- 30 Apr, 2019 4 commits
-
-
Michael Ellerman authored
Merge our topic branch shared with KVM. In particular this includes the rewrite of the idle code into C.
-
Michael Ellerman authored
This is an implementation of commits 53a712ba ("powerpc/powernv/idle: Restore AMR/UAMOR/AMOR after idle") and a3f3072d ("powerpc/powernv/idle: Restore IAMR after idle") using the new C-based idle code. Signed-off-by:
Nicholas Piggin <npiggin@gmail.com> [mpe: Extract from Nick's patch] Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au>
-
Nicholas Piggin authored
Reimplement Book3S idle code in C, moving POWER7/8/9 implementation speific HV idle code to the powernv platform code. Book3S assembly stubs are kept in common code and used only to save the stack frame and non-volatile GPRs before executing architected idle instructions, and restoring the stack and reloading GPRs then returning to C after waking from idle. The complex logic dealing with threads and subcores, locking, SPRs, HMIs, timebase resync, etc., is all done in C which makes it more maintainable. This is not a strict translation to C code, there are some significant differences: - Idle wakeup no longer uses the ->cpu_restore call to reinit SPRs, but saves and restores them itself. - The optimisation where EC=ESL=0 idle modes did not have to save GPRs or change MSR is restored, because it's now simple to do. ESL=1 sleeps that do not lose GPRs can use this optimization too. - KVM secondary entry and cede is now more of a call/return style rather than branchy. nap_state_lost is not required because KVM always returns via NVGPR restoring path. - KVM secondary wakeup from offline sequence is moved entirely into the offline wakeup, which avoids a hwsync in the normal idle wakeup path. Performance measured with context switch ping-pong on different threads or cores, is possibly improved a small amount, 1-3% depending on stop state and core vs thread test for shallow states. Deep states it's in the noise compared with other latencies. KVM improvements: - Idle sleepers now always return to caller rather than branch out to KVM first. - This allows optimisations like very fast return to caller when no state has been lost. - KVM no longer requires nap_state_lost because it controls NVGPR save/restore itself on the way in and out. - The heavy idle wakeup KVM request check can be moved out of the normal host idle code and into the not-performance-critical offline code. - KVM nap code now returns from where it is called, which makes the flow a bit easier to follow. Reviewed-by:
Gautham R. Shenoy <ego@linux.vnet.ibm.com> Signed-off-by:
-
Nicholas Piggin authored
Using a jiffies timer creates a dependency on the tick_do_timer_cpu incrementing jiffies. If that CPU has locked up and jiffies is not incrementing, the watchdog heartbeat timer for all CPUs stops and creates false positives and confusing warnings on local CPUs, and also causes the SMP detector to stop, so the root cause is never detected. Fix this by using hrtimer based timers for the watchdog heartbeat, like the generic kernel hardlockup detector. Cc: Gautham R. Shenoy <ego@linux.vnet.ibm.com> Reported-by:
Ravikumar Bangoria <ravi.bangoria@in.ibm.com> Signed-off-by:
Ravi Bangoria <ravi.bangoria@linux.ibm.com> Reported-by:
-
- 29 Apr, 2019 1 commit
-
-
Nathan Fontenot authored
When removing memory we need to remove the memory from the node it was added to instead of looking up the node it should be in in the device tree. During testing we have seen scenarios where the affinity for a LMB changes due to a partition migration or PRRN event. In these cases the node the LMB exists in may not match the node the device tree indicates it belongs in. This can lead to a system crash when trying to DLPAR remove the LMB after a migration or PRRN event. The current code looks up the node in the device tree to remove the LMB from, the crash occurs when we try to offline this node and it does not have any data, i.e. node_data[nid] == NULL. 36:mon> e cpu 0x36: Vector: 300 (Data Access) at [c0000001828b7810] pc: c00000000036d08c: try_offline_node+0x2c/0x1b0 lr: c0000000003a14ec: remove_memory+0xbc/0x110 sp: c0000001828b7a90 msr: 800000000280b033 dar: 9a28 dsisr: 40000000 current = 0xc0000006329c4c80 paca = 0xc000000007a55200 softe: 0 irq_happened: 0x01 pid = 76926, comm = kworker/u320:3 36:mon> t [link register ] c0000000003a14ec remove_memory+0xbc/0x110 [c0000001828b7a90] c00000000006a1cc arch_remove_memory+0x9c/0xd0 (unreliable) [c0000001828b7ad0] c0000000003a14e0 remove_memory+0xb0/0x110 [c0000001828b7b20] c0000000000c7db4 dlpar_remove_lmb+0x94/0x160 [c0000001828b7b60] c0000000000c8ef8 dlpar_memory+0x7e8/0xd10 [c0000001828b7bf0] c0000000000bf828 handle_dlpar_errorlog+0xf8/0x160 [c0000001828b7c60] c0000000000bf8cc pseries_hp_work_fn+0x3c/0xa0 [c0000001828b7c90] c000000000128cd8 process_one_work+0x298/0x5a0 [c0000001828b7d20] c000000000129068 worker_thread+0x88/0x620 [c0000001828b7dc0] c00000000013223c kthread+0x1ac/0x1c0 [c0000001828b7e30] c00000000000b45c ret_from_kernel_thread+0x5c/0x80 To resolve this we need to track the node a LMB belongs to when it is added to the system so we can remove it from that node instead of the node that the device tree indicates it should belong to. Signed-off-by:Nathan Fontenot <nfont@linux.vnet.ibm.com> Signed-off-by:
-
- 28 Apr, 2019 1 commit
-
-
Colin Ian King authored
There are several identical spelling mistakes in warning messages, fix these. Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
- 21 Apr, 2019 32 commits
-
-
Aneesh Kumar K.V authored
This patch fix the below section mismatch warnings. WARNING: vmlinux.o(.text+0x2d1f44): Section mismatch in reference from the function devm_memremap_pages_release() to the function .meminit.text:arch_remove_memory() WARNING: vmlinux.o(.text+0x2d265c): Section mismatch in reference from the function devm_memremap_pages() to the function .meminit.text:arch_add_memory() Signed-off-by:
Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Signed-off-by:
-
Aneesh Kumar K.V authored
The region actually point to linear map. Rename the #define to clarify thati. Signed-off-by:
-
Aneesh Kumar K.V authored
This helps in debugging. We can look at the dmesg to find out different kernel mapping details. On 4K config this shows kernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000 On 64K config: kernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000 Signed-off-by:
-
Aneesh Kumar K.V authored
This reduces multiple comparisons in get_region_id to a bit shift operation. Signed-off-by:
-
Aneesh Kumar K.V authored
All the regions are now mapped with top nibble 0xc. Hence the region id check is not needed for virt_addr_valid() Signed-off-by:
-
Aneesh Kumar K.V authored
This adds an explicit check in various functions. Signed-off-by:
-
Aneesh Kumar K.V authored
This patch maps vmalloc, IO and vmemap regions in the 0xc address range instead of the current 0xd and 0xf range. This brings the mapping closer to radix translation mode. With hash 64K page size each of this region is 512TB whereas with 4K config we are limited by the max page table range of 64TB and hence there regions are of 16TB size. The kernel mapping is now: On 4K hash kernel_region_map_size = 16TB kernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000 64K hash, 64K radix and 4k radix: kernel_region_map_size = 512TB kernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000 Signed-off-by: -
Aneesh Kumar K.V authored
This makes it easy to update the region mapping in the later patch Signed-off-by:
-
Aneesh Kumar K.V authored
Allocate subpage protect related variables only if we use the feature. This helps in reducing the hash related mm context struct by around 4K Before the patch sizeof(struct hash_mm_context) = 8288 After the patch sizeof(struct hash_mm_context) = 4160 Signed-off-by:
-
Aneesh Kumar K.V authored
Currently, our mm_context_t on book3s64 include all hash specific context details like slice mask and subpage protection details. We can skip allocating these with radix translation. This will help us to save 8K per mm_context with radix translation. With the patch applied we have sizeof(mm_context_t) = 136 sizeof(struct hash_mm_context) = 8288 Signed-off-by:
-
Aneesh Kumar K.V authored
Avoid #ifdef in generic code. Also enables us to do this specific to MMU translation mode on book3s64 Signed-off-by:
-
Aneesh Kumar K.V authored
We want to switch to allocating them runtime only when hash translation is enabled. Add helpers so that both book3s and nohash can be adapted to upcoming change easily. Signed-off-by:
-
Aneesh Kumar K.V authored
Book3s64 always have PPC_MM_SLICES enabled. So remove the unncessary #ifdef Signed-off-by:
-
Aneesh Kumar K.V authored
The current value of MAX_PHYSMEM_BITS cannot work with 32 bit configs. We used to have MAX_PHYSMEM_BITS not defined without SPARSEMEM and 32 bit configs never expected a value to be set for MAX_PHYSMEM_BITS. Dependent code such as zsmalloc derived the right values based on other fields. Instead of finding a value that works with different configs, use new values only for book3s_64. For 64 bit booke, use the definition of MAX_PHYSMEM_BITS as per commit a7df61a0 ("[PATCH] ppc64: Increase sparsemem defaults") That change was done in 2005 and hopefully will work with book3e 64. Fixes: 8bc08689 ("powerpc/mm: Only define MAX_PHYSMEM_BITS in SPARSEMEM configurations") Signed-off-by:
-
Christophe Leroy authored
This patch implements Kernel Userspace Access Protection for book3s/32. Due to limitations of the processor page protection capabilities, the protection is only against writing. read protection cannot be achieved using page protection. The previous patch modifies the page protection so that RW user pages are RW for Key 0 and RO for Key 1, and it sets Key 0 for both user and kernel. This patch changes userspace segment registers are set to Ku 0 and Ks 1. When kernel needs to write to RW pages, the associated segment register is then changed to Ks 0 in order to allow write access to the kernel. In order to avoid having the read all segment registers when locking/unlocking the access, some data is kept in the thread_struct and saved on stack on exceptions. The field identifies both the first unlocked segment and the first segment following the last unlocked one. When no segment is unlocked, it contains value 0. As the hash_page() function is not able to easily determine if a protfault is due to a bad kernel access to userspace, protfaults need to be handled by handle_page_fault when KUAP is set. Signed-off-by:
Christophe Leroy <christophe.leroy@c-s.fr> [mpe: Drop allow_read/write_to/from_user() as they're now in kup.h, and adapt allow_user_access() to do nothing when to == NULL] Signed-off-by:
-
Christophe Leroy authored
This patch prepares Kernel Userspace Access Protection for book3s/32. Due to limitations of the processor page protection capabilities, the protection is only against writing. read protection cannot be achieved using page protection. book3s/32 provides the following values for PP bits: PP00 provides RW for Key 0 and NA for Key 1 PP01 provides RW for Key 0 and RO for Key 1 PP10 provides RW for all PP11 provides RO for all Today PP10 is used for RW pages and PP11 for RO pages, and user segment register's Kp and Ks are set to 1. This patch modifies page protection to use PP01 for RW pages and sets user segment registers to Kp 0 and Ks 0. This will allow to setup Userspace write access protection by settng Ks to 1 in the following patch. Kernel space segment registers remain unchanged. Signed-off-by:
-
Christophe Leroy authored
To implement Kernel Userspace Execution Prevention, this patch sets NX bit on all user segments on kernel entry and clears NX bit on all user segments on kernel exit. Note that powerpc 601 doesn't have the NX bit, so KUEP will not work on it. A warning is displayed at startup. Signed-off-by:
-
Christophe Leroy authored
This patch adds Kernel Userspace Access Protection on the 8xx. When a page is RO or RW, it is set RO or RW for Key 0 and NA for Key 1. Up to now, the User group is defined with Key 0 for both User and Supervisor. By changing the group to Key 0 for User and Key 1 for Supervisor, this patch prevents the Kernel from being able to access user data. At exception entry, the kernel saves SPRN_MD_AP in the regs struct, and reapply the protection. At exception exit it restores SPRN_MD_AP with the value saved on exception entry. Signed-off-by:
-
Christophe Leroy authored
This patch adds Kernel Userspace Execution Prevention on the 8xx. When a page is Executable, it is set Executable for Key 0 and NX for Key 1. Up to now, the User group is defined with Key 0 for both User and Supervisor. By changing the group to Key 0 for User and Key 1 for Supervisor, this patch prevents the Kernel from being able to execute user code. Signed-off-by:
-
Christophe Leroy authored
Since the 8xx implements hardware page table walk assistance, the PGD entries always point to a 4k aligned page, so the 2 upper bits of the APG are not clobbered anymore and remain 0. Therefore only APG0 and APG1 are used and need a definition. We set the other APG to the lowest permission level. Signed-off-by:
-
Christophe Leroy authored
This patch adds ASM macros for saving, restoring and checking the KUAP state, and modifies setup_32 to call them on exceptions from kernel. The macros are defined as empty by default for when CONFIG_PPC_KUAP is not selected and/or for platforms which don't handle (yet) KUAP. Signed-off-by:
-
Christophe Leroy authored
syscalls are from user only, so we can account time without checking whether returning to kernel or user as it will only be user. Signed-off-by:
-
Michael Ellerman authored
When KUAP is enabled we have logic to detect page faults that occur outside of a valid user access region and are blocked by the AMR. What we don't have at the moment is logic to detect a fault *within* a valid user access region, that has been incorrectly blocked by AMR. This is not meant to ever happen, but it can if we incorrectly save/restore the AMR, or if the AMR was overwritten for some other reason. Currently if that happens we assume it's just a regular fault that will be corrected by handling the fault normally, so we just return. But there is nothing the fault handling code can do to fix it, so the fault just happens again and we spin forever, leading to soft lockups. So add some logic to detect that case and WARN() if we ever see it. Arguably it should be a BUG(), but it's more polite to fail the access and let the kernel continue, rather than taking down the box. There should be no data integrity issue with failing the fault rather than BUG'ing, as we're just going to disallow an access that should have been allowed. To make the code a little easier to follow, unroll the condition at the end of bad_kernel_fault() and comment each case, before adding the call to bad_kuap_fault(). Signed-off-by: -
Michael Ellerman authored
Kernel Userspace Access Prevention utilises a feature of the Radix MMU which disallows read and write access to userspace addresses. By utilising this, the kernel is prevented from accessing user data from outside of trusted paths that perform proper safety checks, such as copy_{to/from}_user() and friends. Userspace access is disabled from early boot and is only enabled when performing an operation like copy_{to/from}_user(). The register that controls this (AMR) does not prevent userspace from accessing itself, so there is no need to save and restore when entering and exiting userspace. When entering the kernel from the kernel we save AMR and if it is not blocking user access (because eg. we faulted doing a user access) we reblock user access for the duration of the exception (ie. the page fault) and then restore the AMR when returning back to the kernel. This feature can be tested by using the lkdtm driver (CONFIG_LKDTM=y) and performing the following: # (echo ACCESS_USERSPACE) > [debugfs]/provoke-crash/DIRECT If enabled, this should send SIGSEGV to the thread. We also add paranoid checking of AMR in switch and syscall return under CONFIG_PPC_KUAP_DEBUG. Co-authored-by:Russell Currey <ruscur@russell.cc> Signed-off-by:
-
Russell Currey authored
__patch_instruction() is called in early boot, and uses __put_user_size(), which includes the allow/prevent calls to enforce KUAP, which could either be called too early, or in the Radix case, forced to use "early_" versions of functions just to safely handle this one case. __put_user_asm() does not do this, and thus is safe to use both in early boot, and later on since in this case it should only ever be touching kernel memory. __patch_instruction() was previously refactored to use __put_user_size() in order to be able to return -EFAULT, which would allow the kernel to patch instructions in userspace, which should never happen. This has the functional change of causing faults on userspace addresses if KUAP is turned on, which should never happen in practice. A future enhancement could be to double check the patch address is definitely allowed to be tampered with by the kernel. Signed-off-by:
-
Russell Currey authored
Execution protection already exists on radix, this just refactors the radix init to provide the KUEP setup function instead. Thus, the only functional change is that it can now be disabled. Signed-off-by:
-
Russell Currey authored
Some platforms (i.e. Radix MMU) need per-CPU initialisation for KUP. Any platforms that only want to do KUP initialisation once globally can just check to see if they're running on the boot CPU, or check if whatever setup they need has already been performed. Note that this is only for 64-bit. Signed-off-by:
-
Christophe Leroy authored
This patch implements a framework for Kernel Userspace Access Protection. Then subarches will have the possibility to provide their own implementation by providing setup_kuap() and allow/prevent_user_access(). Some platforms will need to know the area accessed and whether it is accessed from read, write or both. Therefore source, destination and size and handed over to the two functions. mpe: Rename to allow/prevent rather than unlock/lock, and add read/write wrappers. Drop the 32-bit code for now until we have an implementation for it. Add kuap to pt_regs for 64-bit as well as 32-bit. Don't split strings, use pr_crit_ratelimited(). Signed-off-by:
-
Christophe Leroy authored
This patch adds a skeleton for Kernel Userspace Execution Prevention. Then subarches implementing it have to define CONFIG_PPC_HAVE_KUEP and provide setup_kuep() function. Signed-off-by:
-
Christophe Leroy authored
This patch adds a skeleton for Kernel Userspace Protection functionnalities like Kernel Userspace Access Protection and Kernel Userspace Execution Prevention The subsequent implementation of KUAP for radix makes use of a MMU feature in order to patch out assembly when KUAP is disabled or unsupported. This won't work unless there's an entry point for KUP support before the feature magic happens, so for PPC64 setup_kup() is called early in setup. On PPC32, feature_fixup() is done too early to allow the same. Suggested-by:
-
Michael Ellerman authored
In order to implement KUAP (Kernel Userspace Access Protection) on Power9 we will be using the AMR, and therefore indirectly the UAMOR/AMOR. So save/restore these regs in the idle code. Signed-off-by: -
Russell Currey authored
Without restoring the IAMR after idle, execution prevention on POWER9 with Radix MMU is overwritten and the kernel can freely execute userspace without faulting. This is necessary when returning from any stop state that modifies user state, as well as hypervisor state. To test how this fails without this patch, load the lkdtm driver and do the following: $ echo EXEC_USERSPACE > /sys/kernel/debug/provoke-crash/DIRECT which won't fault, then boot the kernel with powersave=off, where it will fault. Applying this patch will fix this. Fixes: 3b10d009 ("powerpc/mm/radix: Prevent kernel execution of user space") Cc: stable@vger.kernel.org # v4.10+ Signed-off-by:
Akshay Adiga <akshay.adiga@linux.vnet.ibm.com> Reviewed-by:
-
- 20 Apr, 2019 2 commits
-
-
Michael Neuling authored
This adds a flag so that the DAWR can be enabled on P9 via: echo Y > /sys/kernel/debug/powerpc/dawr_enable_dangerous The DAWR was previously force disabled on POWER9 in: 96541531 powerpc: Disable DAWR in the base POWER9 CPU features Also see Documentation/powerpc/DAWR-POWER9.txt This is a dangerous setting, USE AT YOUR OWN RISK. Some users may not care about a bad user crashing their box (ie. single user/desktop systems) and really want the DAWR. This allows them to force enable DAWR. This flag can also be used to disable DAWR access. Once this is cleared, all DAWR access should be cleared immediately and your machine once again safe from crashing. Userspace may get confused by toggling this. If DAWR is force enabled/disabled between getting the number of breakpoints (via PTRACE_GETHWDBGINFO) and setting the breakpoint, userspace will get an inconsistent view of what's available. Similarly for guests. For the DAWR to be enabled in a KVM guest, the DAWR needs to be force enabled in the host AND the guest. For this reason, this won't work on POWERVM as it doesn't allow the HCALL to work. Writes of 'Y' to the dawr_enable_dangerous file will fail if the hypervisor doesn't support writing the DAWR. To double check the DAWR is working, run this kernel selftest: tools/testing/selftests/powerpc/ptrace/ptrace-hwbreak.c Any errors/failures/skips mean something is wrong. Signed-off-by:
Michael Neuling <mikey@neuling.org> Signed-off-by:
-
Nathan Lynch authored
Changing the NUMA associations for CPUs and memory at runtime is basically unsupported by the core mm, scheduler etc. We see all manner of crashes, warnings and instability when the pseries code tries to do this. Disable this behavior by default, and document the switch a bit. Signed-off-by:
Nathan Lynch <nathanl@linux.ibm.com> Signed-off-by:
-
