- 03 Dec, 2020 2 commits
-
-
Athira Rajeev authored
On systems without any specific PMU driver support registered, running 'perf record' with —intr-regs will crash ( perf record -I <workload> ). The relevant portion from crash logs and Call Trace: Unable to handle kernel paging request for data at address 0x00000068 Faulting instruction address: 0xc00000000013eb18 Oops: Kernel access of bad area, sig: 11 [#1] CPU: 2 PID: 13435 Comm: kill Kdump: loaded Not tainted 4.18.0-193.el8.ppc64le #1 NIP: c00000000013eb18 LR: c000000000139f2c CTR: c000000000393d80 REGS: c0000004a07ab4f0 TRAP: 0300 Not tainted (4.18.0-193.el8.ppc64le) NIP [c00000000013eb18] is_sier_available+0x18/0x30 LR [c000000000139f2c] perf_reg_value+0x6c/0xb0 Call Trace: [c0000004a07ab770] [c0000004a07ab7c8] 0xc0000004a07ab7c8 (unreliable) [c0000004a07ab7a0] [c0000000003aa77c] perf_output_sample+0x60c/0xac0 [c0000004a07ab840] [c0000000003ab3f0] perf_event_output_forward+0x70/0xb0 [c0000004a07ab8c0] [c00000000039e208] __perf_event_overflow+0x88/0x1a0 [c0000004a07ab910] [c00000000039e42c] perf_swevent_hrtimer+0x10c/0x1d0 [c0000004a07abc50] [c000000000228b9c] __hrtimer_run_queues+0x17c/0x480 [c0000004a07abcf0] [c00000000022aaf4] hrtimer_interrupt+0x144/0x520 [c0000004a07abdd0] [c00000000002a864] timer_interrupt+0x104/0x2f0 [c0000004a07abe30] [c0000000000091c4] decrementer_common+0x114/0x120 When perf record session is started with "-I" option, capturing registers on each sample calls is_sier_available() to check for the SIER (Sample Instruction Event Register) availability in the platform. This function in core-book3s accesses 'ppmu->flags'. If a platform specific PMU driver is not registered, ppmu is set to NULL and accessing its members results in a crash. Fix the crash by returning false in is_sier_available() if ppmu is not set. Fixes: 333804dc ("powerpc/perf: Update perf_regs structure to include SIER") Reported-by:
Sachin Sant <sachinp@linux.vnet.ibm.com> Signed-off-by:
Athira Rajeev <atrajeev@linux.vnet.ibm.com> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/1606185640-1720-1-git-send-email-atrajeev@linux.vnet.ibm.com
-
Alan Modra authored
Use bcl 20,31,0f rather than plain bl to avoid unbalancing the link stack. Update the code to use REL16 relocs, available for ppc64 in 2009 (and ppc32 in 2005). Signed-off-by:
Alan Modra <amodra@gmail.com> [mpe: Incorporate more detail into the change log] Signed-off-by:
-
- 26 Nov, 2020 7 commits
-
-
Bill Wendling authored
The clang toolchain treats inline assembly a bit differently than straight assembly code. In particular, inline assembly doesn't have the complete context available to resolve expressions. This is intentional to avoid divergence in the resulting assembly code. We can work around this issue by borrowing a workaround done for ARM, i.e. not directly testing the labels themselves, but by moving the current output pointer by a value that should always be zero. If this value is not null, then we will trigger a backward move, which is explicitly forbidden. Signed-off-by:
Bill Wendling <morbo@google.com> [mpe: Put it in a macro and only do the workaround for clang] Signed-off-by:
-
Bill Wendling authored
The gcc compiler may not be available if CC is clang. Signed-off-by:
-
Bill Wendling authored
The "-z notext" flag disables reporting an error if DT_TEXTREL is set. ld.lld: error: can't create dynamic relocation R_PPC64_ADDR64 against symbol: _start in readonly segment; recompile object files with -fPIC or pass '-Wl,-z,notext' to allow text relocations in the output >>> defined in >>> referenced by crt0.o:(.text+0x8) in archive arch/powerpc/boot/wrapper.a The BFD linker disables this by default (though it's configurable in current versions). LLD enables this by default. So we add the flag to keep LLD from emitting the error. Signed-off-by: -
Bill Wendling authored
Normally all read-only sections precede SHF_WRITE sections. .dynamic and .got have the SHF_WRITE flag; .dynamic probably because of DT_DEBUG. LLD emits an error when this happens, so use "-z rodynamic" to mark .dynamic as read-only. Signed-off-by:
-
Bill Wendling authored
Both .dynamic and .got are RELRO sections and should be placed together, and LLD emits an error: ld.lld: error: section: .got is not contiguous with other relro sections Place them together to avoid this. Signed-off-by:
-
Oleg Nesterov authored
The commit a8a4b03a ("powerpc: Hard wire PT_SOFTE value to 1 in ptrace & signals") changed ptrace_get_reg(PT_SOFTE) to report 0x1, but PTRACE_GETREGS still copies pt_regs->softe as is. This is not consistent and this breaks the user-regs-peekpoke test from https://sourceware.org/systemtap/wiki/utrace/tests/Reported-by:
Jan Kratochvil <jan.kratochvil@redhat.com> Signed-off-by:
Oleg Nesterov <oleg@redhat.com> Signed-off-by:
-
Oleg Nesterov authored
gpr_get() does membuf_write() twice to override pt_regs->msr in between. We can call membuf_write() once and change ->msr in the kernel buffer, this simplifies the code and the next fix. The patch adds a new simple helper, membuf_at(offs), it returns the new membuf which can be safely used after membuf_write(). Signed-off-by:
-
- 25 Nov, 2020 1 commit
-
-
Michael Ellerman authored
Merge our fixes branch, in particular to bring in the changes for the entry/uaccess flush.
-
- 23 Nov, 2020 2 commits
-
-
Stephen Rothwell authored
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: 9a32a7e7 ("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 authored
From Daniel's cover letter: 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 series flushes the L1 cache on kernel entry (patch 2) and after the kernel performs any user accesses (patch 3). It also adds a self-test and performs some related cleanups.
-
- 19 Nov, 2020 28 commits
-
-
Daniel Axtens authored
pseries|pnv_setup_rfi_flush already does the count cache flush setup, and we just added entry and uaccess flushes. So the name is not very accurate any more. In both platforms we then also immediately setup the STF flush. Rename them to _setup_security_mitigations and fold the STF flush in. Signed-off-by:
Daniel Axtens <dja@axtens.net> Signed-off-by:
-
Daniel Axtens authored
For simplicity in backporting, the original entry_flush test contained a lot of duplicated code from the rfi_flush test. De-duplicate that code. Signed-off-by:
-
Daniel Axtens authored
Add a test modelled on the RFI flush test which counts the number of L1D misses doing a simple syscall with the entry flush on and off. For simplicity of backporting, this test duplicates a lot of code from rfi_flush. We clean that up in the next patch. Signed-off-by:
-
Michael Ellerman authored
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: -
Nicholas Piggin authored
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:
-
Nicholas Piggin authored
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 on kernel entry. This is part of the fix for CVE-2020-4788. Signed-off-by:
-
Russell Currey authored
We are about to add an entry flush. The rfi (exit) flush test measures the number of L1D flushes over a syscall with the RFI flush enabled and disabled. But if the entry flush is also enabled, the effect of enabling and disabling the RFI flush is masked. If there is a debugfs entry for the entry flush, disable it during the RFI flush and restore it later. Reported-by:
Spoorthy S <spoorts2@in.ibm.com> Signed-off-by:
Russell Currey <ruscur@russell.cc> Signed-off-by:
-
David Hildenbrand authored
Let's use alloc_contig_pages() for allocating memory and remove the linear mapping manually via arch_remove_linear_mapping(). Mark all pages PG_offline, such that they will definitely not get touched - e.g., when hibernating. When freeing memory, try to revert what we did. The original idea was discussed in: https://lkml.kernel.org/r/48340e96-7e6b-736f-9e23-d3111b915b6e@redhat.com This is similar to CONFIG_DEBUG_PAGEALLOC handling on other architectures, whereby only single pages are unmapped from the linear mapping. Let's mimic what memory hot(un)plug would do with the linear mapping. We now need MEMORY_HOTPLUG and CONTIG_ALLOC as dependencies. Add a TODO that we want to use __GFP_ZERO for clearing once alloc_contig_pages() understands that. Tested with in QEMU/TCG with 10 GiB of main memory: [root@localhost ~]# echo 0x40000000 > /sys/kernel/debug/powerpc/memtrace/enable [ 105.903043][ T1080] memtrace: Allocated trace memory on node 0 at 0x0000000080000000 [root@localhost ~]# echo 0x40000000 > /sys/kernel/debug/powerpc/memtrace/enable [ 145.042493][ T1080] radix-mmu: Mapped 0x0000000080000000-0x00000000c0000000 with 64.0 KiB pages [ 145.049019][ T1080] memtrace: Freed trace memory back on node 0 [ 145.333960][ T1080] memtrace: Allocated trace memory on node 0 at 0x0000000080000000 [root@localhost ~]# echo 0x80000000 > /sys/kernel/debug/powerpc/memtrace/enable [ 213.606916][ T1080] radix-mmu: Mapped 0x0000000080000000-0x00000000c0000000 with 64.0 KiB pages [ 213.613855][ T1080] memtrace: Freed trace memory back on node 0 [ 214.185094][ T1080] memtrace: Allocated trace memory on node 0 at 0x0000000080000000 [root@localhost ~]# echo 0x100000000 > /sys/kernel/debug/powerpc/memtrace/enable [ 234.874872][ T1080] radix-mmu: Mapped 0x0000000080000000-0x0000000100000000 with 64.0 KiB pages [ 234.886974][ T1080] memtrace: Freed trace memory back on node 0 [ 234.890153][ T1080] memtrace: Failed to allocate trace memory on node 0 [root@localhost ~]# echo 0x40000000 > /sys/kernel/debug/powerpc/memtrace/enable [ 259.490196][ T1080] memtrace: Allocated trace memory on node 0 at 0x0000000080000000 I also made sure allocated memory is properly zeroed. Note 1: We currently won't be allocating from ZONE_MOVABLE - because our pages are not movable. However, as we don't run with any memory hot(un)plug mechanism around, we could make an exception to increase the chance of allocations succeeding. Note 2: PG_reserved isn't sufficient. E.g., kernel_page_present() used along PG_reserved in hibernation code will always return "true" on powerpc, resulting in the pages getting touched. It's too generic - e.g., indicates boot allocations. Note 3: For now, we keep using memory_block_size_bytes() as minimum granularity. Suggested-by:
Michal Hocko <mhocko@kernel.org> Signed-off-by:
David Hildenbrand <david@redhat.com> Reviewed-by:
Oscar Salvador <osalvador@suse.de> Signed-off-by:
-
David Hildenbrand authored
Let's revert what we did in case something goes wrong and we return an error - as already done on arm64 and s390x. Signed-off-by:
-
David Hildenbrand authored
The single caller (arch_remove_linear_mapping()) prints a proper warning when this function fails. No need to eventually crash the kernel - let's drop this WARN_ON. Suggested-by:
-
David Hildenbrand authored
Let's print a warning similar to in arch_add_linear_mapping() instead of WARN_ON_ONCE() and eventually crashing the kernel. Signed-off-by:
-
David Hildenbrand authored
This code currently relies on mem_hotplug_begin()/mem_hotplug_done() - create_section_mapping()/remove_section_mapping() implementations cannot tollerate getting called concurrently. Let's prepare for callers (memtrace) not holding any such locks (and don't force them to mess with memory hotplug locks). Other parts in these functions don't seem to rely on external locking. Signed-off-by:
-
David Hildenbrand authored
We want to stop abusing memory hotplug infrastructure in memtrace code to perform allocations and remove the linear mapping. Instead we will use alloc_contig_pages() and remove the linear mapping manually. Let's factor out creating/removing the linear mapping into arch_create_linear_mapping() / arch_remove_linear_mapping() - so in the future, we might be able to have whole arch_add_memory() / arch_remove_memory() be implemented in common code. Signed-off-by:
-
David Hildenbrand authored
It's very easy to crash the kernel right now by simply trying to enable memtrace concurrently, hammering on the "enable" interface loop.sh: #!/bin/bash dmesg --console-off while true; do echo 0x40000000 > /sys/kernel/debug/powerpc/memtrace/enable done [root@localhost ~]# loop.sh & [root@localhost ~]# loop.sh & Resulting quickly in a kernel crash. Let's properly protect using a mutex. Fixes: 9d5171a8 ("powerpc/powernv: Enable removal of memory for in memory tracing") Cc: stable@vger.kernel.org# v4.14+ Signed-off-by: -
David Hildenbrand authored
We currently leak kernel memory to user space, because memory offlining doesn't do any implicit clearing of memory and we are missing explicit clearing of memory. Let's keep it simple and clear pages before removing the linear mapping. Reproduced in QEMU/TCG with 10 GiB of main memory: [root@localhost ~]# dd obs=9G if=/dev/urandom of=/dev/null [... wait until "free -m" used counter no longer changes and cancel] 19665802+0 records in 1+0 records out 9663676416 bytes (9.7 GB, 9.0 GiB) copied, 135.548 s, 71.3 MB/s [root@localhost ~]# cat /sys/devices/system/memory/block_size_bytes 40000000 [root@localhost ~]# echo 0x40000000 > /sys/kernel/debug/powerpc/memtrace/enable [ 402.978663][ T1086] page:000000001bc4bc74 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x24900 [ 402.980063][ T1086] flags: 0x7ffff000001000(reserved) [ 402.980415][ T1086] raw: 007ffff000001000 c00c000000924008 c00c000000924008 0000000000000000 [ 402.980627][ T1086] raw: 0000000000000000 0000000000000000 00000001ffffffff 0000000000000000 [ 402.980845][ T1086] page dumped because: unmovable page [ 402.989608][ T1086] Offlined Pages 16384 [ 403.324155][ T1086] memtrace: Allocated trace memory on node 0 at 0x0000000200000000 Before this patch: [root@localhost ~]# hexdump -C /sys/kernel/debug/powerpc/memtrace/00000000/trace | head 00000000 c8 25 72 51 4d 26 36 c5 5c c2 56 15 d5 1a cd 10 |.%rQM&6.\.V.....| 00000010 19 b9 50 b2 cb e3 60 b8 ec 0a f3 ec 4b 3c 39 f0 |..P...`.....K<9.|$ 00000020 4e 5a 4c cf bd 26 19 ff 37 79 13 67 24 b7 b8 57 |NZL..&..7y.g$..W|$ 00000030 98 3e f5 be 6f 14 6a bd a4 52 bc 6e e9 e0 c1 5d |.>..o.j..R.n...]|$ 00000040 76 b3 ae b5 88 d7 da e3 64 23 85 2c 10 88 07 b6 |v.......d#.,....|$ 00000050 9a d8 91 de f7 50 27 69 2e 64 9c 6f d3 19 45 79 |.....P'i.d.o..Ey|$ 00000060 6a 6f 8a 61 71 19 1f c7 f1 df 28 26 ca 0f 84 55 |jo.aq.....(&...U|$ 00000070 01 3f be e4 e2 e1 da ff 7b 8c 8e 32 37 b4 24 53 |.?......{..27.$S|$ 00000080 1b 70 30 45 56 e6 8c c4 0e b5 4c fb 9f dd 88 06 |.p0EV.....L.....|$ 00000090 ef c4 18 79 f1 60 b1 5c 79 59 4d f4 36 d7 4a 5c |...y.`.\yYM.6.J\|$ After this patch: [root@localhost ~]# hexdump -C /sys/kernel/debug/powerpc/memtrace/00000000/trace | head 00000000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| * 40000000 Fixes: 9d5171a8 ("powerpc/powernv: Enable removal of memory for in memory tracing") Cc: stable@vger.kernel.org # v4.14+ Reported-by: -
Madhavan Srinivasan authored
In power10 DD1, there is an issue where the SIAR (Sampled Instruction Address Register) is not latching to the sampled address during random sampling. This results in value of 0s in the SIAR. Add a check to use regs->nip when SIAR is zero. Signed-off-by:
Madhavan Srinivasan <maddy@linux.ibm.com> Signed-off-by:
-
Athira Rajeev authored
While setting the processor mode for any sample, perf_get_misc_flags() expects the privilege level to differentiate the userspace and kernel address. On power10 DD1, there is an issue that causes MSR_HV MSR_PR bits of Sampled Instruction Event Register (SIER) not to be set for marked events. Hence add a check to use the address in SIAR (Sampled Instruction Address Register) to identify the privilege level. Signed-off-by:
-
Athira Rajeev authored
In power10 DD1, there is an issue that causes the SIAR_VALID bit of the SIER (Sampled Instruction Event Register) to not be set. But the SIAR_VALID bit is used for fetching the instruction address from the SIAR (Sampled Instruction Address Register), and marked events are sampled only if the SIAR_VALID bit is set. So drop the check for SIAR_VALID and return true always incase of power10 DD1. Signed-off-by:
-
Athira Rajeev authored
Add a new power PMU flag "PPMU_P10_DD1" which can be used to conditionally add any code path for power10 DD1 processor version. Also modify power10 PMU driver code to set this flag only for DD1, based on the Processor Version Register (PVR) value. Signed-off-by:
-
Kaixu Xia authored
Fixes coccicheck warning: ./arch/powerpc/mm/pgtable_32.c:87:11-12: WARNING comparing pointer to 0 Avoid pointer type value compared to 0. Reported-by:
Tosk Robot <tencent_os_robot@tencent.com> Signed-off-by:
Kaixu Xia <kaixuxia@tencent.com> Reviewed-by:
Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by:
-
Christophe Leroy authored
RFI macro is just there to add an infinite loop past rfi in order to avoid prefetch on 40x in half a dozen of places in entry_32 and head_32. Those places are already full of #ifdefs, so just add a few more to explicitely show those loops and remove RFI. Signed-off-by:
-
Christophe Leroy authored
For book3s/32 and for booke, RFI is just an rfi. Only 40x has a non trivial RFI. CONFIG_PPC_RTAS is never selected by 40x platforms. Make it more explicit by replacing RFI by rfi wherever possible. Signed-off-by:
-
Christophe Leroy authored
In head_64.S, we have two places using RFI to return to kernel. Use RFI_TO_KERNEL instead. They are the two only places using RFI on book3s/64, so the RFI macro can go away. Signed-off-by:
-
Kaixu Xia authored
Fix coccicheck warning: arch/powerpc/platforms/powernv/pci-sriov.c:443:7-10: WARNING: Unsigned expression compared with zero: win < 0 arch/powerpc/platforms/powernv/pci-sriov.c:462:7-10: WARNING: Unsigned expression compared with zero: win < 0 Fixes: 39efc03e ("powerpc/powernv/sriov: Move M64 BAR allocation into a helper") Reported-by:
Andrew Donnellan <ajd@linux.ibm.com> Signed-off-by:
-
Zhang Xiaoxu authored
This reverts commit a0ff72f9. Since the commit b015f6bc ("powerpc/pseries: Add cpu DLPAR support for drc-info property"), the 'cpu_drcs' wouldn't be double freed when the 'cpus' node not found. So we needn't apply this patch, otherwise, the memory will be leaked. Fixes: a0ff72f9 ("powerpc/pseries/hotplug-cpu: Remove double free in error path") Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
Zhang Xiaoxu <zhangxiaoxu5@huawei.com> [mpe: Caused by me applying a patch to a function that had changed in the interim] Signed-off-by:
-
Nicholas Piggin authored
read_user_stack_slow that walks user address translation by hand is only required on hash, because a hash fault can not be serviced from "NMI" context (to avoid re-entering the hash code) so the user stack can be mapped into Linux page tables but not accessible by the CPU. Radix MMU mode does not have this restriction. A page fault failure would indicate the page is not accessible via get_user_pages either, so avoid this on radix. Signed-off-by:
-
Youling Tang authored
Use the common INIT_DATA_SECTION rule for the linker script in an effort to regularize the linker script. Signed-off-by:
Youling Tang <tangyouling@loongson.cn> Signed-off-by:
-
Christophe Leroy authored
On 8xx, we get the following features: [ 0.000000] cpu_features = 0x0000000000000100 [ 0.000000] possible = 0x0000000000000120 [ 0.000000] always = 0x0000000000000000 This is not correct. As CONFIG_PPC_8xx is mutually exclusive with all other configurations, the three lines should be equal. The problem is due to CPU_FTRS_GENERIC_32 which is taken when CONFIG_BOOK3S_32 is NOT selected. This CPU_FTRS_GENERIC_32 is pointless because there is no generic configuration supporting all 32 bits but book3s/32. Remove this pointless generic features definition to unbreak the calculation of 'possible' features and 'always' features. Fixes: 76bc080e ("[POWERPC] Make default cputable entries reflect selected CPU family") Signed-off-by:
-
