From 352ccf890a3e91f58bc32503f9afdc161bc2c34a Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 4 Sep 2025 13:39:27 +0200 Subject: [PATCH 1/2] KVM: s390: improve interrupt cpu for wakeup MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Turns out that picking an idle CPU for floating interrupts has some negative side effects. The guest will keep the IO workload on its CPU and rather use an IPI from the interrupt CPU instead of moving workload. For example a guest with 2 vCPUs and 1 fio process might run that fio on vcpu1. If after diag500 both vCPUs are idle then vcpu0 is woken up. The guest will then do an IPI from vcpu0 to vcpu1. So lets change the heuristics and prefer the last CPU that went to sleep. This one is likely still in halt polling and can be woken up quickly. This patch shows significant improvements in terms of bandwidth or cpu consumption for fio and uperf workloads and seems to be a net win. Link: https://lore.kernel.org/linux-s390/20250904113927.119306-1-borntraeger@linux.ibm.com/ Reviewed-by: Christoph Schlameuß Signed-off-by: Christian Borntraeger --- arch/s390/include/asm/kvm_host.h | 2 +- arch/s390/kvm/interrupt.c | 20 +++++++++----------- 2 files changed, 10 insertions(+), 12 deletions(-) diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h index f870d09515cc..95d15416c39d 100644 --- a/arch/s390/include/asm/kvm_host.h +++ b/arch/s390/include/asm/kvm_host.h @@ -356,7 +356,7 @@ struct kvm_s390_float_interrupt { int counters[FIRQ_MAX_COUNT]; struct kvm_s390_mchk_info mchk; struct kvm_s390_ext_info srv_signal; - int next_rr_cpu; + int last_sleep_cpu; struct mutex ais_lock; u8 simm; u8 nimm; diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c index 60c360c18690..b8e6f82e92c3 100644 --- a/arch/s390/kvm/interrupt.c +++ b/arch/s390/kvm/interrupt.c @@ -1322,6 +1322,7 @@ int kvm_s390_handle_wait(struct kvm_vcpu *vcpu) VCPU_EVENT(vcpu, 4, "enabled wait: %llu ns", sltime); no_timer: kvm_vcpu_srcu_read_unlock(vcpu); + vcpu->kvm->arch.float_int.last_sleep_cpu = vcpu->vcpu_idx; kvm_vcpu_halt(vcpu); vcpu->valid_wakeup = false; __unset_cpu_idle(vcpu); @@ -1948,18 +1949,15 @@ static void __floating_irq_kick(struct kvm *kvm, u64 type) if (!online_vcpus) return; - /* find idle VCPUs first, then round robin */ - sigcpu = find_first_bit(kvm->arch.idle_mask, online_vcpus); - if (sigcpu == online_vcpus) { - do { - sigcpu = kvm->arch.float_int.next_rr_cpu++; - kvm->arch.float_int.next_rr_cpu %= online_vcpus; - /* avoid endless loops if all vcpus are stopped */ - if (nr_tries++ >= online_vcpus) - return; - } while (is_vcpu_stopped(kvm_get_vcpu(kvm, sigcpu))); + for (sigcpu = kvm->arch.float_int.last_sleep_cpu; ; sigcpu++) { + sigcpu %= online_vcpus; + dst_vcpu = kvm_get_vcpu(kvm, sigcpu); + if (!is_vcpu_stopped(dst_vcpu)) + break; + /* avoid endless loops if all vcpus are stopped */ + if (nr_tries++ >= online_vcpus) + return; } - dst_vcpu = kvm_get_vcpu(kvm, sigcpu); /* make the VCPU drop out of the SIE, or wake it up if sleeping */ switch (type) { From 5deafa27d9ae040b75d392f60b12e300b42b4792 Mon Sep 17 00:00:00 2001 From: Gautam Gala Date: Wed, 24 Sep 2025 13:26:44 +0200 Subject: [PATCH 2/2] KVM: s390: Fix to clear PTE when discarding a swapped page KVM run fails when guests with 'cmm' cpu feature and host are under memory pressure and use swap heavily. This is because npages becomes ENOMEN (out of memory) in hva_to_pfn_slow() which inturn propagates as EFAULT to qemu. Clearing the page table entry when discarding an address that maps to a swap entry resolves the issue. Fixes: 200197908dc4 ("KVM: s390: Refactor and split some gmap helpers") Cc: stable@vger.kernel.org Suggested-by: Claudio Imbrenda Signed-off-by: Gautam Gala Reviewed-by: Claudio Imbrenda Signed-off-by: Claudio Imbrenda --- arch/s390/include/asm/pgtable.h | 22 ++++++++++++++++++++++ arch/s390/mm/gmap_helpers.c | 12 +++++++++++- arch/s390/mm/pgtable.c | 23 +---------------------- 3 files changed, 34 insertions(+), 23 deletions(-) diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 6d8bc27a366e..324f96485604 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -2010,4 +2010,26 @@ static inline unsigned long gmap_pgste_get_pgt_addr(unsigned long *pgt) return res; } +static inline pgste_t pgste_get_lock(pte_t *ptep) +{ + unsigned long value = 0; +#ifdef CONFIG_PGSTE + unsigned long *ptr = (unsigned long *)(ptep + PTRS_PER_PTE); + + do { + value = __atomic64_or_barrier(PGSTE_PCL_BIT, ptr); + } while (value & PGSTE_PCL_BIT); + value |= PGSTE_PCL_BIT; +#endif + return __pgste(value); +} + +static inline void pgste_set_unlock(pte_t *ptep, pgste_t pgste) +{ +#ifdef CONFIG_PGSTE + barrier(); + WRITE_ONCE(*(unsigned long *)(ptep + PTRS_PER_PTE), pgste_val(pgste) & ~PGSTE_PCL_BIT); +#endif +} + #endif /* _S390_PAGE_H */ diff --git a/arch/s390/mm/gmap_helpers.c b/arch/s390/mm/gmap_helpers.c index a45d417ad951..c382005577bd 100644 --- a/arch/s390/mm/gmap_helpers.c +++ b/arch/s390/mm/gmap_helpers.c @@ -13,6 +13,7 @@ #include #include #include +#include /** * ptep_zap_swap_entry() - discard a swap entry. @@ -45,6 +46,7 @@ void gmap_helper_zap_one_page(struct mm_struct *mm, unsigned long vmaddr) { struct vm_area_struct *vma; spinlock_t *ptl; + pgste_t pgste; pte_t *ptep; mmap_assert_locked(mm); @@ -58,8 +60,16 @@ void gmap_helper_zap_one_page(struct mm_struct *mm, unsigned long vmaddr) ptep = get_locked_pte(mm, vmaddr, &ptl); if (unlikely(!ptep)) return; - if (pte_swap(*ptep)) + if (pte_swap(*ptep)) { + preempt_disable(); + pgste = pgste_get_lock(ptep); + ptep_zap_swap_entry(mm, pte_to_swp_entry(*ptep)); + pte_clear(mm, vmaddr, ptep); + + pgste_set_unlock(ptep, pgste); + preempt_enable(); + } pte_unmap_unlock(ptep, ptl); } EXPORT_SYMBOL_GPL(gmap_helper_zap_one_page); diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 7df70cd8f739..6b92c348b56f 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -23,6 +23,7 @@ #include #include #include +#include #include pgprot_t pgprot_writecombine(pgprot_t prot) @@ -114,28 +115,6 @@ static inline pte_t ptep_flush_lazy(struct mm_struct *mm, return old; } -static inline pgste_t pgste_get_lock(pte_t *ptep) -{ - unsigned long value = 0; -#ifdef CONFIG_PGSTE - unsigned long *ptr = (unsigned long *)(ptep + PTRS_PER_PTE); - - do { - value = __atomic64_or_barrier(PGSTE_PCL_BIT, ptr); - } while (value & PGSTE_PCL_BIT); - value |= PGSTE_PCL_BIT; -#endif - return __pgste(value); -} - -static inline void pgste_set_unlock(pte_t *ptep, pgste_t pgste) -{ -#ifdef CONFIG_PGSTE - barrier(); - WRITE_ONCE(*(unsigned long *)(ptep + PTRS_PER_PTE), pgste_val(pgste) & ~PGSTE_PCL_BIT); -#endif -} - static inline pgste_t pgste_get(pte_t *ptep) { unsigned long pgste = 0;