from Aneesh spotted by lkp@intel.com]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200709131925.922266-5-aneesh.kumar@linux.ibm.com
We can hit the following BUG_ON during memory unplug:
kernel BUG at arch/powerpc/mm/book3s64/pgtable.c:342!
Oops: Exception in kernel mode, sig: 5 [#1]
LE PAGE_SIZE=64K MMU=Radix SMP NR_CPUS=2048 NUMA pSeries
NIP [c000000000093308] pmd_fragment_free+0x48/0xc0
LR [c00000000147bfec] remove_pagetable+0x578/0x60c
Call Trace:
0xc000008050000000 (unreliable)
remove_pagetable+0x384/0x60c
radix__remove_section_mapping+0x18/0x2c
remove_section_mapping+0x1c/0x3c
arch_remove_memory+0x11c/0x180
try_remove_memory+0x120/0x1b0
__remove_memory+0x20/0x40
dlpar_remove_lmb+0xc0/0x114
dlpar_memory+0x8b0/0xb20
handle_dlpar_errorlog+0xc0/0x190
pseries_hp_work_fn+0x2c/0x60
process_one_work+0x30c/0x810
worker_thread+0x98/0x540
kthread+0x1c4/0x1d0
ret_from_kernel_thread+0x5c/0x74
This occurs when unplug is attempted for such memory which has
been mapped using memblock pages as part of early kernel page
table setup. We wouldn't have initialized the PMD or PTE fragment
count for those PMD or PTE pages.
This can be fixed by allocating memory in PAGE_SIZE granularity
during early page table allocation. This makes sure a specific
page is not shared for another memblock allocation and we can
free them correctly on removing page-table pages.
Since we now do PAGE_SIZE allocations for both PUD table and
PMD table (Note that PTE table allocation is already of PAGE_SIZE),
we end up allocating more memory for the same amount of system RAM.
Here is a comparision of how much more we need for a 64T and 2G
system after this patch:
1. 64T system
-------------
64T RAM would need 64G for vmemmap with struct page size being 64B.
128 PUD tables for 64T memory (1G mappings)
1 PUD table and 64 PMD tables for 64G vmemmap (2M mappings)
With default PUD[PMD]_TABLE_SIZE(4K), (128+1+64)*4K=772K
With PAGE_SIZE(64K) table allocations, (128+1+64)*64K=12352K
2. 2G system
------------
2G RAM would need 2M for vmemmap with struct page size being 64B.
1 PUD table for 2G memory (1G mapping)
1 PUD table and 1 PMD table for 2M vmemmap (2M mappings)
With default PUD[PMD]_TABLE_SIZE(4K), (1+1+1)*4K=12K
With new PAGE_SIZE(64K) table allocations, (1+1+1)*64K=192K
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200709131925.922266-2-aneesh.kumar@linux.ibm.com
Pull powerpc updates from Michael Ellerman:
- Support for userspace to send requests directly to the on-chip GZIP
accelerator on Power9.
- Rework of our lockless page table walking (__find_linux_pte()) to
make it safe against parallel page table manipulations without
relying on an IPI for serialisation.
- A series of fixes & enhancements to make our machine check handling
more robust.
- Lots of plumbing to add support for "prefixed" (64-bit) instructions
on Power10.
- Support for using huge pages for the linear mapping on 8xx (32-bit).
- Remove obsolete Xilinx PPC405/PPC440 support, and an associated sound
driver.
- Removal of some obsolete 40x platforms and associated cruft.
- Initial support for booting on Power10.
- Lots of other small features, cleanups & fixes.
Thanks to: Alexey Kardashevskiy, Alistair Popple, Andrew Donnellan,
Andrey Abramov, Aneesh Kumar K.V, Balamuruhan S, Bharata B Rao, Bulent
Abali, Cédric Le Goater, Chen Zhou, Christian Zigotzky, Christophe
JAILLET, Christophe Leroy, Dmitry Torokhov, Emmanuel Nicolet, Erhard F.,
Gautham R. Shenoy, Geoff Levand, George Spelvin, Greg Kurz, Gustavo A.
R. Silva, Gustavo Walbon, Haren Myneni, Hari Bathini, Joel Stanley,
Jordan Niethe, Kajol Jain, Kees Cook, Leonardo Bras, Madhavan
Srinivasan., Mahesh Salgaonkar, Markus Elfring, Michael Neuling, Michal
Simek, Nathan Chancellor, Nathan Lynch, Naveen N. Rao, Nicholas Piggin,
Oliver O'Halloran, Paul Mackerras, Pingfan Liu, Qian Cai, Ram Pai,
Raphael Moreira Zinsly, Ravi Bangoria, Sam Bobroff, Sandipan Das, Segher
Boessenkool, Stephen Rothwell, Sukadev Bhattiprolu, Tyrel Datwyler,
Wolfram Sang, Xiongfeng Wang.
* tag 'powerpc-5.8-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (299 commits)
powerpc/pseries: Make vio and ibmebus initcalls pseries specific
cxl: Remove dead Kconfig options
powerpc: Add POWER10 architected mode
powerpc/dt_cpu_ftrs: Add MMA feature
powerpc/dt_cpu_ftrs: Enable Prefixed Instructions
powerpc/dt_cpu_ftrs: Advertise support for ISA v3.1 if selected
powerpc: Add support for ISA v3.1
powerpc: Add new HWCAP bits
powerpc/64s: Don't set FSCR bits in INIT_THREAD
powerpc/64s: Save FSCR to init_task.thread.fscr after feature init
powerpc/64s: Don't let DT CPU features set FSCR_DSCR
powerpc/64s: Don't init FSCR_DSCR in __init_FSCR()
powerpc/32s: Fix another build failure with CONFIG_PPC_KUAP_DEBUG
powerpc/module_64: Use special stub for _mcount() with -mprofile-kernel
powerpc/module_64: Simplify check for -mprofile-kernel ftrace relocations
powerpc/module_64: Consolidate ftrace code
powerpc/32: Disable KASAN with pages bigger than 16k
powerpc/uaccess: Don't set KUEP by default on book3s/32
powerpc/uaccess: Don't set KUAP by default on book3s/32
powerpc/8xx: Reduce time spent in allow_user_access() and friends
...
Writing the AMR register is documented to require context
synchronizing operations before and after, for it to take effect as
expected. The KUAP restore at interrupt exit time deliberately avoids
the isync after the AMR update because it only needs to take effect
after the context synchronizing RFID that soon follows. Add a comment
for this.
The missing isync before the update doesn't have an obvious
justification, and seems it could theoretically allow a rogue user
access to leak past the AMR update. Add isyncs for these.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200429065654.1677541-3-npiggin@gmail.com
The idea behind this prefetch was to kick off a page table walk before
returning from the fault, getting some pipelining advantage.
But this never showed up any noticable performance advantage, and in
fact with KUAP the prefetches are actually blocked and cause some
kind of micro-architectural fault. Removing this improves page fault
microbenchmark performance by about 9%.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Keep the early return in update_mmu_cache()]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200504122907.49304-1-npiggin@gmail.com
Booting a power9 server with hash MMU could trigger an undefined
behaviour because pud_offset(p4d, 0) will do,
0 >> (PAGE_SHIFT:16 + PTE_INDEX_SIZE:8 + H_PMD_INDEX_SIZE:10)
Fix it by converting pud_index() and friends to static inline
functions.
UBSAN: shift-out-of-bounds in arch/powerpc/mm/ptdump/ptdump.c:282:15
shift exponent 34 is too large for 32-bit type 'int'
CPU: 6 PID: 1 Comm: swapper/0 Not tainted 5.6.0-rc4-next-20200303+ #13
Call Trace:
dump_stack+0xf4/0x164 (unreliable)
ubsan_epilogue+0x18/0x78
__ubsan_handle_shift_out_of_bounds+0x160/0x21c
walk_pagetables+0x2cc/0x700
walk_pud at arch/powerpc/mm/ptdump/ptdump.c:282
(inlined by) walk_pagetables at arch/powerpc/mm/ptdump/ptdump.c:311
ptdump_check_wx+0x8c/0xf0
mark_rodata_ro+0x48/0x80
kernel_init+0x74/0x194
ret_from_kernel_thread+0x5c/0x74
Suggested-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Qian Cai <cai@lca.pw>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Reviewed-by: Christophe Leroy <christophe.leroy@c-s.fr>
Link: https://lore.kernel.org/r/20200306044852.3236-1-cai@lca.pw
Christian reports:
MODPOST vmlinux.o
WARNING: modpost: vmlinux.o(.text.unlikely+0x1a0): Section mismatch in
reference from the function .early_init_mmu() to the function
.init.text:.radix__early_init_mmu()
The function .early_init_mmu() references
the function __init .radix__early_init_mmu().
This is often because .early_init_mmu lacks a __init
annotation or the annotation of .radix__early_init_mmu is wrong.
WARNING: modpost: vmlinux.o(.text.unlikely+0x1ac): Section mismatch in
reference from the function .early_init_mmu() to the function
.init.text:.hash__early_init_mmu()
The function .early_init_mmu() references
the function __init .hash__early_init_mmu().
This is often because .early_init_mmu lacks a __init
annotation or the annotation of .hash__early_init_mmu is wrong.
The compiler is uninlining early_init_mmu and not putting it in an init
section because there is no annotation. Add it.
Reported-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Link: https://lore.kernel.org/r/20200429070247.1678172-1-npiggin@gmail.com
This merges the lockless page table walk rework series from Aneesh.
Because it touches powerpc KVM code we are sharing it with the kvm-ppc
tree in our topic/ppc-kvm branch.
This is the cover letter from Aneesh:
Avoid IPI while updating page table entries.
Problem Summary:
Slow termination of KVM guest with large guest RAM config due to a
large number of IPIs that were caused by clearing level 1 PTE
entries (THP) entries. This is shown in the stack trace below.
- qemu-system-ppc [kernel.vmlinux] [k] smp_call_function_many
- smp_call_function_many
- 36.09% smp_call_function_many
serialize_against_pte_lookup
radix__pmdp_huge_get_and_clear
zap_huge_pmd
unmap_page_range
unmap_vmas
unmap_region
__do_munmap
__vm_munmap
sys_munmap
system_call
__munmap
qemu_ram_munmap
qemu_anon_ram_free
reclaim_ramblock
call_rcu_thread
qemu_thread_start
start_thread
__clone
Why we need to do IPI when clearing PMD entries:
This was added as part of commit: 13bd817bb8 ("powerpc/thp: Serialize pmd clear against a linux page table walk")
serialize_against_pte_lookup makes sure that all parallel lockless
page table walk completes before we convert a PMD pte entry to regular
pmd entry. We end up doing that conversion in the below scenarios
1) __split_huge_zero_page_pmd
2) do_huge_pmd_wp_page_fallback
3) MADV_DONTNEED running parallel to page faults.
local_irq_disable and lockless page table walk:
The lockless page table walk work with the assumption that we can
dereference the page table contents without holding a lock. For this
to work, we need to make sure we read the page table contents
atomically and page table pages are not going to be freed/released
while we are walking the table pages. We can achieve by using a rcu
based freeing for page table pages or if the architecture implements
broadcast tlbie, we can block the IPI as we walk the page table pages.
To support both the above framework, lockless page table walk is done
with irq disabled instead of rcu_read_lock()
We do have two interface for lockless page table walk, gup fast and
__find_linux_pte. This patch series makes __find_linux_pte table walk
safe against the conversion of PMD PTE to regular PMD.
gup fast:
gup fast is already safe against THP split because kernel now
differentiate between a pmd split and a compound page split. gup fast
can run parallel to a pmd split and we prevent a parallel gup fast to
a hugepage split, by freezing the page refcount and failing the
speculative page ref increment.
Similar to how gup is safe against parallel pmd split, this patch
series updates the __find_linux_pte callers to be safe against a
parallel pmd split. We do that by enforcing the following rules.
1) Don't reload the pte value, because that can be updated in
parallel.
2) Code should be able to work with a stale PTE value and not the
recent one. ie, the pte value that we are looking at may not be the
latest value in the page table.
3) Before looking at pte value check for _PAGE_PTE bit. We now do this
as part of pte_present() check.
Performance:
This speeds up Qemu guest RAM del/unplug time as below
128 core, 496GB guest:
Without patch:
munmap start: timer = 13162 ms, PID=7684
munmap finish: timer = 95312 ms, PID=7684 - delta = 82150 ms
With patch (upto removing IPI)
munmap start: timer = 196449 ms, PID=6681
munmap finish: timer = 196488 ms, PID=6681 - delta = 39ms
With patch (with adding the tlb invalidate in pmdp_huge_get_and_clear_full)
munmap start: timer = 196345 ms, PID=6879
munmap finish: timer = 196714 ms, PID=6879 - delta = 369ms
Link: https://lore.kernel.org/r/20200505071729.54912-1-aneesh.kumar@linux.ibm.com
MADV_DONTNEED holds mmap_sem in read mode and that implies a
parallel page fault is possible and the kernel can end up with a level 1 PTE
entry (THP entry) converted to a level 0 PTE entry without flushing
the THP TLB entry.
Most architectures including POWER have issues with kernel instantiating a level
0 PTE entry while holding level 1 TLB entries.
The code sequence I am looking at is
down_read(mmap_sem) down_read(mmap_sem)
zap_pmd_range()
zap_huge_pmd()
pmd lock held
pmd_cleared
table details added to mmu_gather
pmd_unlock()
insert a level 0 PTE entry()
tlb_finish_mmu().
Fix this by forcing a tlb flush before releasing pmd lock if this is
not a fullmm invalidate. We can safely skip this invalidate for
task exit case (fullmm invalidate) because in that case we are sure
there can be no parallel fault handlers.
This do change the Qemu guest RAM del/unplug time as below
128 core, 496GB guest:
Without patch:
munmap start: timer = 196449 ms, PID=6681
munmap finish: timer = 196488 ms, PID=6681 - delta = 39ms
With patch:
munmap start: timer = 196345 ms, PID=6879
munmap finish: timer = 196714 ms, PID=6879 - delta = 369ms
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200505071729.54912-23-aneesh.kumar@linux.ibm.com
This makes the pte_present check stricter by checking for additional _PAGE_PTE
bit. A level 1 pte pointer (THP pte) can be switched to a pointer to level 0 pte
page table page by following two operations.
1) THP split.
2) madvise(MADV_DONTNEED) in parallel to page fault.
A lockless page table walk need to make sure we can handle such changes
gracefully.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200505071729.54912-4-aneesh.kumar@linux.ibm.com
set_thread_uses_vas() sets used_vas flag for a process that opened VAS
window and issue CP_ABORT during context switch for only that process.
In multi-thread application, windows can be shared. For example Thread
A can open a window and Thread B can run COPY/PASTE instructions to
send NX request which may cause corruption or snooping or a covert
channel Also once this flag is set, continue to run CP_ABORT even the
VAS window is closed.
So define vas-windows counter in process mm_context, increment this
counter for each window open and decrement it for window close. If
vas-windows is set, issue CP_ABORT during context switch. It means
clear the foreign real address mapping only if the process / thread
uses COPY/PASTE. Then disable it for that process if windows are not
open.
Moved set_thread_uses_vas() code to vas_tx_win_open() as this
functionality is needed only for userspace open windows. We are adding
VAS userspace support along with this fix. So no need to include this
fix in stable releases.
Fixes: 9d2a4d7133 ("powerpc: Define set_thread_uses_vas()")
Signed-off-by: Haren Myneni <haren@linux.ibm.com>
Reported-by: Nicholas Piggin <npiggin@gmail.com>
Suggested-by: Milton Miller <miltonm@us.ibm.com>
Suggested-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1587017291.2275.1077.camel@hbabu-laptop
Implement the bulk of interrupt return logic in C. The asm return code
must handle a few cases: restoring full GPRs, and emulating stack
store.
The stack store emulation is significantly simplfied, rather than
creating a new return frame and switching to that before performing
the store, it uses the PACA to keep a scratch register around to
perform the store.
The asm return code is moved into 64e for now. The new logic has made
allowance for 64e, but I don't have a full environment that works well
to test it, and even booting in emulated qemu is not great for stress
testing. 64e shouldn't be too far off working with this, given a bit
more testing and auditing of the logic.
This is slightly faster on a POWER9 (page fault speed increases about
1.1%), probably due to reduced mtmsrd.
mpe: Includes fixes from Nick for _TIF_EMULATE_STACK_STORE
handling (including the fast_interrupt_return path), to remove
trace_hardirqs_on(), and fixes the interrupt-return part of the
MSR_VSX restore bug caught by tm-unavailable selftest.
mpe: Incorporate fix from Nick:
The return-to-kernel path has to replay any soft-pending interrupts if
it is returning to a context that had interrupts soft-enabled. It has
to do this carefully and avoid plain enabling interrupts if this is an
irq context, which can cause multiple nesting of interrupts on the
stack, and other unexpected issues.
The code which avoided this case got the soft-mask state wrong, and
marked interrupts as enabled before going around again to retry. This
seems to be mostly harmless except when PREEMPT=y, this calls
preempt_schedule_irq with irqs apparently enabled and runs into a BUG
in kernel/sched/core.c
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200225173541.1549955-29-npiggin@gmail.com
System call entry and particularly exit code is beyond the limit of
what is reasonable to implement in asm.
This conversion moves all conditional branches out of the asm code,
except for the case that all GPRs should be restored at exit.
Null syscall test is about 5% faster after this patch, because the
exit work is handled under local_irq_disable, and the hard mask and
pending interrupt replay is handled after that, which avoids games
with MSR.
mpe: Includes subsequent fixes from Nick:
This fixes 4 issues caught by TM selftests. First was a tm-syscall bug
that hit due to tabort_syscall being called after interrupts were
reconciled (in a subsequent patch), which led to interrupts being
enabled before tabort_syscall was called. Rather than going through an
un-reconciling interrupts for the return, I just go back to putting
the test early in asm, the C-ification of that wasn't a big win
anyway.
Second is the syscall return _TIF_USER_WORK_MASK check would go into
an infinite loop if _TIF_RESTORE_TM became set. The asm code uses
_TIF_USER_WORK_MASK to brach to slowpath which includes
restore_tm_state.
Third is system call return was not calling restore_tm_state, I missed
this completely (alhtough it's in the return from interrupt C
conversion because when the asm syscall code encountered problems it
would branch to the interrupt return code.
Fourth is MSR_VEC missing from restore_math, which was caught by
tm-unavailable selftest taking an unexpected facility unavailable
interrupt when testing VSX unavailble exception with MSR.FP=1
MSR.VEC=1. Fourth case also has a fixup in a subsequent patch.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200225173541.1549955-26-npiggin@gmail.com
