From: Robin Murphy robin.murphy@arm.com

mainline inclusion from mainline-v6.7-rc1 commit 911aa1245da83ff5e76d33bb612d8b5a3f2ec4a5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8PJ8L

Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...

--------------------------------------

The algorithm in the original paper specifies the storage of full magazines in the depot as an unbounded list rather than a fixed-size array. It turns out to be pretty straightforward to do this in our implementation with no significant loss of efficiency. This allows the depot to scale up to the working set sizes of larger systems, while also potentially saving some memory on smaller ones too.

Since this involves touching struct iova_magazine with the requisite care, we may as well reinforce the comment with a proper assertion too.

Reviewed-by: John Garry john.g.garry@oracle.com Reviewed-by: Jerry Snitselaar jsnitsel@redhat.com Signed-off-by: Robin Murphy robin.murphy@arm.com Link: https://lore.kernel.org/r/f597aa72fc3e1d315bc4574af0ce0ebe5c31cd22.169453558... Signed-off-by: Joerg Roedel jroedel@suse.de Signed-off-by: Zhang Zekun zhangzekun11@huawei.com --- drivers/iommu/iova.c | 65 ++++++++++++++++++++++++-------------------- 1 file changed, 35 insertions(+), 30 deletions(-)

diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c index 10b964600948..dd2309e9a6c5 100644 --- a/drivers/iommu/iova.c +++ b/drivers/iommu/iova.c @@ -622,15 +622,19 @@ EXPORT_SYMBOL_GPL(reserve_iova); /* * As kmalloc's buffer size is fixed to power of 2, 127 is chosen to * assure size of 'iova_magazine' to be 1024 bytes, so that no memory - * will be wasted. + * will be wasted. Since only full magazines are inserted into the depot, + * we don't need to waste PFN capacity on a separate list head either. */ #define IOVA_MAG_SIZE 127 -#define MAX_GLOBAL_MAGS 32 /* magazines per bin */

struct iova_magazine { - unsigned long size; + union { + unsigned long size; + struct iova_magazine *next; + }; unsigned long pfns[IOVA_MAG_SIZE]; }; +static_assert(!(sizeof(struct iova_magazine) & (sizeof(struct iova_magazine) - 1)));

struct iova_cpu_rcache { spinlock_t lock; @@ -640,8 +644,7 @@ struct iova_cpu_rcache {

struct iova_rcache { spinlock_t lock; - unsigned long depot_size; - struct iova_magazine *depot[MAX_GLOBAL_MAGS]; + struct iova_magazine *depot; struct iova_cpu_rcache __percpu *cpu_rcaches; };

@@ -717,6 +720,21 @@ static void iova_magazine_push(struct iova_magazine *mag, unsigned long pfn) mag->pfns[mag->size++] = pfn; }

+static struct iova_magazine *iova_depot_pop(struct iova_rcache *rcache) +{ + struct iova_magazine *mag = rcache->depot; + + rcache->depot = mag->next; + mag->size = IOVA_MAG_SIZE; + return mag; +} + +static void iova_depot_push(struct iova_rcache *rcache, struct iova_magazine *mag) +{ + mag->next = rcache->depot; + rcache->depot = mag; +} + int iova_domain_init_rcaches(struct iova_domain *iovad) { unsigned int cpu; @@ -734,7 +752,6 @@ int iova_domain_init_rcaches(struct iova_domain *iovad)

rcache = &iovad->rcaches[i]; spin_lock_init(&rcache->lock); - rcache->depot_size = 0; rcache->cpu_rcaches = __alloc_percpu(sizeof(*cpu_rcache), cache_line_size()); if (!rcache->cpu_rcaches) { @@ -776,7 +793,6 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, struct iova_rcache *rcache, unsigned long iova_pfn) { - struct iova_magazine *mag_to_free = NULL; struct iova_cpu_rcache *cpu_rcache; bool can_insert = false; unsigned long flags; @@ -794,12 +810,7 @@ static bool __iova_rcache_insert(struct iova_domain *iovad,

if (new_mag) { spin_lock(&rcache->lock); - if (rcache->depot_size < MAX_GLOBAL_MAGS) { - rcache->depot[rcache->depot_size++] = - cpu_rcache->loaded; - } else { - mag_to_free = cpu_rcache->loaded; - } + iova_depot_push(rcache, cpu_rcache->loaded); spin_unlock(&rcache->lock);

cpu_rcache->loaded = new_mag; @@ -812,11 +823,6 @@ static bool __iova_rcache_insert(struct iova_domain *iovad,

spin_unlock_irqrestore(&cpu_rcache->lock, flags);

- if (mag_to_free) { - iova_magazine_free_pfns(mag_to_free, iovad); - iova_magazine_free(mag_to_free); - } - return can_insert; }

@@ -854,9 +860,9 @@ static unsigned long __iova_rcache_get(struct iova_rcache *rcache, has_pfn = true; } else { spin_lock(&rcache->lock); - if (rcache->depot_size > 0) { + if (rcache->depot) { iova_magazine_free(cpu_rcache->loaded); - cpu_rcache->loaded = rcache->depot[--rcache->depot_size]; + cpu_rcache->loaded = iova_depot_pop(rcache); has_pfn = true; } spin_unlock(&rcache->lock); @@ -895,9 +901,8 @@ static void free_iova_rcaches(struct iova_domain *iovad) struct iova_rcache *rcache; struct iova_cpu_rcache *cpu_rcache; unsigned int cpu; - int i, j;

- for (i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { + for (int i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { rcache = &iovad->rcaches[i]; if (!rcache->cpu_rcaches) break; @@ -907,8 +912,8 @@ static void free_iova_rcaches(struct iova_domain *iovad) iova_magazine_free(cpu_rcache->prev); } free_percpu(rcache->cpu_rcaches); - for (j = 0; j < rcache->depot_size; ++j) - iova_magazine_free(rcache->depot[j]); + while (rcache->depot) + iova_magazine_free(iova_depot_pop(rcache)); }

kfree(iovad->rcaches); @@ -942,16 +947,16 @@ static void free_global_cached_iovas(struct iova_domain *iovad) { struct iova_rcache *rcache; unsigned long flags; - int i, j;

- for (i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { + for (int i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { rcache = &iovad->rcaches[i]; spin_lock_irqsave(&rcache->lock, flags); - for (j = 0; j < rcache->depot_size; ++j) { - iova_magazine_free_pfns(rcache->depot[j], iovad); - iova_magazine_free(rcache->depot[j]); + while (rcache->depot) { + struct iova_magazine *mag = iova_depot_pop(rcache); + + iova_magazine_free_pfns(mag, iovad); + iova_magazine_free(mag); } - rcache->depot_size = 0; spin_unlock_irqrestore(&rcache->lock, flags); } }

From: Wenkai Lin linwenkai6@hisilicon.com

maillist inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8PSC6

Reference: https://patchwork.kernel.org/project/kvm/patch/20231206005727.46150-1-zhangf...

----------------------------------------------

In the stall model, invalid transactions were expected to be stalled and aborted by the IOPF handler.

However, when killing a test case with a huge amount of data, the accelerator streamline can not stop until all data is consumed even if the page fault handler reports errors. As a result, the kill may take a long time, about 10 seconds with numerous iopf interrupts.

So disable stall for quiet_cd in the non-force stall model, since force stall model (STALL_MODEL==0b10) requires CD.S must be 1.

Signed-off-by: Zhangfei Gao zhangfei.gao@linaro.org Signed-off-by: Wenkai Lin linwenkai6@hisilicon.com Suggested-by: Jean-Philippe Brucker jean-philippe@linaro.org Reviewed-by: Jason Gunthorpe jgg@nvidia.com Reviewed-by: Jean-Philippe Brucker jean-philippe@linaro.org Link: https://lore.kernel.org/r/20231206005727.46150-1-zhangfei.gao@linaro.org Signed-off-by: Will Deacon will@kernel.org Signed-off-by: Zhang Zekun zhangzekun11@huawei.com --- drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 3 +++ 1 file changed, 3 insertions(+)

diff --git a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c index 4d417b4243eb..5055a66644af 100644 --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c @@ -1070,6 +1070,7 @@ int arm_smmu_write_ctx_desc(struct arm_smmu_domain *smmu_domain, int ssid, u64 val; bool cd_live; __le64 *cdptr; + struct arm_smmu_device *smmu = smmu_domain->smmu;

if (WARN_ON(ssid >= (1 << smmu_domain->s1_cfg.s1cdmax))) return -E2BIG; @@ -1084,6 +1085,8 @@ int arm_smmu_write_ctx_desc(struct arm_smmu_domain *smmu_domain, int ssid, if (!cd) { /* (5) */ val = 0; } else if (cd == &quiet_cd) { /* (4) */ + if (!(smmu->features & ARM_SMMU_FEAT_STALL_FORCE)) + val &= ~(CTXDESC_CD_0_S | CTXDESC_CD_0_R); val |= CTXDESC_CD_0_TCR_EPD0; } else if (cd_live) { /* (3) */ val &= ~CTXDESC_CD_0_ASID;