| Message ID | 20260104133457.57742-1-luxu.kernel@bytedance.com |
|---|---|
| State | New |
| Headers | show |
| Series | [v5] irqchip/riscv-imsic: Adjust the number of available guest irq files | expand |
On 1/4/2026 9:34 PM, Xu Lu wrote: > Currently, KVM assumes the minimum of implemented HGEIE bits and > "BIT(gc->guest_index_bits) - 1" as the number of guest files available > across all CPUs. This will not work when CPUs have different number > of guest files because KVM may incorrectly allocate a guest file on a > CPU with fewer guest files. > > To address above, during initialization, calculate the number of > available guest interrupt files according to MMIO resources and > constrain the number of guest interrupt files that can be allocated > by KVM. > > Signed-off-by: Xu Lu <luxu.kernel@bytedance.com> > --- > arch/riscv/kvm/aia.c | 2 +- > drivers/irqchip/irq-riscv-imsic-state.c | 12 +++++++++++- > include/linux/irqchip/riscv-imsic.h | 3 +++ > 3 files changed, 15 insertions(+), 2 deletions(-) Reviewed-by: Nutty Liu <nutty.liu@hotmail.com> Thanks, Nutty > diff --git a/arch/riscv/kvm/aia.c b/arch/riscv/kvm/aia.c > index dad3181856600..cac3c2b51d724 100644 > --- a/arch/riscv/kvm/aia.c > +++ b/arch/riscv/kvm/aia.c > @@ -630,7 +630,7 @@ int kvm_riscv_aia_init(void) > */ > if (gc) > kvm_riscv_aia_nr_hgei = min((ulong)kvm_riscv_aia_nr_hgei, > - BIT(gc->guest_index_bits) - 1); > + gc->nr_guest_files); > else > kvm_riscv_aia_nr_hgei = 0; > > diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-riscv-imsic-state.c > index dc95ad856d80a..e8f20efb028be 100644 > --- a/drivers/irqchip/irq-riscv-imsic-state.c > +++ b/drivers/irqchip/irq-riscv-imsic-state.c > @@ -794,7 +794,7 @@ static int __init imsic_parse_fwnode(struct fwnode_handle *fwnode, > > int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > { > - u32 i, j, index, nr_parent_irqs, nr_mmios, nr_handlers = 0; > + u32 i, j, index, nr_parent_irqs, nr_mmios, nr_guest_files, nr_handlers = 0; > struct imsic_global_config *global; > struct imsic_local_config *local; > void __iomem **mmios_va = NULL; > @@ -888,6 +888,7 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > } > > /* Configure handlers for target CPUs */ > + global->nr_guest_files = BIT(global->guest_index_bits) - 1; > for (i = 0; i < nr_parent_irqs; i++) { > rc = imsic_get_parent_hartid(fwnode, i, &hartid); > if (rc) { > @@ -928,6 +929,15 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > local->msi_pa = mmios[index].start + reloff; > local->msi_va = mmios_va[index] + reloff; > > + /* > + * KVM uses global->nr_guest_files to determine the available guest > + * interrupt files on each CPU. Take the minimum number of guest > + * interrupt files across all CPUs to avoid KVM incorrectly allocating > + * an unexisted or unmapped guest interrupt file on some CPUs. > + */ > + nr_guest_files = (resource_size(&mmios[index]) - reloff) / IMSIC_MMIO_PAGE_SZ - 1; > + global->nr_guest_files = min(global->nr_guest_files, nr_guest_files); > + > nr_handlers++; > } > > diff --git a/include/linux/irqchip/riscv-imsic.h b/include/linux/irqchip/riscv-imsic.h > index 7494952c55187..43aed52385008 100644 > --- a/include/linux/irqchip/riscv-imsic.h > +++ b/include/linux/irqchip/riscv-imsic.h > @@ -69,6 +69,9 @@ struct imsic_global_config { > /* Number of guest interrupt identities */ > u32 nr_guest_ids; > > + /* Number of guest interrupt files per core */ > + u32 nr_guest_files; > + > /* Per-CPU IMSIC addresses */ > struct imsic_local_config __percpu *local; > };
On Sun, Jan 4, 2026 at 7:05 PM Xu Lu <luxu.kernel@bytedance.com> wrote: > > Currently, KVM assumes the minimum of implemented HGEIE bits and > "BIT(gc->guest_index_bits) - 1" as the number of guest files available > across all CPUs. This will not work when CPUs have different number > of guest files because KVM may incorrectly allocate a guest file on a > CPU with fewer guest files. > > To address above, during initialization, calculate the number of > available guest interrupt files according to MMIO resources and > constrain the number of guest interrupt files that can be allocated > by KVM. > > Signed-off-by: Xu Lu <luxu.kernel@bytedance.com> Please carry Reviewed-by and Acked-by tags obtained in previous revisions. Next time, I will not take the patch if previous tags are missing. Queued this patch for Linux-6.20. Regards, Anup > --- > arch/riscv/kvm/aia.c | 2 +- > drivers/irqchip/irq-riscv-imsic-state.c | 12 +++++++++++- > include/linux/irqchip/riscv-imsic.h | 3 +++ > 3 files changed, 15 insertions(+), 2 deletions(-) > > diff --git a/arch/riscv/kvm/aia.c b/arch/riscv/kvm/aia.c > index dad3181856600..cac3c2b51d724 100644 > --- a/arch/riscv/kvm/aia.c > +++ b/arch/riscv/kvm/aia.c > @@ -630,7 +630,7 @@ int kvm_riscv_aia_init(void) > */ > if (gc) > kvm_riscv_aia_nr_hgei = min((ulong)kvm_riscv_aia_nr_hgei, > - BIT(gc->guest_index_bits) - 1); > + gc->nr_guest_files); > else > kvm_riscv_aia_nr_hgei = 0; > > diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-riscv-imsic-state.c > index dc95ad856d80a..e8f20efb028be 100644 > --- a/drivers/irqchip/irq-riscv-imsic-state.c > +++ b/drivers/irqchip/irq-riscv-imsic-state.c > @@ -794,7 +794,7 @@ static int __init imsic_parse_fwnode(struct fwnode_handle *fwnode, > > int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > { > - u32 i, j, index, nr_parent_irqs, nr_mmios, nr_handlers = 0; > + u32 i, j, index, nr_parent_irqs, nr_mmios, nr_guest_files, nr_handlers = 0; > struct imsic_global_config *global; > struct imsic_local_config *local; > void __iomem **mmios_va = NULL; > @@ -888,6 +888,7 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > } > > /* Configure handlers for target CPUs */ > + global->nr_guest_files = BIT(global->guest_index_bits) - 1; > for (i = 0; i < nr_parent_irqs; i++) { > rc = imsic_get_parent_hartid(fwnode, i, &hartid); > if (rc) { > @@ -928,6 +929,15 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > local->msi_pa = mmios[index].start + reloff; > local->msi_va = mmios_va[index] + reloff; > > + /* > + * KVM uses global->nr_guest_files to determine the available guest > + * interrupt files on each CPU. Take the minimum number of guest > + * interrupt files across all CPUs to avoid KVM incorrectly allocating > + * an unexisted or unmapped guest interrupt file on some CPUs. > + */ > + nr_guest_files = (resource_size(&mmios[index]) - reloff) / IMSIC_MMIO_PAGE_SZ - 1; > + global->nr_guest_files = min(global->nr_guest_files, nr_guest_files); > + > nr_handlers++; > } > > diff --git a/include/linux/irqchip/riscv-imsic.h b/include/linux/irqchip/riscv-imsic.h > index 7494952c55187..43aed52385008 100644 > --- a/include/linux/irqchip/riscv-imsic.h > +++ b/include/linux/irqchip/riscv-imsic.h > @@ -69,6 +69,9 @@ struct imsic_global_config { > /* Number of guest interrupt identities */ > u32 nr_guest_ids; > > + /* Number of guest interrupt files per core */ > + u32 nr_guest_files; > + > /* Per-CPU IMSIC addresses */ > struct imsic_local_config __percpu *local; > }; > -- > 2.20.1 > >
On Mon, Jan 26, 2026 at 6:54 PM Anup Patel <anup@brainfault.org> wrote: > > On Sun, Jan 4, 2026 at 7:05 PM Xu Lu <luxu.kernel@bytedance.com> wrote: > > > > Currently, KVM assumes the minimum of implemented HGEIE bits and > > "BIT(gc->guest_index_bits) - 1" as the number of guest files available > > across all CPUs. This will not work when CPUs have different number > > of guest files because KVM may incorrectly allocate a guest file on a > > CPU with fewer guest files. > > > > To address above, during initialization, calculate the number of > > available guest interrupt files according to MMIO resources and > > constrain the number of guest interrupt files that can be allocated > > by KVM. > > > > Signed-off-by: Xu Lu <luxu.kernel@bytedance.com> > > Please carry Reviewed-by and Acked-by tags obtained in previous > revisions. Next time, I will not take the patch if previous tags are > missing. Sorry about that. I thought the Reviewed-by and Acked-by tags belong to the previous version so didn't carry them. > > Queued this patch for Linux-6.20. Do I still need to resend the patch with Reviewed-by and Acked-by tags? Best regards, Xu Lu > > Regards, > Anup > > > --- > > arch/riscv/kvm/aia.c | 2 +- > > drivers/irqchip/irq-riscv-imsic-state.c | 12 +++++++++++- > > include/linux/irqchip/riscv-imsic.h | 3 +++ > > 3 files changed, 15 insertions(+), 2 deletions(-) > > > > diff --git a/arch/riscv/kvm/aia.c b/arch/riscv/kvm/aia.c > > index dad3181856600..cac3c2b51d724 100644 > > --- a/arch/riscv/kvm/aia.c > > +++ b/arch/riscv/kvm/aia.c > > @@ -630,7 +630,7 @@ int kvm_riscv_aia_init(void) > > */ > > if (gc) > > kvm_riscv_aia_nr_hgei = min((ulong)kvm_riscv_aia_nr_hgei, > > - BIT(gc->guest_index_bits) - 1); > > + gc->nr_guest_files); > > else > > kvm_riscv_aia_nr_hgei = 0; > > > > diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-riscv-imsic-state.c > > index dc95ad856d80a..e8f20efb028be 100644 > > --- a/drivers/irqchip/irq-riscv-imsic-state.c > > +++ b/drivers/irqchip/irq-riscv-imsic-state.c > > @@ -794,7 +794,7 @@ static int __init imsic_parse_fwnode(struct fwnode_handle *fwnode, > > > > int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > > { > > - u32 i, j, index, nr_parent_irqs, nr_mmios, nr_handlers = 0; > > + u32 i, j, index, nr_parent_irqs, nr_mmios, nr_guest_files, nr_handlers = 0; > > struct imsic_global_config *global; > > struct imsic_local_config *local; > > void __iomem **mmios_va = NULL; > > @@ -888,6 +888,7 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > > } > > > > /* Configure handlers for target CPUs */ > > + global->nr_guest_files = BIT(global->guest_index_bits) - 1; > > for (i = 0; i < nr_parent_irqs; i++) { > > rc = imsic_get_parent_hartid(fwnode, i, &hartid); > > if (rc) { > > @@ -928,6 +929,15 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > > local->msi_pa = mmios[index].start + reloff; > > local->msi_va = mmios_va[index] + reloff; > > > > + /* > > + * KVM uses global->nr_guest_files to determine the available guest > > + * interrupt files on each CPU. Take the minimum number of guest > > + * interrupt files across all CPUs to avoid KVM incorrectly allocating > > + * an unexisted or unmapped guest interrupt file on some CPUs. > > + */ > > + nr_guest_files = (resource_size(&mmios[index]) - reloff) / IMSIC_MMIO_PAGE_SZ - 1; > > + global->nr_guest_files = min(global->nr_guest_files, nr_guest_files); > > + > > nr_handlers++; > > } > > > > diff --git a/include/linux/irqchip/riscv-imsic.h b/include/linux/irqchip/riscv-imsic.h > > index 7494952c55187..43aed52385008 100644 > > --- a/include/linux/irqchip/riscv-imsic.h > > +++ b/include/linux/irqchip/riscv-imsic.h > > @@ -69,6 +69,9 @@ struct imsic_global_config { > > /* Number of guest interrupt identities */ > > u32 nr_guest_ids; > > > > + /* Number of guest interrupt files per core */ > > + u32 nr_guest_files; > > + > > /* Per-CPU IMSIC addresses */ > > struct imsic_local_config __percpu *local; > > }; > > -- > > 2.20.1 > > > >
On Mon, Jan 26, 2026 at 4:37 PM Xu Lu <luxu.kernel@bytedance.com> wrote: > > On Mon, Jan 26, 2026 at 6:54 PM Anup Patel <anup@brainfault.org> wrote: > > > > On Sun, Jan 4, 2026 at 7:05 PM Xu Lu <luxu.kernel@bytedance.com> wrote: > > > > > > Currently, KVM assumes the minimum of implemented HGEIE bits and > > > "BIT(gc->guest_index_bits) - 1" as the number of guest files available > > > across all CPUs. This will not work when CPUs have different number > > > of guest files because KVM may incorrectly allocate a guest file on a > > > CPU with fewer guest files. > > > > > > To address above, during initialization, calculate the number of > > > available guest interrupt files according to MMIO resources and > > > constrain the number of guest interrupt files that can be allocated > > > by KVM. > > > > > > Signed-off-by: Xu Lu <luxu.kernel@bytedance.com> > > > > Please carry Reviewed-by and Acked-by tags obtained in previous > > revisions. Next time, I will not take the patch if previous tags are > > missing. > > Sorry about that. I thought the Reviewed-by and Acked-by tags belong > to the previous version so didn't carry them. > > > > > Queued this patch for Linux-6.20. > > Do I still need to resend the patch with Reviewed-by and Acked-by tags? I have included the tags at the time of the merging patch. Regards, Anup > > Best regards, > Xu Lu > > > > > Regards, > > Anup > > > > > --- > > > arch/riscv/kvm/aia.c | 2 +- > > > drivers/irqchip/irq-riscv-imsic-state.c | 12 +++++++++++- > > > include/linux/irqchip/riscv-imsic.h | 3 +++ > > > 3 files changed, 15 insertions(+), 2 deletions(-) > > > > > > diff --git a/arch/riscv/kvm/aia.c b/arch/riscv/kvm/aia.c > > > index dad3181856600..cac3c2b51d724 100644 > > > --- a/arch/riscv/kvm/aia.c > > > +++ b/arch/riscv/kvm/aia.c > > > @@ -630,7 +630,7 @@ int kvm_riscv_aia_init(void) > > > */ > > > if (gc) > > > kvm_riscv_aia_nr_hgei = min((ulong)kvm_riscv_aia_nr_hgei, > > > - BIT(gc->guest_index_bits) - 1); > > > + gc->nr_guest_files); > > > else > > > kvm_riscv_aia_nr_hgei = 0; > > > > > > diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-riscv-imsic-state.c > > > index dc95ad856d80a..e8f20efb028be 100644 > > > --- a/drivers/irqchip/irq-riscv-imsic-state.c > > > +++ b/drivers/irqchip/irq-riscv-imsic-state.c > > > @@ -794,7 +794,7 @@ static int __init imsic_parse_fwnode(struct fwnode_handle *fwnode, > > > > > > int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > > > { > > > - u32 i, j, index, nr_parent_irqs, nr_mmios, nr_handlers = 0; > > > + u32 i, j, index, nr_parent_irqs, nr_mmios, nr_guest_files, nr_handlers = 0; > > > struct imsic_global_config *global; > > > struct imsic_local_config *local; > > > void __iomem **mmios_va = NULL; > > > @@ -888,6 +888,7 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > > > } > > > > > > /* Configure handlers for target CPUs */ > > > + global->nr_guest_files = BIT(global->guest_index_bits) - 1; > > > for (i = 0; i < nr_parent_irqs; i++) { > > > rc = imsic_get_parent_hartid(fwnode, i, &hartid); > > > if (rc) { > > > @@ -928,6 +929,15 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) > > > local->msi_pa = mmios[index].start + reloff; > > > local->msi_va = mmios_va[index] + reloff; > > > > > > + /* > > > + * KVM uses global->nr_guest_files to determine the available guest > > > + * interrupt files on each CPU. Take the minimum number of guest > > > + * interrupt files across all CPUs to avoid KVM incorrectly allocating > > > + * an unexisted or unmapped guest interrupt file on some CPUs. > > > + */ > > > + nr_guest_files = (resource_size(&mmios[index]) - reloff) / IMSIC_MMIO_PAGE_SZ - 1; > > > + global->nr_guest_files = min(global->nr_guest_files, nr_guest_files); > > > + > > > nr_handlers++; > > > } > > > > > > diff --git a/include/linux/irqchip/riscv-imsic.h b/include/linux/irqchip/riscv-imsic.h > > > index 7494952c55187..43aed52385008 100644 > > > --- a/include/linux/irqchip/riscv-imsic.h > > > +++ b/include/linux/irqchip/riscv-imsic.h > > > @@ -69,6 +69,9 @@ struct imsic_global_config { > > > /* Number of guest interrupt identities */ > > > u32 nr_guest_ids; > > > > > > + /* Number of guest interrupt files per core */ > > > + u32 nr_guest_files; > > > + > > > /* Per-CPU IMSIC addresses */ > > > struct imsic_local_config __percpu *local; > > > }; > > > -- > > > 2.20.1 > > > > > >
diff --git a/arch/riscv/kvm/aia.c b/arch/riscv/kvm/aia.c index dad3181856600..cac3c2b51d724 100644 --- a/arch/riscv/kvm/aia.c +++ b/arch/riscv/kvm/aia.c @@ -630,7 +630,7 @@ int kvm_riscv_aia_init(void) */ if (gc) kvm_riscv_aia_nr_hgei = min((ulong)kvm_riscv_aia_nr_hgei, - BIT(gc->guest_index_bits) - 1); + gc->nr_guest_files); else kvm_riscv_aia_nr_hgei = 0; diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-riscv-imsic-state.c index dc95ad856d80a..e8f20efb028be 100644 --- a/drivers/irqchip/irq-riscv-imsic-state.c +++ b/drivers/irqchip/irq-riscv-imsic-state.c @@ -794,7 +794,7 @@ static int __init imsic_parse_fwnode(struct fwnode_handle *fwnode, int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) { - u32 i, j, index, nr_parent_irqs, nr_mmios, nr_handlers = 0; + u32 i, j, index, nr_parent_irqs, nr_mmios, nr_guest_files, nr_handlers = 0; struct imsic_global_config *global; struct imsic_local_config *local; void __iomem **mmios_va = NULL; @@ -888,6 +888,7 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) } /* Configure handlers for target CPUs */ + global->nr_guest_files = BIT(global->guest_index_bits) - 1; for (i = 0; i < nr_parent_irqs; i++) { rc = imsic_get_parent_hartid(fwnode, i, &hartid); if (rc) { @@ -928,6 +929,15 @@ int __init imsic_setup_state(struct fwnode_handle *fwnode, void *opaque) local->msi_pa = mmios[index].start + reloff; local->msi_va = mmios_va[index] + reloff; + /* + * KVM uses global->nr_guest_files to determine the available guest + * interrupt files on each CPU. Take the minimum number of guest + * interrupt files across all CPUs to avoid KVM incorrectly allocating + * an unexisted or unmapped guest interrupt file on some CPUs. + */ + nr_guest_files = (resource_size(&mmios[index]) - reloff) / IMSIC_MMIO_PAGE_SZ - 1; + global->nr_guest_files = min(global->nr_guest_files, nr_guest_files); + nr_handlers++; } diff --git a/include/linux/irqchip/riscv-imsic.h b/include/linux/irqchip/riscv-imsic.h index 7494952c55187..43aed52385008 100644 --- a/include/linux/irqchip/riscv-imsic.h +++ b/include/linux/irqchip/riscv-imsic.h @@ -69,6 +69,9 @@ struct imsic_global_config { /* Number of guest interrupt identities */ u32 nr_guest_ids; + /* Number of guest interrupt files per core */ + u32 nr_guest_files; + /* Per-CPU IMSIC addresses */ struct imsic_local_config __percpu *local; };
Currently, KVM assumes the minimum of implemented HGEIE bits and "BIT(gc->guest_index_bits) - 1" as the number of guest files available across all CPUs. This will not work when CPUs have different number of guest files because KVM may incorrectly allocate a guest file on a CPU with fewer guest files. To address above, during initialization, calculate the number of available guest interrupt files according to MMIO resources and constrain the number of guest interrupt files that can be allocated by KVM. Signed-off-by: Xu Lu <luxu.kernel@bytedance.com> --- arch/riscv/kvm/aia.c | 2 +- drivers/irqchip/irq-riscv-imsic-state.c | 12 +++++++++++- include/linux/irqchip/riscv-imsic.h | 3 +++ 3 files changed, 15 insertions(+), 2 deletions(-)