- 30 Oct, 2012 10 commits
-
-
Paul Mackerras authored
Currently the code that accounts stolen time tends to overestimate the stolen time, and will sometimes report more stolen time in a DTL (dispatch trace log) entry than has elapsed since the last DTL entry. This can cause guests to underflow the user or system time measured for some tasks, leading to ridiculous CPU percentages and total runtimes being reported by top and other utilities. In addition, the current code was designed for the previous policy where a vcore would only run when all the vcpus in it were runnable, and so only counted stolen time on a per-vcore basis. Now that a vcore can run while some of the vcpus in it are doing other things in the kernel (e.g. handling a page fault), we need to count the time when a vcpu task is preempted while it is not running as part of a vcore as stolen also. To do this, we bring back the BUSY_IN_HOST vcpu state and extend the vcpu_load/put functions to count preemption time while the vcpu is in that state. Handling the transitions between the RUNNING and BUSY_IN_HOST states requires checking and updating two variables (accumulated time stolen and time last preempted), so we add a new spinlock, vcpu->arch.tbacct_lock. This protects both the per-vcpu stolen/preempt-time variables, and the per-vcore variables while this vcpu is running the vcore. Finally, we now don't count time spent in userspace as stolen time. The task could be executing in userspace on behalf of the vcpu, or it could be preempted, or the vcpu could be genuinely stopped. Since we have no way of dividing up the time between these cases, we don't count any of it as stolen. Signed-off-by:
Paul Mackerras <paulus@samba.org> Signed-off-by:
Alexander Graf <agraf@suse.de>
-
Paul Mackerras authored
Currently the Book3S HV code implements a policy on multi-threaded processors (i.e. POWER7) that requires all of the active vcpus in a virtual core to be ready to run before we run the virtual core. However, that causes problems on reset, because reset stops all vcpus except vcpu 0, and can also reduce throughput since all four threads in a virtual core have to wait whenever any one of them hits a hypervisor page fault. This relaxes the policy, allowing the virtual core to run as soon as any vcpu in it is runnable. With this, the KVMPPC_VCPU_STOPPED state and the KVMPPC_VCPU_BUSY_IN_HOST state have been combined into a single KVMPPC_VCPU_NOTREADY state, since we no longer need to distinguish between them. Signed-off-by:
-
Paul Mackerras authored
If a thread in a virtual core becomes runnable while other threads in the same virtual core are already running in the guest, it is possible for the latecomer to join the others on the core without first pulling them all out of the guest. Currently this only happens rarely, when a vcpu is first started. This fixes some bugs and omissions in the code in this case. First, we need to check for VPA updates for the latecomer and make a DTL entry for it. Secondly, if it comes along while the master vcpu is doing a VPA update, we don't need to do anything since the master will pick it up in kvmppc_run_core. To handle this correctly we introduce a new vcore state, VCORE_STARTING. Thirdly, there is a race because we currently clear the hardware thread's hwthread_req before waiting to see it get to nap. A latecomer thread could have its hwthread_req cleared before it gets to test it, and therefore never increment the nap_count, leading to messages about wait_for_nap timeouts. Signed-off-by:
-
Paul Mackerras authored
There were a few places where we were traversing the list of runnable threads in a virtual core, i.e. vc->runnable_threads, without holding the vcore spinlock. This extends the places where we hold the vcore spinlock to cover everywhere that we traverse that list. Since we possibly need to sleep inside kvmppc_book3s_hv_page_fault, this moves the call of it from kvmppc_handle_exit out to kvmppc_vcpu_run, where we don't hold the vcore lock. In kvmppc_vcore_blocked, we don't actually need to check whether all vcpus are ceded and don't have any pending exceptions, since the caller has already done that. The caller (kvmppc_run_vcpu) wasn't actually checking for pending exceptions, so we add that. The change of if to while in kvmppc_run_vcpu is to make sure that we never call kvmppc_remove_runnable() when the vcore state is RUNNING or EXITING. Signed-off-by:
-
Paul Mackerras authored
Subsequent patches implementing in-kernel XICS emulation will make it possible for IPIs to arrive at secondary threads at arbitrary times. This fixes some races in how we start the secondary threads, which if not fixed could lead to occasional crashes of the host kernel. This makes sure that (a) we have grabbed all the secondary threads, and verified that they are no longer in the kernel, before we start any thread, (b) that the secondary thread loads its vcpu pointer after clearing the IPI that woke it up (so we don't miss a wakeup), and (c) that the secondary thread clears its vcpu pointer before incrementing the nap count. It also removes unnecessary setting of the vcpu and vcore pointers in the paca in kvmppc_core_vcpu_load. Signed-off-by:
-
Paul Mackerras authored
When a Book3S HV KVM guest is running, we need the host to be in single-thread mode, that is, all of the cores (or at least all of the cores where the KVM guest could run) to be running only one active hardware thread. This is because of the hardware restriction in POWER processors that all of the hardware threads in the core must be in the same logical partition. Complying with this restriction is much easier if, from the host kernel's point of view, only one hardware thread is active. This adds two hooks in the SMP hotplug code to allow the KVM code to make sure that secondary threads (i.e. hardware threads other than thread 0) cannot come online while any KVM guest exists. The KVM code still has to check that any core where it runs a guest has the secondary threads offline, but having done that check it can now be sure that they will not come online while the guest is running. Signed-off-by:
Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by:
-
Alexander Graf authored
The new uapi framework splits kernel internal and user space exported bits of header files more cleanly. Adjust the ePAPR header accordingly. Signed-off-by: -
Alexander Graf authored
The mtspr/mfspr emulation code became quite big over time. Move it into its own function so things stay more readable. Signed-off-by: -
Alexander Graf authored
All user space offloaded instruction emulation needs to reenter kvm to produce consistent state again. Fix the section in the documentation to mention all of them. Signed-off-by: -
Alexander Graf authored
When remembering the direction of a DCR transaction, we should write to the same variable that we interpret on later when doing vcpu_run again. Signed-off-by:
-
- 23 Oct, 2012 1 commit
-
-
Christoffer Dall authored
The mmu_notifier_retry is not specific to any vcpu (and never will be) so only take struct kvm as a parameter. The motivation is the ARM mmu code that needs to call this from somewhere where we long let go of the vcpu pointer. Signed-off-by:
Christoffer Dall <c.dall@virtualopensystems.com> Signed-off-by:
Avi Kivity <avi@redhat.com>
-
- 22 Oct, 2012 1 commit
-
-
Borislav Petkov authored
Use __func__ instead of the function name in svm_hardware_enable since those things tend to get out of sync. This also slims down printk line length in conjunction with using pr_err. No functionality change. Cc: Joerg Roedel <joro@8bytes.org> Cc: Avi Kivity <avi@redhat.com> Signed-off-by:
Borislav Petkov <bp@alien8.de> Signed-off-by:
-
- 18 Oct, 2012 3 commits
-
-
Xiao Guangrong authored
The #PF with PFEC.RSV = 1 indicates that the guest is accessing MMIO, we can not fix it if it is caused by delivery event. Reporting internal error for this case Signed-off-by:
Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com> Signed-off-by:
-
Xiao Guangrong authored
VM exits during Event Delivery is really unexpected if it is not caused by Exceptions/EPT-VIOLATION/TASK_SWITCH, we'd better to report an internal and freeze the guest, the VMM has the chance to check the guest Signed-off-by:
-
Gleb Natapov authored
Signed-off-by:
Gleb Natapov <gleb@redhat.com> Signed-off-by:
-
- 17 Oct, 2012 4 commits
-
-
Xiao Guangrong authored
The only difference between FNAME(update_pte) and FNAME(pte_prefetch) is that the former is allowed to prefetch gfn from dirty logged slot, so introduce a common function to prefetch spte Signed-off-by:
-
Xiao Guangrong authored
The function does not depend on guest mmu mode, move it out from paging_tmpl.h Signed-off-by:
-
Xiao Guangrong authored
Let it return emulate state instead of spte like __direct_map Signed-off-by:
-
Xiao Guangrong authored
Remove mmu_is_invalid and use is_invalid_pfn instead Signed-off-by:
-
- 10 Oct, 2012 5 commits
-
-
http://github.com/agraf/linux-2.6Marcelo Tosatti authored
* 'for-upstream' of http://github.com/agraf/linux-2.6: (56 commits) arch/powerpc/kvm/e500_tlb.c: fix error return code KVM: PPC: Book3S HV: Provide a way for userspace to get/set per-vCPU areas KVM: PPC: Book3S: Get/set guest FP regs using the GET/SET_ONE_REG interface KVM: PPC: Book3S: Get/set guest SPRs using the GET/SET_ONE_REG interface KVM: PPC: set IN_GUEST_MODE before checking requests KVM: PPC: e500: MMU API: fix leak of shared_tlb_pages KVM: PPC: e500: fix allocation size error on g2h_tlb1_map KVM: PPC: Book3S HV: Fix calculation of guest phys address for MMIO emulation KVM: PPC: Book3S HV: Remove bogus update of physical thread IDs KVM: PPC: Book3S HV: Fix updates of vcpu->cpu KVM: Move some PPC ioctl definitions to the correct place KVM: PPC: Book3S HV: Handle memory slot deletion and modification correctly KVM: PPC: Move kvm->arch.slot_phys into memslot.arch KVM: PPC: Book3S HV: Take the SRCU read lock before looking up memslots KVM: PPC: bookehv: Allow duplicate calls of DO_KVM macro KVM: PPC: BookE: Support FPU on non-hv systems KVM: PPC: 440: Implement mfdcrx KVM: PPC: 440: Implement mtdcrx Document IACx/DACx registers access using ONE_REG API KVM: PPC: E500: Remove E500_TLB_DIRTY flag ...
-
Christian Borntraeger authored
Newer kernels (linux-next with the transparent huge page patches) use rrbm if the feature is announced via feature bit 66. RRBM will cause intercepts, so KVM does not handle it right now, causing an illegal instruction in the guest. The easy solution is to disable the feature bit for the guest. This fixes bugs like: Kernel BUG at 0000000000124c2a [verbose debug info unavailable] illegal operation: 0001 [#1] SMP Modules linked in: virtio_balloon virtio_net ipv6 autofs4 CPU: 0 Not tainted 3.5.4 #1 Process fmempig (pid: 659, task: 000000007b712fd0, ksp: 000000007bed3670) Krnl PSW : 0704d00180000000 0000000000124c2a (pmdp_clear_flush_young+0x5e/0x80) R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:1 PM:0 EA:3 00000000003cc000 0000000000000004 0000000000000000 0000000079800000 0000000000040000 0000000000000000 000000007bed3918 000000007cf40000 0000000000000001 000003fff7f00000 000003d281a94000 000000007bed383c 000000007bed3918 00000000005ecbf8 00000000002314a6 000000007bed36e0 Krnl Code:>0000000000124c2a: b9810025 ogr %r2,%r5 0000000000124c2e: 41343000 la %r3,0(%r4,%r3) 0000000000124c32: a716fffa brct %r1,124c26 0000000000124c36: b9010022 lngr %r2,%r2 0000000000124c3a: e3d0f0800004 lg %r13,128(%r15) 0000000000124c40: eb22003f000c srlg %r2,%r2,63 [ 2150.713198] Call Trace: [ 2150.713223] ([<00000000002312c4>] page_referenced_one+0x6c/0x27c) [ 2150.713749] [<0000000000233812>] page_referenced+0x32a/0x410 [...] CC: stable@vger.kernel.org CC: Alex Graf <agraf@suse.de> Signed-off-by:Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by:
Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by:
Marcelo Tosatti <mtosatti@redhat.com>
-
Cornelia Huck authored
Signed-off-by:
Cornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by:
-
Jason J. Herne authored
EXTERNAL_CALL and EMERGENCY type interrupts need to preserve their interrupt code parameter when being injected from user space. Signed-off-by:
Jason J. Herne <jjherne@us.ibm.com> Reviewed-by:
-
Shuah Khan authored
Change existing kernel error message to include return value from iommu_attach_device() when it fails. This will help debug device assignment failures more effectively. Signed-off-by:
Shuah Khan <shuah.khan@hp.com> Signed-off-by:
-
- 08 Oct, 2012 2 commits
-
-
Jan Kiszka authored
No users outside of kvm/x86.c. Signed-off-by:
Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by:
-
Jan Kiszka authored
There are no external callers of this function as there is no concept of resetting a vcpu from generic code. Signed-off-by:
-
- 05 Oct, 2012 14 commits
-
-
Julia Lawall authored
Convert a 0 error return code to a negative one, as returned elsewhere in the function. A new label is also added to avoid freeing things that are known to not yet be allocated. A simplified version of the semantic match that finds the first problem is as follows: (http://coccinelle.lip6.fr/) // <smpl> @@ identifier ret; expression e,e1,e2,e3,e4,x; @@ ( if (\(ret != 0\|ret < 0\) || ...) { ... return ...; } | ret = 0 ) ... when != ret = e1 *x = \(kmalloc\|kzalloc\|kcalloc\|devm_kzalloc\|ioremap\|ioremap_nocache\|devm_ioremap\|devm_ioremap_nocache\)(...); ... when != x = e2 when != ret = e3 *if (x == NULL || ...) { ... when != ret = e4 * return ret; } // </smpl> Signed-off-by:
Julia Lawall <julia@diku.dk> Signed-off-by:
-
Paul Mackerras authored
The PAPR paravirtualization interface lets guests register three different types of per-vCPU buffer areas in its memory for communication with the hypervisor. These are called virtual processor areas (VPAs). Currently the hypercalls to register and unregister VPAs are handled by KVM in the kernel, and userspace has no way to know about or save and restore these registrations across a migration. This adds "register" codes for these three areas that userspace can use with the KVM_GET/SET_ONE_REG ioctls to see what addresses have been registered, and to register or unregister them. This will be needed for guest hibernation and migration, and is also needed so that userspace can unregister them on reset (otherwise we corrupt guest memory after reboot by writing to the VPAs registered by the previous kernel). The "register" for the VPA is a 64-bit value containing the address, since the length of the VPA is fixed. The "registers" for the SLB shadow buffer and dispatch trace log (DTL) are 128 bits long, consisting of the guest physical address in the high (first) 64 bits and the length in the low 64 bits. This also fixes a bug where we were calling init_vpa unconditionally, leading to an oops when unregistering the VPA. Signed-off-by:
-
Paul Mackerras authored
This enables userspace to get and set all the guest floating-point state using the KVM_[GS]ET_ONE_REG ioctls. The floating-point state includes all of the traditional floating-point registers and the FPSCR (floating point status/control register), all the VMX/Altivec vector registers and the VSCR (vector status/control register), and on POWER7, the vector-scalar registers (note that each FP register is the high-order half of the corresponding VSR). Most of these are implemented in common Book 3S code, except for VSX on POWER7. Because HV and PR differ in how they store the FP and VSX registers on POWER7, the code for these cases is not common. On POWER7, the FP registers are the upper halves of the VSX registers vsr0 - vsr31. PR KVM stores vsr0 - vsr31 in two halves, with the upper halves in the arch.fpr[] array and the lower halves in the arch.vsr[] array, whereas HV KVM on POWER7 stores the whole VSX register in arch.vsr[]. Signed-off-by:
-
Paul Mackerras authored
This enables userspace to get and set various SPRs (special-purpose registers) using the KVM_[GS]ET_ONE_REG ioctls. With this, userspace can get and set all the SPRs that are part of the guest state, either through the KVM_[GS]ET_REGS ioctls, the KVM_[GS]ET_SREGS ioctls, or the KVM_[GS]ET_ONE_REG ioctls. The SPRs that are added here are: - DABR: Data address breakpoint register - DSCR: Data stream control register - PURR: Processor utilization of resources register - SPURR: Scaled PURR - DAR: Data address register - DSISR: Data storage interrupt status register - AMR: Authority mask register - UAMOR: User authority mask override register - MMCR0, MMCR1, MMCRA: Performance monitor unit control registers - PMC1..PMC8: Performance monitor unit counter registers In order to reduce code duplication between PR and HV KVM code, this moves the kvm_vcpu_ioctl_[gs]et_one_reg functions into book3s.c and centralizes the copying between user and kernel space there. The registers that are handled differently between PR and HV, and those that exist only in one flavor, are handled in kvmppc_[gs]et_one_reg() functions that are specific to each flavor. Signed-off-by:
-
Scott Wood authored
Avoid a race as described in the code comment. Also remove a related smp_wmb() from booke's kvmppc_prepare_to_enter(). I can't see any reason for it, and the book3s_pr version doesn't have it. Signed-off-by:
Scott Wood <scottwood@freescale.com> Signed-off-by:
-
Scott Wood authored
This was found by kmemleak. Signed-off-by:
-
Scott Wood authored
We were only allocating half the bytes we need, which was made more obvious by a recent fix to the memset in clear_tlb1_bitmap(). Signed-off-by:
-
Paul Mackerras authored
In the case where the host kernel is using a 64kB base page size and the guest uses a 4k HPTE (hashed page table entry) to map an emulated MMIO device, we were calculating the guest physical address wrongly. We were calculating a gfn as the guest physical address shifted right 16 bits (PAGE_SHIFT) but then only adding back in 12 bits from the effective address, since the HPTE had a 4k page size. Thus the gpa reported to userspace was missing 4 bits. Instead, we now compute the guest physical address from the HPTE without reference to the host page size, and then compute the gfn by shifting the gpa right PAGE_SHIFT bits. Reported-by:
Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by:
-
Paul Mackerras authored
When making a vcpu non-runnable we incorrectly changed the thread IDs of all other threads on the core, just remove that code. Signed-off-by:
-
Paul Mackerras authored
This removes the powerpc "generic" updates of vcpu->cpu in load and put, and moves them to the various backends. The reason is that "HV" KVM does its own sauce with that field and the generic updates might corrupt it. The field contains the CPU# of the -first- HW CPU of the core always for all the VCPU threads of a core (the one that's online from a host Linux perspective). However, the preempt notifiers are going to be called on the threads VCPUs when they are running (due to them sleeping on our private waitqueue) causing unload to be called, potentially clobbering the value. Signed-off-by:
-
Paul Mackerras authored
This moves the definitions of KVM_CREATE_SPAPR_TCE and KVM_ALLOCATE_RMA in include/linux/kvm.h from the section listing the vcpu ioctls to the section listing VM ioctls, as these are both implemented and documented as VM ioctls. Fortunately there is no actual collision of ioctl numbers at this point. Moving these to the correct section will reduce the probability of a future collision. This does not change the user/kernel ABI at all. Signed-off-by:
-
Paul Mackerras authored
This adds an implementation of kvm_arch_flush_shadow_memslot for Book3S HV, and arranges for kvmppc_core_commit_memory_region to flush the dirty log when modifying an existing slot. With this, we can handle deletion and modification of memory slots. kvm_arch_flush_shadow_memslot calls kvmppc_core_flush_memslot, which on Book3S HV now traverses the reverse map chains to remove any HPT (hashed page table) entries referring to pages in the memslot. This gets called by generic code whenever deleting a memslot or changing the guest physical address for a memslot. We flush the dirty log in kvmppc_core_commit_memory_region for consistency with what x86 does. We only need to flush when an existing memslot is being modified, because for a new memslot the rmap array (which stores the dirty bits) is all zero, meaning that every page is considered clean already, and when deleting a memslot we obviously don't care about the dirty bits any more. Signed-off-by:
-
Paul Mackerras authored
Now that we have an architecture-specific field in the kvm_memory_slot structure, we can use it to store the array of page physical addresses that we need for Book3S HV KVM on PPC970 processors. This reduces the size of struct kvm_arch for Book3S HV, and also reduces the size of struct kvm_arch_memory_slot for other PPC KVM variants since the fields in it are now only compiled in for Book3S HV. This necessitates making the kvm_arch_create_memslot and kvm_arch_free_memslot operations specific to each PPC KVM variant. That in turn means that we now don't allocate the rmap arrays on Book3S PR and Book E. Since we now unpin pages and free the slot_phys array in kvmppc_core_free_memslot, we no longer need to do it in kvmppc_core_destroy_vm, since the generic code takes care to free all the memslots when destroying a VM. We now need the new memslot to be passed in to kvmppc_core_prepare_memory_region, since we need to initialize its arch.slot_phys member on Book3S HV. Signed-off-by:
-
Paul Mackerras authored
The generic KVM code uses SRCU (sleeping RCU) to protect accesses to the memslots data structures against updates due to userspace adding, modifying or removing memory slots. We need to do that too, both to avoid accessing stale copies of the memslots and to avoid lockdep warnings. This therefore adds srcu_read_lock/unlock pairs around code that accesses and uses memslots. Since the real-mode handlers for H_ENTER, H_REMOVE and H_BULK_REMOVE need to access the memslots, and we don't want to call the SRCU code in real mode (since we have no assurance that it would only access the linear mapping), we hold the SRCU read lock for the VM while in the guest. This does mean that adding or removing memory slots while some vcpus are executing in the guest will block for up to two jiffies. This tradeoff is acceptable since adding/removing memory slots only happens rarely, while H_ENTER/H_REMOVE/H_BULK_REMOVE are performance-critical hot paths. Signed-off-by:
-
