| Message ID | 20220707145248.458771-2-apatel@ventanamicro.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | KVM RISC-V Svpbmt support | expand |
On Thu, Jul 7, 2022 at 7:53 AM Anup Patel <apatel@ventanamicro.com> wrote: > > The kvm_mmu_topup_memory_cache() always uses GFP_KERNEL_ACCOUNT for > memory allocation which prevents it's use in atomic context. To address > this limitation of kvm_mmu_topup_memory_cache(), we add gfp_custom flag > in struct kvm_mmu_memory_cache. When the gfp_custom flag is set to some > GFP_xyz flags, the kvm_mmu_topup_memory_cache() will use that instead of > GFP_KERNEL_ACCOUNT. > > Signed-off-by: Anup Patel <apatel@ventanamicro.com> > --- > include/linux/kvm_types.h | 1 + > virt/kvm/kvm_main.c | 4 +++- > 2 files changed, 4 insertions(+), 1 deletion(-) > > diff --git a/include/linux/kvm_types.h b/include/linux/kvm_types.h > index ac1ebb37a0ff..1dcfba68076a 100644 > --- a/include/linux/kvm_types.h > +++ b/include/linux/kvm_types.h > @@ -87,6 +87,7 @@ struct gfn_to_pfn_cache { > struct kvm_mmu_memory_cache { > int nobjs; > gfp_t gfp_zero; > + gfp_t gfp_custom; > struct kmem_cache *kmem_cache; > void *objects[KVM_ARCH_NR_OBJS_PER_MEMORY_CACHE]; > }; > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c > index a49df8988cd6..e3a6f7647474 100644 > --- a/virt/kvm/kvm_main.c > +++ b/virt/kvm/kvm_main.c > @@ -386,7 +386,9 @@ int kvm_mmu_topup_memory_cache(struct kvm_mmu_memory_cache *mc, int min) > if (mc->nobjs >= min) > return 0; > while (mc->nobjs < ARRAY_SIZE(mc->objects)) { > - obj = mmu_memory_cache_alloc_obj(mc, GFP_KERNEL_ACCOUNT); > + obj = mmu_memory_cache_alloc_obj(mc, (mc->gfp_custom) ? > + mc->gfp_custom : > + GFP_KERNEL_ACCOUNT); > if (!obj) > return mc->nobjs >= min ? 0 : -ENOMEM; > mc->objects[mc->nobjs++] = obj; > -- > 2.34.1 > Acked-by: Atish Patra <atishp@rivosinc.com>
Hi Paolo, On Wed, Jul 13, 2022 at 6:57 AM Atish Patra <atishp@atishpatra.org> wrote: > > On Thu, Jul 7, 2022 at 7:53 AM Anup Patel <apatel@ventanamicro.com> wrote: > > > > The kvm_mmu_topup_memory_cache() always uses GFP_KERNEL_ACCOUNT for > > memory allocation which prevents it's use in atomic context. To address > > this limitation of kvm_mmu_topup_memory_cache(), we add gfp_custom flag > > in struct kvm_mmu_memory_cache. When the gfp_custom flag is set to some > > GFP_xyz flags, the kvm_mmu_topup_memory_cache() will use that instead of > > GFP_KERNEL_ACCOUNT. > > > > Signed-off-by: Anup Patel <apatel@ventanamicro.com> > > --- > > include/linux/kvm_types.h | 1 + > > virt/kvm/kvm_main.c | 4 +++- > > 2 files changed, 4 insertions(+), 1 deletion(-) > > > > diff --git a/include/linux/kvm_types.h b/include/linux/kvm_types.h > > index ac1ebb37a0ff..1dcfba68076a 100644 > > --- a/include/linux/kvm_types.h > > +++ b/include/linux/kvm_types.h > > @@ -87,6 +87,7 @@ struct gfn_to_pfn_cache { > > struct kvm_mmu_memory_cache { > > int nobjs; > > gfp_t gfp_zero; > > + gfp_t gfp_custom; > > struct kmem_cache *kmem_cache; > > void *objects[KVM_ARCH_NR_OBJS_PER_MEMORY_CACHE]; > > }; > > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c > > index a49df8988cd6..e3a6f7647474 100644 > > --- a/virt/kvm/kvm_main.c > > +++ b/virt/kvm/kvm_main.c > > @@ -386,7 +386,9 @@ int kvm_mmu_topup_memory_cache(struct kvm_mmu_memory_cache *mc, int min) > > if (mc->nobjs >= min) > > return 0; > > while (mc->nobjs < ARRAY_SIZE(mc->objects)) { > > - obj = mmu_memory_cache_alloc_obj(mc, GFP_KERNEL_ACCOUNT); > > + obj = mmu_memory_cache_alloc_obj(mc, (mc->gfp_custom) ? > > + mc->gfp_custom : > > + GFP_KERNEL_ACCOUNT); > > if (!obj) > > return mc->nobjs >= min ? 0 : -ENOMEM; > > mc->objects[mc->nobjs++] = obj; > > -- > > 2.34.1 > > > > Acked-by: Atish Patra <atishp@rivosinc.com> I have queued this patch for 5.20. Please let me know if you are not okay or need changes in this patch. Thanks, Anup
diff --git a/include/linux/kvm_types.h b/include/linux/kvm_types.h index ac1ebb37a0ff..1dcfba68076a 100644 --- a/include/linux/kvm_types.h +++ b/include/linux/kvm_types.h @@ -87,6 +87,7 @@ struct gfn_to_pfn_cache { struct kvm_mmu_memory_cache { int nobjs; gfp_t gfp_zero; + gfp_t gfp_custom; struct kmem_cache *kmem_cache; void *objects[KVM_ARCH_NR_OBJS_PER_MEMORY_CACHE]; }; diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index a49df8988cd6..e3a6f7647474 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -386,7 +386,9 @@ int kvm_mmu_topup_memory_cache(struct kvm_mmu_memory_cache *mc, int min) if (mc->nobjs >= min) return 0; while (mc->nobjs < ARRAY_SIZE(mc->objects)) { - obj = mmu_memory_cache_alloc_obj(mc, GFP_KERNEL_ACCOUNT); + obj = mmu_memory_cache_alloc_obj(mc, (mc->gfp_custom) ? + mc->gfp_custom : + GFP_KERNEL_ACCOUNT); if (!obj) return mc->nobjs >= min ? 0 : -ENOMEM; mc->objects[mc->nobjs++] = obj;
The kvm_mmu_topup_memory_cache() always uses GFP_KERNEL_ACCOUNT for memory allocation which prevents it's use in atomic context. To address this limitation of kvm_mmu_topup_memory_cache(), we add gfp_custom flag in struct kvm_mmu_memory_cache. When the gfp_custom flag is set to some GFP_xyz flags, the kvm_mmu_topup_memory_cache() will use that instead of GFP_KERNEL_ACCOUNT. Signed-off-by: Anup Patel <apatel@ventanamicro.com> --- include/linux/kvm_types.h | 1 + virt/kvm/kvm_main.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-)