- 04 Jan, 2018 5 commits
-
-
Thomas Gleixner authored
The recent changes for PTI touch cpu_tlbstate from various tlb_flush inlines. cpu_tlbstate is exported as GPL symbol, so this causes a regression when building out of tree drivers for certain graphics cards. Aside of that the export was wrong since it was introduced as it should have been EXPORT_PER_CPU_SYMBOL_GPL(). Use the correct PER_CPU export and drop the _GPL to restore the previous state which allows users to utilize the cards they payed for. As always I'm really thrilled to make this kind of change to support the #friends (or however the hot hashtag of today is spelled) from that closet sauce graphics corp. Fixes: 1e02ce4c ("x86: Store a per-cpu shadow copy of CR4") Fixes: 6fd166aa ("x86/mm: Use/Fix PCID to optimize user/kernel switches") Reported-by:
Kees Cook <keescook@google.com> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: stable@vger.kernel.org
-
Peter Zijlstra authored
Thomas reported the following warning: BUG: using smp_processor_id() in preemptible [00000000] code: ovsdb-server/4498 caller is native_flush_tlb_single+0x57/0xc0 native_flush_tlb_single+0x57/0xc0 __set_pte_vaddr+0x2d/0x40 set_pte_vaddr+0x2f/0x40 cea_set_pte+0x30/0x40 ds_update_cea.constprop.4+0x4d/0x70 reserve_ds_buffers+0x159/0x410 x86_reserve_hardware+0x150/0x160 x86_pmu_event_init+0x3e/0x1f0 perf_try_init_event+0x69/0x80 perf_event_alloc+0x652/0x740 SyS_perf_event_open+0x3f6/0xd60 do_syscall_64+0x5c/0x190 set_pte_vaddr is used to map the ds buffers into the cpu entry area, but there are two problems with that: 1) The resulting flush is not supposed to be called in preemptible context 2) The cpu entry area is supposed to be per CPU, but the debug store buffers are mapped for all CPUs so these mappings need to be flushed globally. Add the necessary preemption protection across the mapping code and flush TLBs globally. Fixes: c1961a46 ("x86/events/intel/ds: Map debug buffers in cpu_entry_area") Reported-by:Thomas Zeitlhofer <thomas.zeitlhofer+lkml@ze-it.at> Signed-off-by:
Peter Zijlstra <peterz@infradead.org> Signed-off-by:
-
Thomas Gleixner authored
vaddr_end for KASLR is only documented in the KASLR code itself and is adjusted depending on config options. So it's not surprising that a change of the memory layout causes KASLR to have the wrong vaddr_end. This can map arbitrary stuff into other areas causing hard to understand problems. Remove the whole ifdef magic and define the start of the cpu_entry_area to be the end of the KASLR vaddr range. Add documentation to that effect. Fixes: 92a0f81d ("x86/cpu_entry_area: Move it out of the fixmap") Reported-by:
Benjamin Gilbert <benjamin.gilbert@coreos.com> Signed-off-by:
-
Thomas Gleixner authored
There is no reason for 4 and 5 level pagetables to have a different layout. It just makes determining vaddr_end for KASLR harder than necessary. Fixes: 92a0f81d ("x86/cpu_entry_area: Move it out of the fixmap") Signed-off-by:
-
Andrey Ryabinin authored
Since f06bdd40 ("x86/mm: Adapt MODULES_END based on fixmap section size") kasan_mem_to_shadow(MODULES_END) could be not aligned to a page boundary. So passing page unaligned address to kasan_populate_zero_shadow() have two possible effects: 1) It may leave one page hole in supposed to be populated area. After commit 21506525 ("x86/kasan/64: Teach KASAN about the cpu_entry_area") that hole happens to be in the shadow covering fixmap area and leads to crash: BUG: unable to handle kernel paging request at fffffbffffe8ee04 RIP: 0010:check_memory_region+0x5c/0x190 Call Trace: <NMI> memcpy+0x1f/0x50 ghes_copy_tofrom_phys+0xab/0x180 ghes_read_estatus+0xfb/0x280 ghes_notify_nmi+0x2b2/0x410 nmi_handle+0x115/0x2c0 default_do_nmi+0x57/0x110 do_nmi+0xf8/0x150 end_repeat_nmi+0x1a/0x1e Note, the crash likely disappeared after commit 92a0f81d, which changed kasan_populate_zero_shadow() call the way it was before commit 21506525. 2) Attempt to load module near MODULES_END will fail, because __vmalloc_node_range() called from kasan_module_alloc() will hit the WARN_ON(!pte_none(*pte)) in the vmap_pte_range() and bail out with error. To fix this we need to make kasan_mem_to_shadow(MODULES_END) page aligned which means that MODULES_END should be 8*PAGE_SIZE aligned. The whole point of commit f06bdd40 was to move MODULES_END down if NR_CPUS is big, so the cpu_entry_area takes a lot of space. But since 92a0f81d ("x86/cpu_entry_area: Move it out of the fixmap") the cpu_entry_area is no longer in fixmap, so we could just set MODULES_END to a fixed 8*PAGE_SIZE aligned address. Fixes: f06bdd40 ("x86/mm: Adapt MODULES_END based on fixmap section size") Reported-by:
Jakub Kicinski <kubakici@wp.pl> Signed-off-by:
Andrey Ryabinin <aryabinin@virtuozzo.com> Signed-off-by:
-
- 03 Jan, 2018 7 commits
-
-
Nick Desaulniers authored
cpu_tss_rw is declared with DECLARE_PER_CPU_PAGE_ALIGNED but then defined with DEFINE_PER_CPU_SHARED_ALIGNED leading to section mismatch warnings. Use DEFINE_PER_CPU_PAGE_ALIGNED consistently. This is necessary because it's mapped to the cpu entry area and must be page aligned. [ tglx: Massaged changelog a bit ] Fixes: 1a935bc3 ("x86/entry: Move SYSENTER_stack to the beginning of struct tss_struct") Suggested-by:
Nick Desaulniers <ndesaulniers@google.com> Signed-off-by:
-
Thomas Gleixner authored
The preparation for PTI which added CR3 switching to the entry code misplaced the CR3 switch in entry_SYSCALL_compat(). With PTI enabled the entry code tries to access a per cpu variable after switching to kernel GS. This fails because that variable is not mapped to user space. This results in a double fault and in the worst case a kernel crash. Move the switch ahead of the access and clobber RSP which has been saved already. Fixes: 8a09317b ("x86/mm/pti: Prepare the x86/entry assembly code for entry/exit CR3 switching") Reported-by:
Lars Wendler <wendler.lars@web.de> Reported-by:
Laura Abbott <labbott@redhat.com> Signed-off-by:
-
Josh Poimboeuf authored
In the stack dump code, if the frame after the starting pt_regs is also a regs frame, the registers don't get printed. Fix that. Reported-by:
Andy Lutomirski <luto@amacapital.net> Tested-by:
Alexander Tsoy <alexander@tsoy.me> Signed-off-by:
Josh Poimboeuf <jpoimboe@redhat.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Toralf Förster <toralf.foerster@gmx.de> Cc: stable@vger.kernel.org Fixes: 3b3fa11b ("x86/dumpstack: Print any pt_regs found on the stack") Link: http://lkml.kernel.org/r/396f84491d2f0ef64eda4217a2165f5712f6a115.1514736742.git.jpoimboe@redhat.comSigned-off-by:
Ingo Molnar <mingo@kernel.org>
-
Josh Poimboeuf authored
The show_regs_safe() logic is wrong. When there's an iret stack frame, it prints the entire pt_regs -- most of which is random stack data -- instead of just the five registers at the end. show_regs_safe() is also poorly named: the on_stack() checks aren't for safety. Rename the function to show_regs_if_on_stack() and add a comment to explain why the checks are needed. These issues were introduced with the "partial register dump" feature of the following commit: b02fcf9b ("x86/unwinder: Handle stack overflows more gracefully") That patch had gone through a few iterations of development, and the above issues were artifacts from a previous iteration of the patch where 'regs' pointed directly to the iret frame rather than to the (partially empty) pt_regs. Tested-by:
-
Thomas Gleixner authored
Meelis reported that his K8 Athlon64 emits MCE warnings when PTI is enabled: [Hardware Error]: Error Addr: 0x0000ffff81e000e0 [Hardware Error]: MC1 Error: L1 TLB multimatch. [Hardware Error]: cache level: L1, tx: INSN The address is in the entry area, which is mapped into kernel _AND_ user space. That's special because we switch CR3 while we are executing there. User mapping: 0xffffffff81e00000-0xffffffff82000000 2M ro PSE GLB x pmd Kernel mapping: 0xffffffff81000000-0xffffffff82000000 16M ro PSE x pmd So the K8 is complaining that the TLB entries differ. They differ in the GLB bit. Drop the GLB bit when installing the user shared mapping. Fixes: 6dc72c3c ("x86/mm/pti: Share entry text PMD") Reported-by:
Meelis Roos <mroos@linux.ee> Signed-off-by:
-
Tom Lendacky authored
AMD processors are not subject to the types of attacks that the kernel page table isolation feature protects against. The AMD microarchitecture does not allow memory references, including speculative references, that access higher privileged data when running in a lesser privileged mode when that access would result in a page fault. Disable page table isolation by default on AMD processors by not setting the X86_BUG_CPU_INSECURE feature, which controls whether X86_FEATURE_PTI is set. Signed-off-by:
Tom Lendacky <thomas.lendacky@amd.com> Signed-off-by:
Borislav Petkov <bp@suse.de> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20171227054354.20369.94587.stgit@tlendack-t1.amdoffice.net
-
Thomas Gleixner authored
This really want's to be enabled by default. Users who know what they are doing can disable it either in the config or on the kernel command line. Signed-off-by:
-
- 31 Dec, 2017 4 commits
-
-
Thomas Gleixner authored
Andy prefers to be paranoid about the pagetable free in the error path of write_ldt(). Make it conditional and warn whenever the installment of a secondary LDT fails. Requested-by:
-
Thomas Gleixner authored
The error path in write_ldt() tries to free 'old_ldt' instead of the newly allocated 'new_ldt', resulting in a memory leak. It also misses to clean up a half populated LDT pagetable, which is not a leak as it gets cleaned up when the process exits. Free both the potentially half populated LDT pagetable and the newly allocated LDT struct. This can be done unconditionally because once an LDT is mapped subsequent maps will succeed, because the PTE page is already populated and the two LDTs fit into that single page. Reported-by:
Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by:
-
Thomas Gleixner authored
The preempt_disable/enable() pair in __native_flush_tlb() was added in commit: 5cf0791d ("x86/mm: Disable preemption during CR3 read+write") ... to protect the UP variant of flush_tlb_mm_range(). That preempt_disable/enable() pair should have been added to the UP variant of flush_tlb_mm_range() instead. The UP variant was removed with commit: ce4a4e56 ("x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code") ... but the preempt_disable/enable() pair stayed around. The latest change to __native_flush_tlb() in commit: 6fd166aa ("x86/mm: Use/Fix PCID to optimize user/kernel switches") ... added an access to a per CPU variable outside the preempt disabled regions, which makes no sense at all. __native_flush_tlb() must always be called with at least preemption disabled. Remove the preempt_disable/enable() pair and add a WARN_ON_ONCE() to catch bad callers independent of the smp_processor_id() debugging. Signed-off-by:
-
Thomas Gleixner authored
smpboot_setup_warm_reset_vector() and smpboot_restore_warm_reset_vector() invoke local_flush_tlb() for no obvious reason. Digging in history revealed that the original code in the 2.1 era added those because the code manipulated a swapper_pg_dir pagetable entry. The pagetable manipulation was removed long ago in the 2.3 timeframe, but the TLB flush invocations stayed around forever. Remove them along with the pointless pr_debug()s which come from the same 2.1 change. Reported-by:
Dominik Brodowski <linux@dominikbrodowski.net> Signed-off-by:
-
- 23 Dec, 2017 24 commits
-
-
Thomas Gleixner authored
Now that the LDT mapping is in a known area when PAGE_TABLE_ISOLATION is enabled its a primary target for attacks, if a user space interface fails to validate a write address correctly. That can never happen, right? The SDM states: If the segment descriptors in the GDT or an LDT are placed in ROM, the processor can enter an indefinite loop if software or the processor attempts to update (write to) the ROM-based segment descriptors. To prevent this problem, set the accessed bits for all segment descriptors placed in a ROM. Also, remove operating-system or executive code that attempts to modify segment descriptors located in ROM. So its a valid approach to set the ACCESS bit when setting up the LDT entry and to map the table RO. Fixup the selftest so it can handle that new mode. Remove the manual ACCESS bit setter in set_tls_desc() as this is now pointless. Folded the patch from Peter Ziljstra. Signed-off-by: -
Thomas Gleixner authored
Add two debugfs files which allow to dump the pagetable of the current task. current_kernel dumps the regular page table. This is the page table which is normally shared between kernel and user space. If kernel page table isolation is enabled this is the kernel space mapping. If kernel page table isolation is enabled the second file, current_user, dumps the user space page table. These files allow to verify the resulting page tables for page table isolation, but even in the normal case its useful to be able to inspect user space page tables of current for debugging purposes. Signed-off-by:
-
Thomas Gleixner authored
ptdump_walk_pgd_level_checkwx() checks the kernel page table for WX pages, but does not check the PAGE_TABLE_ISOLATION user space page table. Restructure the code so that dmesg output is selected by an explicit argument and not implicit via checking the pgd argument for !NULL. Add the check for the user space page table. Signed-off-by:
-
Borislav Petkov authored
The upcoming support for dumping the kernel and the user space page tables of the current process would create more random files in the top level debugfs directory. Add a page table directory and move the existing file to it. Signed-off-by:
-
Dave Hansen authored
Finally allow CONFIG_PAGE_TABLE_ISOLATION to be enabled. PARAVIRT generally requires that the kernel not manage its own page tables. It also means that the hypervisor and kernel must agree wholeheartedly about what format the page tables are in and what they contain. PAGE_TABLE_ISOLATION, unfortunately, changes the rules and they can not be used together. I've seen conflicting feedback from maintainers lately about whether they want the Kconfig magic to go first or last in a patch series. It's going last here because the partially-applied series leads to kernels that can not boot in a bunch of cases. I did a run through the entire series with CONFIG_PAGE_TABLE_ISOLATION=y to look for build errors, though. [ tglx: Removed SMP and !PARAVIRT dependencies as they not longer exist ] Signed-off-by:
Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by:
-
Vlastimil Babka authored
CONFIG_PAGE_TABLE_ISOLATION is relatively new and intrusive feature that may still have some corner cases which could take some time to manifest and be fixed. It would be useful to have Oops messages indicate whether it was enabled for building the kernel, and whether it was disabled during boot. Example of fully enabled: Oops: 0001 [#1] SMP PTI Example of enabled during build, but disabled during boot: Oops: 0001 [#1] SMP NOPTI We can decide to remove this after the feature has been tested in the field long enough. [ tglx: Made it use boot_cpu_has() as requested by Borislav ] Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Signed-off-by:
Eduardo Valentin <eduval@amazon.com> Acked-by:
Dave Hansen <dave.hansen@intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Andy Lutomirsky <luto@kernel.org> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: David Laight <David.Laight@aculab.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: Greg KH <gregkh@linuxfoundation.org> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Juergen Gross <jgross@suse.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Will Deacon <will.deacon@arm.com> Cc: aliguori@amazon.com Cc: bpetkov@suse.de Cc: daniel.gruss@iaik.tugraz.at Cc: hughd@google.com Cc: jkosina@suse.cz Cc: keescook@google.com Signed-off-by:
-
Peter Zijlstra authored
Ideally we'd also use sparse to enforce this separation so it becomes much more difficult to mess up. Signed-off-by:
-
Dave Hansen authored
This uses INVPCID to shoot down individual lines of the user mapping instead of marking the entire user map as invalid. This could/might/possibly be faster. This for sure needs tlb_single_page_flush_ceiling to be redetermined; esp. since INVPCID is _slow_. A detailed performance analysis is available here: https://lkml.kernel.org/r/3062e486-3539-8a1f-5724-16199420be71@intel.com [ Peterz: Split out from big combo patch ] Signed-off-by:
-
Peter Zijlstra authored
Most NMI/paranoid exceptions will not in fact change pagetables and would thus not require TLB flushing, however RESTORE_CR3 uses flushing CR3 writes. Restores to kernel PCIDs can be NOFLUSH, because we explicitly flush the kernel mappings and now that we track which user PCIDs need flushing we can avoid those too when possible. This does mean RESTORE_CR3 needs an additional scratch_reg, luckily both sites have plenty available. Signed-off-by:
-
Peter Zijlstra authored
We can use PCID to retain the TLBs across CR3 switches; including those now part of the user/kernel switch. This increases performance of kernel entry/exit at the cost of more expensive/complicated TLB flushing. Now that we have two address spaces, one for kernel and one for user space, we need two PCIDs per mm. We use the top PCID bit to indicate a user PCID (just like we use the PFN LSB for the PGD). Since we do TLB invalidation from kernel space, the existing code will only invalidate the kernel PCID, we augment that by marking the corresponding user PCID invalid, and upon switching back to userspace, use a flushing CR3 write for the switch. In order to access the user_pcid_flush_mask we use PER_CPU storage, which means the previously established SWAPGS vs CR3 ordering is now mandatory and required. Having to do this memory access does require additional registers, most sites have a functioning stack and we can spill one (RAX), sites without functional stack need to otherwise provide the second scratch register. Note: PCID is generally available on Intel Sandybridge and later CPUs. Note: Up until this point TLB flushing was broken in this series. Based-on-code-from: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by:
-
Dave Hansen authored
In preparation to adding additional PCID flushing, abstract the loading of a new ASID into CR3. [ PeterZ: Split out from big combo patch ] Signed-off-by:
-
Dave Hansen authored
If changing the page tables in such a way that an invalidation of all contexts (aka. PCIDs / ASIDs) is required, they can be actively invalidated by: 1. INVPCID for each PCID (works for single pages too). 2. Load CR3 with each PCID without the NOFLUSH bit set 3. Load CR3 with the NOFLUSH bit set for each and do INVLPG for each address. But, none of these are really feasible since there are ~6 ASIDs (12 with PAGE_TABLE_ISOLATION) at the time that invalidation is required. Instead of actively invalidating them, invalidate the *current* context and also mark the cpu_tlbstate _quickly_ to indicate future invalidation to be required. At the next context-switch, look for this indicator ('invalidate_other' being set) invalidate all of the cpu_tlbstate.ctxs[] entries. This ensures that any future context switches will do a full flush of the TLB, picking up the previous changes. [ tglx: Folded more fixups from Peter ] Signed-off-by: -
Andy Lutomirski authored
Make VSYSCALLs work fully in PTI mode by mapping them properly to the user space visible page tables. [ tglx: Hide unused functions (Patch by Arnd Bergmann) ] Signed-off-by:
Andy Lutomirski <luto@kernel.org> Signed-off-by:
-
Andy Lutomirski authored
With PTI enabled, the LDT must be mapped in the usermode tables somewhere. The LDT is per process, i.e. per mm. An earlier approach mapped the LDT on context switch into a fixmap area, but that's a big overhead and exhausted the fixmap space when NR_CPUS got big. Take advantage of the fact that there is an address space hole which provides a completely unused pgd. Use this pgd to manage per-mm LDT mappings. This has a down side: the LDT isn't (currently) randomized, and an attack that can write the LDT is instant root due to call gates (thanks, AMD, for leaving call gates in AMD64 but designing them wrong so they're only useful for exploits). This can be mitigated by making the LDT read-only or randomizing the mapping, either of which is strightforward on top of this patch. This will significantly slow down LDT users, but that shouldn't matter for important workloads -- the LDT is only used by DOSEMU(2), Wine, and very old libc implementations. [ tglx: Cleaned it up. ] Signed-off-by:
-
Andy Lutomirski authored
Shrink vmalloc space from 16384TiB to 12800TiB to enlarge the hole starting at 0xff90000000000000 to be a full PGD entry. A subsequent patch will use this hole for the pagetable isolation LDT alias. Signed-off-by:
-
Hugh Dickins authored
The BTS and PEBS buffers both have their virtual addresses programmed into the hardware. This means that any access to them is performed via the page tables. The times that the hardware accesses these are entirely dependent on how the performance monitoring hardware events are set up. In other words, there is no way for the kernel to tell when the hardware might access these buffers. To avoid perf crashes, place 'debug_store' allocate pages and map them into the cpu_entry_area. The PEBS fixup buffer does not need this treatment. [ tglx: Got rid of the kaiser_add_mapping() complication ] Signed-off-by:
Hugh Dickins <hughd@google.com> Signed-off-by:
-
Thomas Gleixner authored
The Intel PEBS/BTS debug store is a design trainwreck as it expects virtual addresses which must be visible in any execution context. So it is required to make these mappings visible to user space when kernel page table isolation is active. Provide enough room for the buffer mappings in the cpu_entry_area so the buffers are available in the user space visible page tables. At the point where the kernel side entry area is populated there is no buffer available yet, but the kernel PMD must be populated. To achieve this set the entries for these buffers to non present. Signed-off-by:
-
Andy Lutomirski authored
Map the ESPFIX pages into user space when PTI is enabled. Signed-off-by:
-
Thomas Gleixner authored
Share the entry text PMD of the kernel mapping with the user space mapping. If large pages are enabled this is a single PMD entry and at the point where it is copied into the user page table the RW bit has not been cleared yet. Clear it right away so the user space visible map becomes RX. Signed-off-by:
-
Thomas Gleixner authored
The (irq)entry text must be visible in the user space page tables. To allow simple PMD based sharing, make the entry text PMD aligned. Signed-off-by:
-
Andy Lutomirski authored
Share the cpu entry area so the user space and kernel space page tables have the same P4D page. Signed-off-by:
-
Thomas Gleixner authored
Force the entry through the trampoline only when PTI is active. Otherwise go through the normal entry code. Signed-off-by:
-
Andy Lutomirski authored
Provide infrastructure to: - find a kernel PMD for a mapping which must be visible to user space for the entry/exit code to work. - walk an address range and share the kernel PMD with it. This reuses a small part of the original KAISER patches to populate the user space page table. [ tglx: Made it universally usable so it can be used for any kind of shared mapping. Add a mechanism to clear specific bits in the user space visible PMD entry. Folded Andys simplifactions ] Originally-by:
-
Dave Hansen authored
In clone_pgd_range() copy the init user PGDs which cover the kernel half of the address space, so a process has all the required kernel mappings visible. [ tglx: Split out from the big kaiser dump and folded Andys simplification ] Signed-off-by:
-
