From 551407d426a3f5e0146235a7acdeb66ad66338eb Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:09:35 +0800 Subject: [PATCH 01/18] iommu: Move IOMMU_DOMAIN_BLOCKED global statics to ops->blocked_domain ANBZ: #21698 commit 06c375053cefc3a2f383d200596abe5ab3fb35f9 upstream. Following the pattern of identity domains, just assign the BLOCKED domain global statics to a value in ops. Update the core code to use the global static directly. Update powerpc to use the new scheme and remove its empty domain_alloc callback. Reviewed-by: Lu Baolu Signed-off-by: Jason Gunthorpe Reviewed-by: Kevin Tian Acked-by: Sven Peter Link: https://lore.kernel.org/r/1-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- arch/powerpc/kernel/iommu.c | 9 +-------- drivers/iommu/iommu.c | 2 ++ include/linux/iommu.h | 3 +++ 3 files changed, 6 insertions(+), 8 deletions(-) diff --git a/arch/powerpc/kernel/iommu.c b/arch/powerpc/kernel/iommu.c index b1e38b78ed52..33a1b2a1bf95 100644 --- a/arch/powerpc/kernel/iommu.c +++ b/arch/powerpc/kernel/iommu.c @@ -1327,13 +1327,6 @@ static bool spapr_tce_iommu_capable(struct device *dev, enum iommu_cap cap) return false; } -static struct iommu_domain *spapr_tce_iommu_domain_alloc(unsigned int type) -{ - if (type != IOMMU_DOMAIN_BLOCKED) - return NULL; - return &spapr_tce_blocked_domain; -} - static struct iommu_device *spapr_tce_iommu_probe_device(struct device *dev) { struct pci_dev *pdev; @@ -1368,8 +1361,8 @@ static struct iommu_group *spapr_tce_iommu_device_group(struct device *dev) static const struct iommu_ops spapr_tce_iommu_ops = { .default_domain = &spapr_tce_platform_domain, + .blocked_domain = &spapr_tce_blocked_domain, .capable = spapr_tce_iommu_capable, - .domain_alloc = spapr_tce_iommu_domain_alloc, .probe_device = spapr_tce_iommu_probe_device, .release_device = spapr_tce_iommu_release_device, .device_group = spapr_tce_iommu_device_group, diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c index 8ba53882d2f0..b85e62bf16d3 100644 --- a/drivers/iommu/iommu.c +++ b/drivers/iommu/iommu.c @@ -2070,6 +2070,8 @@ static struct iommu_domain *__iommu_domain_alloc(const struct iommu_ops *ops, if (alloc_type == IOMMU_DOMAIN_IDENTITY && ops->identity_domain) return ops->identity_domain; + else if (alloc_type == IOMMU_DOMAIN_BLOCKED && ops->blocked_domain) + return ops->blocked_domain; else if (type & __IOMMU_DOMAIN_PAGING && ops->domain_alloc_paging) domain = ops->domain_alloc_paging(dev); else if (ops->domain_alloc) diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 2e37fae3e776..e98ec027dce0 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -315,6 +315,8 @@ struct iommu_dirty_ops { * @owner: Driver module providing these ops * @identity_domain: An always available, always attachable identity * translation. + * @blocked_domain: An always available, always attachable blocking + * translation. * @default_domain: If not NULL this will always be set as the default domain. * This should be an IDENTITY/BLOCKED/PLATFORM domain. * Do not use in new drivers. @@ -354,6 +356,7 @@ struct iommu_ops { unsigned long pgsize_bitmap; struct module *owner; struct iommu_domain *identity_domain; + struct iommu_domain *blocked_domain; struct iommu_domain *default_domain; CK_KABI_RESERVE(1) -- Gitee From 62ac2b21ad935ee7eb747be57730d732d3aec883 Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:15:55 +0800 Subject: [PATCH 02/18] iommu/vt-d: Update the definition of the blocking domain ANBZ: #21698 commit 7b6dd84e703147f5c71f163b84b0cb729a320541 upstream. iommu/vt-d: Update the definition of the blocking domain The global static should pre-define the type and the NOP free function can be now left as NULL. Reviewed-by: Lu Baolu Signed-off-by: Jason Gunthorpe Reviewed-by: Kevin Tian Acked-by: Sven Peter Link: https://lore.kernel.org/r/2-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 6638b8287b0e..574c2dc65316 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -4090,9 +4090,9 @@ static int blocking_domain_attach_dev(struct iommu_domain *domain, } static struct iommu_domain blocking_domain = { + .type = IOMMU_DOMAIN_BLOCKED, .ops = &(const struct iommu_domain_ops) { .attach_dev = blocking_domain_attach_dev, - .free = intel_iommu_domain_free } }; @@ -4179,7 +4179,7 @@ intel_iommu_domain_alloc_user(struct device *dev, u32 flags) static void intel_iommu_domain_free(struct iommu_domain *domain) { - if (domain != &si_domain->domain && domain != &blocking_domain) + if (domain != &si_domain->domain) domain_exit(to_dmar_domain(domain)); } -- Gitee From 6f3a85e8bb348f7ca869e413839b8ac64d3e373c Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:20:53 +0800 Subject: [PATCH 03/18] iommu/vt-d: Use ops->blocked_domain ANBZ: #21698 commit 06c375053cefc3a2f383d200596abe5ab3fb35f9 upstream. Trivially migrate to the ops->blocked_domain for the existing global static. Reviewed-by: Lu Baolu Signed-off-by: Jason Gunthorpe Reviewed-by: Kevin Tian Acked-by: Sven Peter Link: https://lore.kernel.org/r/3-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 574c2dc65316..40fd449274bb 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -4102,8 +4102,6 @@ static struct iommu_domain *intel_iommu_domain_alloc(unsigned type) struct iommu_domain *domain; switch (type) { - case IOMMU_DOMAIN_BLOCKED: - return &blocking_domain; case IOMMU_DOMAIN_DMA: case IOMMU_DOMAIN_UNMANAGED: dmar_domain = alloc_domain(type); @@ -4994,6 +4992,7 @@ static const struct iommu_dirty_ops intel_dirty_ops = { }; const struct iommu_ops intel_iommu_ops = { + .blocked_domain = &blocking_domain, .capable = intel_iommu_capable, .hw_info = intel_iommu_hw_info, .domain_alloc = intel_iommu_domain_alloc, -- Gitee From d31ea6d8c12820129e3d2c477ae927d900e65d5f Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:26:45 +0800 Subject: [PATCH 04/18] iommufd: Convert to alloc_domain_paging() ANBZ: #21698 commit 13fbceb1b8e9d1101d9dcd5ceec483cb6d203a3c upstream. Move the global static blocked domain to the ops and convert the unmanaged domain to domain_alloc_paging. Signed-off-by: Jason Gunthorpe Reviewed-by: Kevin Tian Acked-by: Sven Peter Link: https://lore.kernel.org/r/4-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/iommufd/selftest.c | 15 +++++---------- 1 file changed, 5 insertions(+), 10 deletions(-) diff --git a/drivers/iommu/iommufd/selftest.c b/drivers/iommu/iommufd/selftest.c index 4eed87f936b4..8ea74157495d 100644 --- a/drivers/iommu/iommufd/selftest.c +++ b/drivers/iommu/iommufd/selftest.c @@ -228,16 +228,10 @@ const struct iommu_dirty_ops dirty_ops = { static const struct iommu_ops mock_ops; -static struct iommu_domain *mock_domain_alloc(unsigned int iommu_domain_type) +static struct iommu_domain *mock_domain_alloc_paging(struct device *dev) { struct mock_iommu_domain *mock; - if (iommu_domain_type == IOMMU_DOMAIN_BLOCKED) - return &mock_blocking_domain; - - if (iommu_domain_type != IOMMU_DOMAIN_UNMANAGED) - return NULL; - mock = kzalloc(sizeof(*mock), GFP_KERNEL); if (!mock) return NULL; @@ -245,7 +239,7 @@ static struct iommu_domain *mock_domain_alloc(unsigned int iommu_domain_type) mock->domain.geometry.aperture_end = MOCK_APERTURE_LAST; mock->domain.pgsize_bitmap = MOCK_IO_PAGE_SIZE; mock->domain.ops = mock_ops.default_domain_ops; - mock->domain.type = iommu_domain_type; + mock->domain.type = IOMMU_DOMAIN_UNMANAGED; xa_init(&mock->pfns); return &mock->domain; } @@ -264,7 +258,7 @@ mock_domain_alloc_user(struct device *dev, u32 flags) (mdev->flags & MOCK_FLAGS_DEVICE_NO_DIRTY)) return ERR_PTR(-EOPNOTSUPP); - domain = mock_domain_alloc(IOMMU_DOMAIN_UNMANAGED); + domain = mock_domain_alloc_paging(NULL); if (domain && !(mdev->flags & MOCK_FLAGS_DEVICE_NO_DIRTY)) domain->dirty_ops = &dirty_ops; if (!domain) @@ -417,10 +411,11 @@ static const struct iommu_ops mock_ops = { * because it is zero. */ .default_domain = &mock_blocking_domain, + .blocked_domain = &mock_blocking_domain, .owner = THIS_MODULE, .pgsize_bitmap = MOCK_IO_PAGE_SIZE, .hw_info = mock_domain_hw_info, - .domain_alloc = mock_domain_alloc, + .domain_alloc_paging = mock_domain_alloc_paging, .domain_alloc_user = mock_domain_alloc_user, .capable = mock_domain_capable, .device_group = generic_device_group, -- Gitee From e1a76e3f3ca53421c7f5c4874b73eb6bf0f588fb Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:32:05 +0800 Subject: [PATCH 05/18] iommu/dart: Use static global identity domains ANBZ: #21698 commit 7993085d8d5d271a83045c5c7d01982a7e89a52e upstream. Move to the new static global for identity domains. Move the identity specific code to apple_dart_attach_dev_identity(). Reviewed-by: Janne Grunau Signed-off-by: Jason Gunthorpe Acked-by: Sven Peter Link: https://lore.kernel.org/r/5-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/apple-dart.c | 39 +++++++++++++++++++++++++++----------- 1 file changed, 28 insertions(+), 11 deletions(-) diff --git a/drivers/iommu/apple-dart.c b/drivers/iommu/apple-dart.c index 2674fb4cd6f5..4624fa004a7a 100644 --- a/drivers/iommu/apple-dart.c +++ b/drivers/iommu/apple-dart.c @@ -659,11 +659,7 @@ static int apple_dart_attach_dev(struct iommu_domain *domain, struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); struct apple_dart_domain *dart_domain = to_dart_domain(domain); - if (cfg->stream_maps[0].dart->force_bypass && - domain->type != IOMMU_DOMAIN_IDENTITY) - return -EINVAL; - if (!cfg->stream_maps[0].dart->supports_bypass && - domain->type == IOMMU_DOMAIN_IDENTITY) + if (cfg->stream_maps[0].dart->force_bypass) return -EINVAL; ret = apple_dart_finalize_domain(domain, cfg); @@ -683,15 +679,35 @@ static int apple_dart_attach_dev(struct iommu_domain *domain, for_each_stream_map(i, cfg, stream_map) apple_dart_hw_disable_dma(stream_map); break; - case IOMMU_DOMAIN_IDENTITY: - for_each_stream_map(i, cfg, stream_map) - apple_dart_hw_enable_bypass(stream_map); - break; } return ret; } +static int apple_dart_attach_dev_identity(struct iommu_domain *domain, + struct device *dev) +{ + struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); + struct apple_dart_stream_map *stream_map; + int i; + + if (!cfg->stream_maps[0].dart->supports_bypass) + return -EINVAL; + + for_each_stream_map(i, cfg, stream_map) + apple_dart_hw_enable_bypass(stream_map); + return 0; +} + +static const struct iommu_domain_ops apple_dart_identity_ops = { + .attach_dev = apple_dart_attach_dev_identity, +}; + +static struct iommu_domain apple_dart_identity_domain = { + .type = IOMMU_DOMAIN_IDENTITY, + .ops = &apple_dart_identity_ops, +}; + static struct iommu_device *apple_dart_probe_device(struct device *dev) { struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); @@ -721,7 +737,7 @@ static struct iommu_domain *apple_dart_domain_alloc(unsigned int type) struct apple_dart_domain *dart_domain; if (type != IOMMU_DOMAIN_DMA && type != IOMMU_DOMAIN_UNMANAGED && - type != IOMMU_DOMAIN_IDENTITY && type != IOMMU_DOMAIN_BLOCKED) + type != IOMMU_DOMAIN_BLOCKED) return NULL; dart_domain = kzalloc(sizeof(*dart_domain), GFP_KERNEL); @@ -731,7 +747,7 @@ static struct iommu_domain *apple_dart_domain_alloc(unsigned int type) mutex_init(&dart_domain->init_lock); /* no need to allocate pgtbl_ops or do any other finalization steps */ - if (type == IOMMU_DOMAIN_IDENTITY || type == IOMMU_DOMAIN_BLOCKED) + if (type == IOMMU_DOMAIN_BLOCKED) dart_domain->finalized = true; return &dart_domain->domain; @@ -946,6 +962,7 @@ static void apple_dart_get_resv_regions(struct device *dev, } static const struct iommu_ops apple_dart_iommu_ops = { + .identity_domain = &apple_dart_identity_domain, .domain_alloc = apple_dart_domain_alloc, .probe_device = apple_dart_probe_device, .release_device = apple_dart_release_device, -- Gitee From 3425a033ab0137e004283083f11b47c2b1d75db7 Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:48:29 +0800 Subject: [PATCH 06/18] iommu/dart: Move the blocked domain support to a global static ANBZ: #21698 commit 17ef8d6876e01165836790d83d908eb21da3f08e upstream. Move to the new static global for blocked domains. Move the blocked specific code to apple_dart_attach_dev_blocked(). Reviewed-by: Janne Grunau Signed-off-by: Jason Gunthorpe Acked-by: Sven Peter Link: https://lore.kernel.org/r/6-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/apple-dart.c | 53 +++++++++++++++++++++++--------------- 1 file changed, 32 insertions(+), 21 deletions(-) diff --git a/drivers/iommu/apple-dart.c b/drivers/iommu/apple-dart.c index 4624fa004a7a..a18518962e72 100644 --- a/drivers/iommu/apple-dart.c +++ b/drivers/iommu/apple-dart.c @@ -666,22 +666,13 @@ static int apple_dart_attach_dev(struct iommu_domain *domain, if (ret) return ret; - switch (domain->type) { - default: - ret = apple_dart_domain_add_streams(dart_domain, cfg); - if (ret) - return ret; - - for_each_stream_map(i, cfg, stream_map) - apple_dart_setup_translation(dart_domain, stream_map); - break; - case IOMMU_DOMAIN_BLOCKED: - for_each_stream_map(i, cfg, stream_map) - apple_dart_hw_disable_dma(stream_map); - break; - } + ret = apple_dart_domain_add_streams(dart_domain, cfg); + if (ret) + return ret; - return ret; + for_each_stream_map(i, cfg, stream_map) + apple_dart_setup_translation(dart_domain, stream_map); + return 0; } static int apple_dart_attach_dev_identity(struct iommu_domain *domain, @@ -708,6 +699,30 @@ static struct iommu_domain apple_dart_identity_domain = { .ops = &apple_dart_identity_ops, }; +static int apple_dart_attach_dev_blocked(struct iommu_domain *domain, + struct device *dev) +{ + struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); + struct apple_dart_stream_map *stream_map; + int i; + + if (cfg->stream_maps[0].dart->force_bypass) + return -EINVAL; + + for_each_stream_map(i, cfg, stream_map) + apple_dart_hw_disable_dma(stream_map); + return 0; +} + +static const struct iommu_domain_ops apple_dart_blocked_ops = { + .attach_dev = apple_dart_attach_dev_blocked, +}; + +static struct iommu_domain apple_dart_blocked_domain = { + .type = IOMMU_DOMAIN_BLOCKED, + .ops = &apple_dart_blocked_ops, +}; + static struct iommu_device *apple_dart_probe_device(struct device *dev) { struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); @@ -736,8 +751,7 @@ static struct iommu_domain *apple_dart_domain_alloc(unsigned int type) { struct apple_dart_domain *dart_domain; - if (type != IOMMU_DOMAIN_DMA && type != IOMMU_DOMAIN_UNMANAGED && - type != IOMMU_DOMAIN_BLOCKED) + if (type != IOMMU_DOMAIN_DMA && type != IOMMU_DOMAIN_UNMANAGED) return NULL; dart_domain = kzalloc(sizeof(*dart_domain), GFP_KERNEL); @@ -746,10 +760,6 @@ static struct iommu_domain *apple_dart_domain_alloc(unsigned int type) mutex_init(&dart_domain->init_lock); - /* no need to allocate pgtbl_ops or do any other finalization steps */ - if (type == IOMMU_DOMAIN_BLOCKED) - dart_domain->finalized = true; - return &dart_domain->domain; } @@ -963,6 +973,7 @@ static void apple_dart_get_resv_regions(struct device *dev, static const struct iommu_ops apple_dart_iommu_ops = { .identity_domain = &apple_dart_identity_domain, + .blocked_domain = &apple_dart_blocked_domain, .domain_alloc = apple_dart_domain_alloc, .probe_device = apple_dart_probe_device, .release_device = apple_dart_release_device, -- Gitee From dd00e5fa20f6856028b23ce0886db45c0143d3e8 Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:52:40 +0800 Subject: [PATCH 07/18] iommu/dart: Convert to domain_alloc_paging() ANBZ: #21698 commit 9c3ef90c4ccbee0a000b0f6b1b28c38773de87c7 upstream. Since the IDENTITY and BLOCKED behaviors were moved to global statics all that remains is the paging domain. Rename to apple_dart_attach_dev_paging() and remove the left over type check. Reviewed-by: Janne Grunau Signed-off-by: Jason Gunthorpe Acked-by: Sven Peter Link: https://lore.kernel.org/r/7-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/apple-dart.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/drivers/iommu/apple-dart.c b/drivers/iommu/apple-dart.c index a18518962e72..a5b8b98f7d16 100644 --- a/drivers/iommu/apple-dart.c +++ b/drivers/iommu/apple-dart.c @@ -651,8 +651,8 @@ static int apple_dart_domain_add_streams(struct apple_dart_domain *domain, true); } -static int apple_dart_attach_dev(struct iommu_domain *domain, - struct device *dev) +static int apple_dart_attach_dev_paging(struct iommu_domain *domain, + struct device *dev) { int ret, i; struct apple_dart_stream_map *stream_map; @@ -747,13 +747,10 @@ static void apple_dart_release_device(struct device *dev) kfree(cfg); } -static struct iommu_domain *apple_dart_domain_alloc(unsigned int type) +static struct iommu_domain *apple_dart_domain_alloc_paging(struct device *dev) { struct apple_dart_domain *dart_domain; - if (type != IOMMU_DOMAIN_DMA && type != IOMMU_DOMAIN_UNMANAGED) - return NULL; - dart_domain = kzalloc(sizeof(*dart_domain), GFP_KERNEL); if (!dart_domain) return NULL; @@ -974,7 +971,7 @@ static void apple_dart_get_resv_regions(struct device *dev, static const struct iommu_ops apple_dart_iommu_ops = { .identity_domain = &apple_dart_identity_domain, .blocked_domain = &apple_dart_blocked_domain, - .domain_alloc = apple_dart_domain_alloc, + .domain_alloc_paging = apple_dart_domain_alloc_paging, .probe_device = apple_dart_probe_device, .release_device = apple_dart_release_device, .device_group = apple_dart_device_group, @@ -984,7 +981,7 @@ static const struct iommu_ops apple_dart_iommu_ops = { .pgsize_bitmap = -1UL, /* Restricted during dart probe */ .owner = THIS_MODULE, .default_domain_ops = &(const struct iommu_domain_ops) { - .attach_dev = apple_dart_attach_dev, + .attach_dev = apple_dart_attach_dev_paging, .map_pages = apple_dart_map_pages, .unmap_pages = apple_dart_unmap_pages, .flush_iotlb_all = apple_dart_flush_iotlb_all, -- Gitee From 51399e78ebdf02c1327fd0ee54d19471700fbde9 Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 19:57:59 +0800 Subject: [PATCH 08/18] iommu/dart: Call apple_dart_finalize_domain() as part of alloc_paging() ANBZ: #21698 commit 482feb5c649261cd2a7ad02e4ca63c159d6ec795 upstream. In many cases the dev argument will now be !NULL so we should use it to finalize the domain at allocation. Make apple_dart_finalize_domain() accept the correct type. Reviewed-by: Janne Grunau Signed-off-by: Jason Gunthorpe Acked-by: Sven Peter Link: https://lore.kernel.org/r/8-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/apple-dart.c | 28 +++++++++++++++++++--------- 1 file changed, 19 insertions(+), 9 deletions(-) diff --git a/drivers/iommu/apple-dart.c b/drivers/iommu/apple-dart.c index a5b8b98f7d16..8d386182ca31 100644 --- a/drivers/iommu/apple-dart.c +++ b/drivers/iommu/apple-dart.c @@ -568,10 +568,9 @@ apple_dart_setup_translation(struct apple_dart_domain *domain, stream_map->dart->hw->invalidate_tlb(stream_map); } -static int apple_dart_finalize_domain(struct iommu_domain *domain, +static int apple_dart_finalize_domain(struct apple_dart_domain *dart_domain, struct apple_dart_master_cfg *cfg) { - struct apple_dart_domain *dart_domain = to_dart_domain(domain); struct apple_dart *dart = cfg->stream_maps[0].dart; struct io_pgtable_cfg pgtbl_cfg; int ret = 0; @@ -597,17 +596,18 @@ static int apple_dart_finalize_domain(struct iommu_domain *domain, .iommu_dev = dart->dev, }; - dart_domain->pgtbl_ops = - alloc_io_pgtable_ops(dart->hw->fmt, &pgtbl_cfg, domain); + dart_domain->pgtbl_ops = alloc_io_pgtable_ops(dart->hw->fmt, &pgtbl_cfg, + &dart_domain->domain); if (!dart_domain->pgtbl_ops) { ret = -ENOMEM; goto done; } - domain->pgsize_bitmap = pgtbl_cfg.pgsize_bitmap; - domain->geometry.aperture_start = 0; - domain->geometry.aperture_end = (dma_addr_t)DMA_BIT_MASK(dart->ias); - domain->geometry.force_aperture = true; + dart_domain->domain.pgsize_bitmap = pgtbl_cfg.pgsize_bitmap; + dart_domain->domain.geometry.aperture_start = 0; + dart_domain->domain.geometry.aperture_end = + (dma_addr_t)DMA_BIT_MASK(dart->ias); + dart_domain->domain.geometry.force_aperture = true; dart_domain->finalized = true; @@ -662,7 +662,7 @@ static int apple_dart_attach_dev_paging(struct iommu_domain *domain, if (cfg->stream_maps[0].dart->force_bypass) return -EINVAL; - ret = apple_dart_finalize_domain(domain, cfg); + ret = apple_dart_finalize_domain(dart_domain, cfg); if (ret) return ret; @@ -757,6 +757,16 @@ static struct iommu_domain *apple_dart_domain_alloc_paging(struct device *dev) mutex_init(&dart_domain->init_lock); + if (dev) { + struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); + int ret; + + ret = apple_dart_finalize_domain(dart_domain, cfg); + if (ret) { + kfree(dart_domain); + return ERR_PTR(ret); + } + } return &dart_domain->domain; } -- Gitee From 6b502b9c2754bfd917a7c9d33abcd7a57d2d4bd0 Mon Sep 17 00:00:00 2001 From: gaorui Date: Mon, 9 Jun 2025 20:02:50 +0800 Subject: [PATCH 09/18] iommu/dart: Remove the force_bypass variable ANBZ: #21698 commit bbc70e0aec287e164344b1a071bd46466a4f29b3 upstream. This flag just caches if the IO page size is larger than the CPU PAGE_SIZE. This only needs to be checked in two places so remove the confusingly named cache. dart would like to not support paging domains at all if the IO page size is larger than the CPU page size. In this case we should ideally fail domain_alloc_paging(), as there is no point in creating a domain that can never be attached. Move the test into apple_dart_finalize_domain(). The check in apple_dart_mod_streams() will prevent the domain from being attached to the wrong dart There is no HW limitation that prevents BLOCKED domains from working, remove that test. The check in apple_dart_of_xlate() is redundant since immediately after the pgsize is checked. Remove it. Remove the variable. Suggested-by: Janne Grunau Signed-off-by: Jason Gunthorpe Reviewed-by: Janne Grunau Acked-by: Sven Peter Link: https://lore.kernel.org/r/9-v2-bff223cf6409+282-dart_paging_jgg@nvidia.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/apple-dart.c | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-) diff --git a/drivers/iommu/apple-dart.c b/drivers/iommu/apple-dart.c index 8d386182ca31..f0de11c97d0b 100644 --- a/drivers/iommu/apple-dart.c +++ b/drivers/iommu/apple-dart.c @@ -196,7 +196,6 @@ struct apple_dart_hw { * @lock: lock for hardware operations involving this dart * @pgsize: pagesize supported by this DART * @supports_bypass: indicates if this DART supports bypass mode - * @force_bypass: force bypass mode due to pagesize mismatch? * @sid2group: maps stream ids to iommu_groups * @iommu: iommu core device */ @@ -217,7 +216,6 @@ struct apple_dart { u32 pgsize; u32 num_streams; u32 supports_bypass : 1; - u32 force_bypass : 1; struct iommu_group *sid2group[DART_MAX_STREAMS]; struct iommu_device iommu; @@ -576,6 +574,9 @@ static int apple_dart_finalize_domain(struct apple_dart_domain *dart_domain, int ret = 0; int i, j; + if (dart->pgsize > PAGE_SIZE) + return -EINVAL; + mutex_lock(&dart_domain->init_lock); if (dart_domain->finalized) @@ -659,9 +660,6 @@ static int apple_dart_attach_dev_paging(struct iommu_domain *domain, struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); struct apple_dart_domain *dart_domain = to_dart_domain(domain); - if (cfg->stream_maps[0].dart->force_bypass) - return -EINVAL; - ret = apple_dart_finalize_domain(dart_domain, cfg); if (ret) return ret; @@ -706,9 +704,6 @@ static int apple_dart_attach_dev_blocked(struct iommu_domain *domain, struct apple_dart_stream_map *stream_map; int i; - if (cfg->stream_maps[0].dart->force_bypass) - return -EINVAL; - for_each_stream_map(i, cfg, stream_map) apple_dart_hw_disable_dma(stream_map); return 0; @@ -802,8 +797,6 @@ static int apple_dart_of_xlate(struct device *dev, struct of_phandle_args *args) if (cfg_dart) { if (cfg_dart->supports_bypass != dart->supports_bypass) return -EINVAL; - if (cfg_dart->force_bypass != dart->force_bypass) - return -EINVAL; if (cfg_dart->pgsize != dart->pgsize) return -EINVAL; } @@ -945,7 +938,7 @@ static int apple_dart_def_domain_type(struct device *dev) { struct apple_dart_master_cfg *cfg = dev_iommu_priv_get(dev); - if (cfg->stream_maps[0].dart->force_bypass) + if (cfg->stream_maps[0].dart->pgsize > PAGE_SIZE) return IOMMU_DOMAIN_IDENTITY; if (!cfg->stream_maps[0].dart->supports_bypass) return IOMMU_DOMAIN_DMA; @@ -1145,8 +1138,6 @@ static int apple_dart_probe(struct platform_device *pdev) goto err_clk_disable; } - dart->force_bypass = dart->pgsize > PAGE_SIZE; - ret = apple_dart_hw_reset(dart); if (ret) goto err_clk_disable; @@ -1170,7 +1161,8 @@ static int apple_dart_probe(struct platform_device *pdev) dev_info( &pdev->dev, "DART [pagesize %x, %d streams, bypass support: %d, bypass forced: %d] initialized\n", - dart->pgsize, dart->num_streams, dart->supports_bypass, dart->force_bypass); + dart->pgsize, dart->num_streams, dart->supports_bypass, + dart->pgsize > PAGE_SIZE); return 0; err_sysfs_remove: -- Gitee From ea610d912e90631d8dd0c00f65dd14274c480fb6 Mon Sep 17 00:00:00 2001 From: gaorui Date: Tue, 10 Jun 2025 20:08:25 +0800 Subject: [PATCH 10/18] iommu/vt-d: Refactor device_to_iommu() to retrieve iommu directly ANBZ: #21698 commit 1903ef8f0d77680e6eb36bd0cc3ea3aa6e95a050 upstream. The device_to_iommu() helper was originally designed to look up the DMAR ACPI table to retrieve the iommu device and the request ID for a given device. However, it was also being used in other places where there was no need to lookup the ACPI table at all. Retrieve the iommu device directly from the per-device iommu private data in functions called after device is probed. Rename the original device_to_iommu() function to a more meaningful name, device_lookup_iommu(), to avoid mis-using it. Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20231116015048.29675-2-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 39 +++++++++++++------------------------ drivers/iommu/intel/iommu.h | 1 - drivers/iommu/intel/svm.c | 20 +++---------------- 3 files changed, 16 insertions(+), 44 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 40fd449274bb..7374c45a3fc4 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -704,7 +704,7 @@ static bool iommu_is_dummy(struct intel_iommu *iommu, struct device *dev) return false; } -struct intel_iommu *device_to_iommu(struct device *dev, u8 *bus, u8 *devfn) +static struct intel_iommu *device_lookup_iommu(struct device *dev, u8 *bus, u8 *devfn) { struct dmar_drhd_unit *drhd = NULL; struct pci_dev *pdev = NULL; @@ -2096,14 +2096,11 @@ static int domain_context_mapping_cb(struct pci_dev *pdev, static int domain_context_mapping(struct dmar_domain *domain, struct device *dev) { + struct device_domain_info *info = dev_iommu_priv_get(dev); struct domain_context_mapping_data data; + struct intel_iommu *iommu = info->iommu; + u8 bus = info->bus, devfn = info->devfn; struct pasid_table *table; - struct intel_iommu *iommu; - u8 bus, devfn; - - iommu = device_to_iommu(dev, &bus, &devfn); - if (!iommu) - return -ENODEV; table = intel_pasid_get_table(dev); @@ -2459,15 +2456,10 @@ static int dmar_domain_attach_device(struct dmar_domain *domain, struct device *dev) { struct device_domain_info *info = dev_iommu_priv_get(dev); - struct intel_iommu *iommu; + struct intel_iommu *iommu = info->iommu; unsigned long flags; - u8 bus, devfn; int ret; - iommu = device_to_iommu(dev, &bus, &devfn); - if (!iommu) - return -ENODEV; - ret = domain_attach_iommu(domain, iommu); if (ret) return ret; @@ -3776,7 +3768,7 @@ static inline int acpi_rmrr_andd_probe(struct device *dev) ret = iommu_probe_device(dev); - iommu = device_to_iommu(dev, &bus, &devfn); + iommu = device_lookup_iommu(dev, &bus, &devfn); if (!iommu) { pr_info("dpoint-- cannot get acpi device corresponding iommu\n"); return -EINVAL; @@ -4136,18 +4128,15 @@ static struct iommu_domain *intel_iommu_domain_alloc(unsigned type) static struct iommu_domain * intel_iommu_domain_alloc_user(struct device *dev, u32 flags) { + struct device_domain_info *info = dev_iommu_priv_get(dev); struct iommu_domain *domain; - struct intel_iommu *iommu; + struct intel_iommu *iommu = info->iommu; bool dirty_tracking; if (flags & (~(IOMMU_HWPT_ALLOC_NEST_PARENT | IOMMU_HWPT_ALLOC_DIRTY_TRACKING))) return ERR_PTR(-EOPNOTSUPP); - iommu = device_to_iommu(dev, NULL, NULL); - if (!iommu) - return ERR_PTR(-ENODEV); - if ((flags & IOMMU_HWPT_ALLOC_NEST_PARENT) && !nested_supported(iommu)) return ERR_PTR(-EOPNOTSUPP); @@ -4184,14 +4173,11 @@ static void intel_iommu_domain_free(struct iommu_domain *domain) static int prepare_domain_attach_device(struct iommu_domain *domain, struct device *dev) { + struct device_domain_info *info = dev_iommu_priv_get(dev); struct dmar_domain *dmar_domain = to_dmar_domain(domain); - struct intel_iommu *iommu; + struct intel_iommu *iommu = info->iommu; int addr_width; - iommu = device_to_iommu(dev, NULL, NULL); - if (!iommu) - return -ENODEV; - if (dmar_domain->force_snooping && !ecap_sc_support(iommu->ecap)) return -EINVAL; @@ -4468,7 +4454,7 @@ static struct iommu_device *intel_iommu_probe_device(struct device *dev) u8 bus, devfn; int ret; - iommu = device_to_iommu(dev, &bus, &devfn); + iommu = device_lookup_iommu(dev, &bus, &devfn); if (!iommu || !iommu->iommu.ops) return ERR_PTR(-ENODEV); @@ -4801,8 +4787,9 @@ static void intel_iommu_iotlb_sync_map(struct iommu_domain *domain, static void intel_iommu_remove_dev_pasid(struct device *dev, ioasid_t pasid) { - struct intel_iommu *iommu = device_to_iommu(dev, NULL, NULL); + struct device_domain_info *info = dev_iommu_priv_get(dev); struct dev_pasid_info *curr, *dev_pasid = NULL; + struct intel_iommu *iommu = info->iommu; struct dmar_domain *dmar_domain; struct iommu_domain *domain; unsigned long flags; diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index 49ea164cb006..14223c606e8a 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -862,7 +862,6 @@ int dmar_ir_support(void); void *alloc_pgtable_page(int node, gfp_t gfp); void free_pgtable_page(void *vaddr); void iommu_flush_write_buffer(struct intel_iommu *iommu); -struct intel_iommu *device_to_iommu(struct device *dev, u8 *bus, u8 *devfn); #ifdef CONFIG_INTEL_IOMMU_SVM void intel_svm_check(struct intel_iommu *iommu); diff --git a/drivers/iommu/intel/svm.c b/drivers/iommu/intel/svm.c index 6010b93c514c..06a442f22bd7 100644 --- a/drivers/iommu/intel/svm.c +++ b/drivers/iommu/intel/svm.c @@ -392,14 +392,9 @@ static int intel_svm_bind_mm(struct intel_iommu *iommu, struct device *dev, void intel_svm_remove_dev_pasid(struct device *dev, u32 pasid) { struct intel_svm_dev *sdev; - struct intel_iommu *iommu; struct intel_svm *svm; struct mm_struct *mm; - iommu = device_to_iommu(dev, NULL, NULL); - if (!iommu) - return; - if (pasid_to_svm_sdev(dev, pasid, &svm, &sdev)) return; mm = svm->mm; @@ -750,25 +745,16 @@ int intel_svm_page_response(struct device *dev, struct iommu_fault_event *evt, struct iommu_page_response *msg) { + struct device_domain_info *info = dev_iommu_priv_get(dev); + struct intel_iommu *iommu = info->iommu; + u8 bus = info->bus, devfn = info->devfn; struct iommu_fault_page_request *prm; - struct intel_iommu *iommu; bool private_present; bool pasid_present; bool last_page; - u8 bus, devfn; int ret = 0; u16 sid; - if (!dev || !dev_is_pci(dev)) - return -ENODEV; - - iommu = device_to_iommu(dev, &bus, &devfn); - if (!iommu) - return -ENODEV; - - if (!msg || !evt) - return -EINVAL; - prm = &evt->fault.prm; sid = PCI_DEVID(bus, devfn); pasid_present = prm->flags & IOMMU_FAULT_PAGE_REQUEST_PASID_VALID; -- Gitee From 917c04666a24bcf28f8c3ba5504a714fe4d7f439 Mon Sep 17 00:00:00 2001 From: gaorui Date: Tue, 10 Jun 2025 20:12:24 +0800 Subject: [PATCH 11/18] iommu/vt-d: Remove unused parameter of intel_pasid_setup_pass_through() ANBZ: #21698 commit 47642bdd5a25cd46228221719af2902b583bd36c upstream. The domain parameter of this helper is unused and can be deleted to avoid dead code. Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20231116015048.29675-3-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 5 ++--- drivers/iommu/intel/pasid.c | 1 - drivers/iommu/intel/pasid.h | 1 - 3 files changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 7374c45a3fc4..1ae5262535de 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -2472,7 +2472,7 @@ static int dmar_domain_attach_device(struct dmar_domain *domain, if (sm_supported(iommu) && !dev_is_real_dma_subdevice(dev)) { /* Setup the PASID entry for requests without PASID: */ if (hw_pass_through && domain_type_is_si(domain)) - ret = intel_pasid_setup_pass_through(iommu, domain, + ret = intel_pasid_setup_pass_through(iommu, dev, IOMMU_NO_PASID); else if (domain->use_first_level) ret = domain_setup_first_level(iommu, domain, dev, @@ -4859,8 +4859,7 @@ static int intel_iommu_set_dev_pasid(struct iommu_domain *domain, goto out_free; if (domain_type_is_si(dmar_domain)) - ret = intel_pasid_setup_pass_through(iommu, dmar_domain, - dev, pasid); + ret = intel_pasid_setup_pass_through(iommu, dev, pasid); else if (dmar_domain->use_first_level) ret = domain_setup_first_level(iommu, dmar_domain, dev, pasid); diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index 06ea2dd53542..b49eceded041 100644 --- a/drivers/iommu/intel/pasid.c +++ b/drivers/iommu/intel/pasid.c @@ -752,7 +752,6 @@ int intel_pasid_setup_dirty_tracking(struct intel_iommu *iommu, * Set up the scalable mode pasid entry for passthrough translation type. */ int intel_pasid_setup_pass_through(struct intel_iommu *iommu, - struct dmar_domain *domain, struct device *dev, u32 pasid) { u16 did = FLPT_DEFAULT_DID; diff --git a/drivers/iommu/intel/pasid.h b/drivers/iommu/intel/pasid.h index 958050b093aa..55c3e28f0b7a 100644 --- a/drivers/iommu/intel/pasid.h +++ b/drivers/iommu/intel/pasid.h @@ -111,7 +111,6 @@ int intel_pasid_setup_dirty_tracking(struct intel_iommu *iommu, struct device *dev, u32 pasid, bool enabled); int intel_pasid_setup_pass_through(struct intel_iommu *iommu, - struct dmar_domain *domain, struct device *dev, u32 pasid); void intel_pasid_tear_down_entry(struct intel_iommu *iommu, struct device *dev, u32 pasid, -- Gitee From 465b96cccb8bb1d131f24bc1a864debb99febfa9 Mon Sep 17 00:00:00 2001 From: gaorui Date: Tue, 10 Jun 2025 20:17:07 +0800 Subject: [PATCH 12/18] iommu/vt-d: Remove unused vcmd interfaces ANBZ: #21698 commit d2b66903464ec29e54ac0991fa43c0da2c2a4892 upstream. Commit 99b5726b4423 ("iommu: Remove ioasid infrastructure") has removed ioasid allocation interfaces from the iommu subsystem. As a result, these vcmd interfaces have become obsolete. Remove them to avoid dead code. Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20231116015048.29675-4-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/debugfs.c | 3 -- drivers/iommu/intel/iommu.h | 3 -- drivers/iommu/intel/pasid.c | 57 ----------------------------------- drivers/iommu/intel/pasid.h | 12 -------- 4 files changed, 75 deletions(-) diff --git a/drivers/iommu/intel/debugfs.c b/drivers/iommu/intel/debugfs.c index 1f925285104e..8ab7b3fbe3b1 100644 --- a/drivers/iommu/intel/debugfs.c +++ b/drivers/iommu/intel/debugfs.c @@ -106,9 +106,6 @@ static const struct iommu_regset iommu_regs_64[] = { IOMMU_REGSET_ENTRY(MTRR_PHYSMASK8), IOMMU_REGSET_ENTRY(MTRR_PHYSBASE9), IOMMU_REGSET_ENTRY(MTRR_PHYSMASK9), - IOMMU_REGSET_ENTRY(VCCAP), - IOMMU_REGSET_ENTRY(VCMD), - IOMMU_REGSET_ENTRY(VCRSP), }; static int iommu_regset_show(struct seq_file *m, void *unused) diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index 14223c606e8a..2a8dd87c3b4f 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -139,9 +139,6 @@ #define DMAR_ECEO_REG 0x408 #define DMAR_ECRSP_REG 0x410 #define DMAR_ECCAP_REG 0x430 -#define DMAR_VCCAP_REG 0xe30 /* Virtual command capability register */ -#define DMAR_VCMD_REG 0xe00 /* Virtual command register */ -#define DMAR_VCRSP_REG 0xe10 /* Virtual command response register */ #define DMAR_IQER_REG_IQEI(reg) FIELD_GET(GENMASK_ULL(3, 0), reg) #define DMAR_IQER_REG_ITESID(reg) FIELD_GET(GENMASK_ULL(47, 32), reg) diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index b49eceded041..0bf3c0565f23 100644 --- a/drivers/iommu/intel/pasid.c +++ b/drivers/iommu/intel/pasid.c @@ -26,63 +26,6 @@ */ u32 intel_pasid_max_id = PASID_MAX; -int vcmd_alloc_pasid(struct intel_iommu *iommu, u32 *pasid) -{ - unsigned long flags; - u8 status_code; - int ret = 0; - u64 res; - - raw_spin_lock_irqsave(&iommu->register_lock, flags); - dmar_writeq(iommu->reg + DMAR_VCMD_REG, VCMD_CMD_ALLOC); - IOMMU_WAIT_OP(iommu, DMAR_VCRSP_REG, dmar_readq, - !(res & VCMD_VRSP_IP), res); - raw_spin_unlock_irqrestore(&iommu->register_lock, flags); - - status_code = VCMD_VRSP_SC(res); - switch (status_code) { - case VCMD_VRSP_SC_SUCCESS: - *pasid = VCMD_VRSP_RESULT_PASID(res); - break; - case VCMD_VRSP_SC_NO_PASID_AVAIL: - pr_info("IOMMU: %s: No PASID available\n", iommu->name); - ret = -ENOSPC; - break; - default: - ret = -ENODEV; - pr_warn("IOMMU: %s: Unexpected error code %d\n", - iommu->name, status_code); - } - - return ret; -} - -void vcmd_free_pasid(struct intel_iommu *iommu, u32 pasid) -{ - unsigned long flags; - u8 status_code; - u64 res; - - raw_spin_lock_irqsave(&iommu->register_lock, flags); - dmar_writeq(iommu->reg + DMAR_VCMD_REG, - VCMD_CMD_OPERAND(pasid) | VCMD_CMD_FREE); - IOMMU_WAIT_OP(iommu, DMAR_VCRSP_REG, dmar_readq, - !(res & VCMD_VRSP_IP), res); - raw_spin_unlock_irqrestore(&iommu->register_lock, flags); - - status_code = VCMD_VRSP_SC(res); - switch (status_code) { - case VCMD_VRSP_SC_SUCCESS: - break; - case VCMD_VRSP_SC_INVALID_PASID: - pr_info("IOMMU: %s: Invalid PASID\n", iommu->name); - break; - default: - pr_warn("IOMMU: %s: Unexpected error code %d\n", - iommu->name, status_code); - } -} - /* * Per device pasid table management: */ diff --git a/drivers/iommu/intel/pasid.h b/drivers/iommu/intel/pasid.h index 55c3e28f0b7a..16e559b03f64 100644 --- a/drivers/iommu/intel/pasid.h +++ b/drivers/iommu/intel/pasid.h @@ -22,16 +22,6 @@ #define is_pasid_enabled(entry) (((entry)->lo >> 3) & 0x1) #define get_pasid_dir_size(entry) (1 << ((((entry)->lo >> 9) & 0x7) + 7)) -/* Virtual command interface for enlightened pasid management. */ -#define VCMD_CMD_ALLOC 0x1 -#define VCMD_CMD_FREE 0x2 -#define VCMD_VRSP_IP 0x1 -#define VCMD_VRSP_SC(e) (((e) & 0xff) >> 1) -#define VCMD_VRSP_SC_SUCCESS 0 -#define VCMD_VRSP_SC_NO_PASID_AVAIL 16 -#define VCMD_VRSP_SC_INVALID_PASID 16 -#define VCMD_VRSP_RESULT_PASID(e) (((e) >> 16) & 0xfffff) -#define VCMD_CMD_OPERAND(e) ((e) << 16) /* * Domain ID reserved for pasid entries programmed for first-level * only and pass-through transfer modes. @@ -115,8 +105,6 @@ int intel_pasid_setup_pass_through(struct intel_iommu *iommu, void intel_pasid_tear_down_entry(struct intel_iommu *iommu, struct device *dev, u32 pasid, bool fault_ignore); -int vcmd_alloc_pasid(struct intel_iommu *iommu, u32 *pasid); -void vcmd_free_pasid(struct intel_iommu *iommu, u32 pasid); void intel_pasid_setup_page_snoop_control(struct intel_iommu *iommu, struct device *dev, u32 pasid); #endif /* __INTEL_PASID_H */ -- Gitee From ed602ca0bbbaca918843cce17d85ab16a2614cfd Mon Sep 17 00:00:00 2001 From: gaorui Date: Tue, 10 Jun 2025 20:40:59 +0800 Subject: [PATCH 13/18] iommu/vt-d: Move inline helpers to header files ANBZ: #21698 commit 80b79e141da7251d6ea82573f058ee9418896465 upstream. Move inline helpers to header files so that other files can use them without duplicating the code. Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20231116015048.29675-5-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 204 +++-------------------------------- drivers/iommu/intel/iommu.h | 175 ++++++++++++++++++++++++++++++ drivers/iommu/intel/pasid.c | 198 +--------------------------------- drivers/iommu/intel/pasid.h | 210 ++++++++++++++++++++++++++++++++++++ 4 files changed, 400 insertions(+), 387 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 1ae5262535de..5e92543a390a 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -46,9 +46,6 @@ #define DEFAULT_DOMAIN_ADDRESS_WIDTH 57 -#define MAX_AGAW_WIDTH 64 -#define MAX_AGAW_PFN_WIDTH (MAX_AGAW_WIDTH - VTD_PAGE_SHIFT) - #define __DOMAIN_MAX_PFN(gaw) ((((uint64_t)1) << ((gaw) - VTD_PAGE_SHIFT)) - 1) #define __DOMAIN_MAX_ADDR(gaw) ((((uint64_t)1) << (gaw)) - 1) @@ -63,74 +60,6 @@ #define IOVA_PFN(addr) ((addr) >> PAGE_SHIFT) -/* page table handling */ -#define LEVEL_STRIDE (9) -#define LEVEL_MASK (((u64)1 << LEVEL_STRIDE) - 1) - -static inline int agaw_to_level(int agaw) -{ - return agaw + 2; -} - -static inline int agaw_to_width(int agaw) -{ - return min_t(int, 30 + agaw * LEVEL_STRIDE, MAX_AGAW_WIDTH); -} - -static inline int width_to_agaw(int width) -{ - return DIV_ROUND_UP(width - 30, LEVEL_STRIDE); -} - -static inline unsigned int level_to_offset_bits(int level) -{ - return (level - 1) * LEVEL_STRIDE; -} - -static inline int pfn_level_offset(u64 pfn, int level) -{ - return (pfn >> level_to_offset_bits(level)) & LEVEL_MASK; -} - -static inline u64 level_mask(int level) -{ - return -1ULL << level_to_offset_bits(level); -} - -static inline u64 level_size(int level) -{ - return 1ULL << level_to_offset_bits(level); -} - -static inline u64 align_to_level(u64 pfn, int level) -{ - return (pfn + level_size(level) - 1) & level_mask(level); -} - -static inline unsigned long lvl_to_nr_pages(unsigned int lvl) -{ - return 1UL << min_t(int, (lvl - 1) * LEVEL_STRIDE, MAX_AGAW_PFN_WIDTH); -} - -/* VT-d pages must always be _smaller_ than MM pages. Otherwise things - are never going to work. */ -static inline unsigned long mm_to_dma_pfn_start(unsigned long mm_pfn) -{ - return mm_pfn << (PAGE_SHIFT - VTD_PAGE_SHIFT); -} -static inline unsigned long mm_to_dma_pfn_end(unsigned long mm_pfn) -{ - return ((mm_pfn + 1) << (PAGE_SHIFT - VTD_PAGE_SHIFT)) - 1; -} -static inline unsigned long page_to_dma_pfn(struct page *pg) -{ - return mm_to_dma_pfn_start(page_to_pfn(pg)); -} -static inline unsigned long virt_to_dma_pfn(void *p) -{ - return page_to_dma_pfn(virt_to_page(p)); -} - static void __init check_tylersburg_isoch(void); static int rwbf_quirk; @@ -168,78 +97,6 @@ static phys_addr_t root_entry_uctp(struct root_entry *re) return re->hi & VTD_PAGE_MASK; } -static inline void context_set_present(struct context_entry *context) -{ - context->lo |= 1; -} - -static inline void context_set_fault_enable(struct context_entry *context) -{ - context->lo &= (((u64)-1) << 2) | 1; -} - -static inline void context_set_translation_type(struct context_entry *context, - unsigned long value) -{ - context->lo &= (((u64)-1) << 4) | 3; - context->lo |= (value & 3) << 2; -} - -static inline void context_set_address_root(struct context_entry *context, - unsigned long value) -{ - context->lo &= ~VTD_PAGE_MASK; - context->lo |= value & VTD_PAGE_MASK; -} - -static inline void context_set_address_width(struct context_entry *context, - unsigned long value) -{ - context->hi |= value & 7; -} - -static inline void context_set_domain_id(struct context_entry *context, - unsigned long value) -{ - context->hi |= (value & ((1 << 16) - 1)) << 8; -} - -static inline void context_set_pasid(struct context_entry *context) -{ - context->lo |= CONTEXT_PASIDE; -} - -static inline int context_domain_id(struct context_entry *c) -{ - return((c->hi >> 8) & 0xffff); -} - -static inline void context_clear_entry(struct context_entry *context) -{ - context->lo = 0; - context->hi = 0; -} - -static inline bool context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) -{ - if (!iommu->copied_tables) - return false; - - return test_bit(((long)bus << 8) | devfn, iommu->copied_tables); -} - -static inline void -set_context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) -{ - set_bit(((long)bus << 8) | devfn, iommu->copied_tables); -} - -static inline void -clear_context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) -{ - clear_bit(((long)bus << 8) | devfn, iommu->copied_tables); -} - /* * This domain is a statically identity mapping domain. * 1. This domain creats a static 1:1 mapping to all usable memory. @@ -384,13 +241,12 @@ void free_pgtable_page(void *vaddr) free_page((unsigned long)vaddr); } -static inline int domain_type_is_si(struct dmar_domain *domain) +static int domain_type_is_si(struct dmar_domain *domain) { return domain->domain.type == IOMMU_DOMAIN_IDENTITY; } -static inline int domain_pfn_supported(struct dmar_domain *domain, - unsigned long pfn) +static int domain_pfn_supported(struct dmar_domain *domain, unsigned long pfn) { int addr_width = agaw_to_width(domain->agaw) - VTD_PAGE_SHIFT; @@ -452,7 +308,7 @@ int iommu_calculate_agaw(struct intel_iommu *iommu) return __iommu_calculate_agaw(iommu, DEFAULT_DOMAIN_ADDRESS_WIDTH); } -static inline bool iommu_paging_structure_coherency(struct intel_iommu *iommu) +static bool iommu_paging_structure_coherency(struct intel_iommu *iommu) { return sm_supported(iommu) ? ecap_smpwc(iommu->ecap) : ecap_coherent(iommu->ecap); @@ -1587,9 +1443,8 @@ static void iommu_flush_iotlb_psi(struct intel_iommu *iommu, } /* Notification for newly created mappings */ -static inline void __mapping_notify_one(struct intel_iommu *iommu, - struct dmar_domain *domain, - unsigned long pfn, unsigned int pages) +static void __mapping_notify_one(struct intel_iommu *iommu, struct dmar_domain *domain, + unsigned long pfn, unsigned int pages) { /* * It's a non-present to present mapping. Only flush if caching mode @@ -1858,7 +1713,7 @@ static void domain_detach_iommu(struct dmar_domain *domain, spin_unlock(&iommu->lock); } -static inline int guestwidth_to_adjustwidth(int gaw) +static int guestwidth_to_adjustwidth(int gaw) { int agaw; int r = (gaw - 12) % 9; @@ -1892,7 +1747,7 @@ static void domain_exit(struct dmar_domain *domain) * Value of X in the PDTS field of a scalable mode context entry * indicates PASID directory with 2^(X + 7) entries. */ -static inline unsigned long context_get_sm_pds(struct pasid_table *table) +static unsigned long context_get_sm_pds(struct pasid_table *table) { unsigned long pds, max_pde; @@ -1904,38 +1759,6 @@ static inline unsigned long context_get_sm_pds(struct pasid_table *table) return pds - 7; } -/* - * Set the RID_PASID field of a scalable mode context entry. The - * IOMMU hardware will use the PASID value set in this field for - * DMA translations of DMA requests without PASID. - */ -static inline void -context_set_sm_rid2pasid(struct context_entry *context, unsigned long pasid) -{ - context->hi |= pasid & ((1 << 20) - 1); -} - -/* - * Set the DTE(Device-TLB Enable) field of a scalable mode context - * entry. - */ -static inline void context_set_sm_dte(struct context_entry *context) -{ - context->lo |= BIT_ULL(2); -} - -/* - * Set the PRE(Page Request Enable) field of a scalable mode context - * entry. - */ -static inline void context_set_sm_pre(struct context_entry *context) -{ - context->lo |= BIT_ULL(4); -} - -/* Convert value to context PASID directory size field coding. */ -#define context_pdts(pds) (((pds) & 0x7) << 9) - static int domain_context_mapping_one(struct dmar_domain *domain, struct intel_iommu *iommu, struct pasid_table *table, @@ -2117,18 +1940,15 @@ domain_context_mapping(struct dmar_domain *domain, struct device *dev) } /* Returns a number of VTD pages, but aligned to MM page size */ -static inline unsigned long aligned_nrpages(unsigned long host_addr, - size_t size) +static unsigned long aligned_nrpages(unsigned long host_addr, size_t size) { host_addr &= ~PAGE_MASK; return PAGE_ALIGN(host_addr + size) >> VTD_PAGE_SHIFT; } /* Return largest possible superpage level for a given mapping */ -static inline int hardware_largepage_caps(struct dmar_domain *domain, - unsigned long iov_pfn, - unsigned long phy_pfn, - unsigned long pages) +static int hardware_largepage_caps(struct dmar_domain *domain, unsigned long iov_pfn, + unsigned long phy_pfn, unsigned long pages) { int support, level = 1; unsigned long pfnmerge; @@ -3635,7 +3455,7 @@ void intel_iommu_shutdown(void) up_write(&dmar_global_lock); } -static inline struct intel_iommu *dev_to_intel_iommu(struct device *dev) +static struct intel_iommu *dev_to_intel_iommu(struct device *dev) { struct iommu_device *iommu_dev = dev_to_iommu_device(dev); @@ -3714,7 +3534,7 @@ const struct attribute_group *intel_iommu_groups[] = { NULL, }; -static inline bool has_external_pci(void) +static bool has_external_pci(void) { struct pci_dev *pdev = NULL; diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index 2a8dd87c3b4f..dc5a07b0c2d9 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -820,6 +820,181 @@ static inline bool context_present(struct context_entry *context) return (context->lo & 1); } +#define LEVEL_STRIDE (9) +#define LEVEL_MASK (((u64)1 << LEVEL_STRIDE) - 1) +#define MAX_AGAW_WIDTH (64) +#define MAX_AGAW_PFN_WIDTH (MAX_AGAW_WIDTH - VTD_PAGE_SHIFT) + +static inline int agaw_to_level(int agaw) +{ + return agaw + 2; +} + +static inline int agaw_to_width(int agaw) +{ + return min_t(int, 30 + agaw * LEVEL_STRIDE, MAX_AGAW_WIDTH); +} + +static inline int width_to_agaw(int width) +{ + return DIV_ROUND_UP(width - 30, LEVEL_STRIDE); +} + +static inline unsigned int level_to_offset_bits(int level) +{ + return (level - 1) * LEVEL_STRIDE; +} + +static inline int pfn_level_offset(u64 pfn, int level) +{ + return (pfn >> level_to_offset_bits(level)) & LEVEL_MASK; +} + +static inline u64 level_mask(int level) +{ + return -1ULL << level_to_offset_bits(level); +} + +static inline u64 level_size(int level) +{ + return 1ULL << level_to_offset_bits(level); +} + +static inline u64 align_to_level(u64 pfn, int level) +{ + return (pfn + level_size(level) - 1) & level_mask(level); +} + +static inline unsigned long lvl_to_nr_pages(unsigned int lvl) +{ + return 1UL << min_t(int, (lvl - 1) * LEVEL_STRIDE, MAX_AGAW_PFN_WIDTH); +} + +/* VT-d pages must always be _smaller_ than MM pages. Otherwise things + are never going to work. */ +static inline unsigned long mm_to_dma_pfn_start(unsigned long mm_pfn) +{ + return mm_pfn << (PAGE_SHIFT - VTD_PAGE_SHIFT); +} +static inline unsigned long mm_to_dma_pfn_end(unsigned long mm_pfn) +{ + return ((mm_pfn + 1) << (PAGE_SHIFT - VTD_PAGE_SHIFT)) - 1; +} +static inline unsigned long page_to_dma_pfn(struct page *pg) +{ + return mm_to_dma_pfn_start(page_to_pfn(pg)); +} +static inline unsigned long virt_to_dma_pfn(void *p) +{ + return page_to_dma_pfn(virt_to_page(p)); +} + +static inline void context_set_present(struct context_entry *context) +{ + context->lo |= 1; +} + +static inline void context_set_fault_enable(struct context_entry *context) +{ + context->lo &= (((u64)-1) << 2) | 1; +} + +static inline void context_set_translation_type(struct context_entry *context, + unsigned long value) +{ + context->lo &= (((u64)-1) << 4) | 3; + context->lo |= (value & 3) << 2; +} + +static inline void context_set_address_root(struct context_entry *context, + unsigned long value) +{ + context->lo &= ~VTD_PAGE_MASK; + context->lo |= value & VTD_PAGE_MASK; +} + +static inline void context_set_address_width(struct context_entry *context, + unsigned long value) +{ + context->hi |= value & 7; +} + +static inline void context_set_domain_id(struct context_entry *context, + unsigned long value) +{ + context->hi |= (value & ((1 << 16) - 1)) << 8; +} + +static inline void context_set_pasid(struct context_entry *context) +{ + context->lo |= CONTEXT_PASIDE; +} + +static inline int context_domain_id(struct context_entry *c) +{ + return((c->hi >> 8) & 0xffff); +} + +static inline void context_clear_entry(struct context_entry *context) +{ + context->lo = 0; + context->hi = 0; +} + +#ifdef CONFIG_INTEL_IOMMU +static inline bool context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) +{ + if (!iommu->copied_tables) + return false; + + return test_bit(((long)bus << 8) | devfn, iommu->copied_tables); +} + +static inline void +set_context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) +{ + set_bit(((long)bus << 8) | devfn, iommu->copied_tables); +} + +static inline void +clear_context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) +{ + clear_bit(((long)bus << 8) | devfn, iommu->copied_tables); +} +#endif /* CONFIG_INTEL_IOMMU */ + +/* + * Set the RID_PASID field of a scalable mode context entry. The + * IOMMU hardware will use the PASID value set in this field for + * DMA translations of DMA requests without PASID. + */ +static inline void +context_set_sm_rid2pasid(struct context_entry *context, unsigned long pasid) +{ + context->hi |= pasid & ((1 << 20) - 1); +} + +/* + * Set the DTE(Device-TLB Enable) field of a scalable mode context + * entry. + */ +static inline void context_set_sm_dte(struct context_entry *context) +{ + context->lo |= BIT_ULL(2); +} + +/* + * Set the PRE(Page Request Enable) field of a scalable mode context + * entry. + */ +static inline void context_set_sm_pre(struct context_entry *context) +{ + context->lo |= BIT_ULL(4); +} + +/* Convert value to context PASID directory size field coding. */ +#define context_pdts(pds) (((pds) & 0x7) << 9) + struct dmar_drhd_unit *dmar_find_matched_drhd_unit(struct pci_dev *dev); int dmar_enable_qi(struct intel_iommu *iommu); diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index 0bf3c0565f23..066f6837a07c 100644 --- a/drivers/iommu/intel/pasid.c +++ b/drivers/iommu/intel/pasid.c @@ -173,30 +173,6 @@ static struct pasid_entry *intel_pasid_get_entry(struct device *dev, u32 pasid) /* * Interfaces for PASID table entry manipulation: */ -static inline void pasid_clear_entry(struct pasid_entry *pe) -{ - WRITE_ONCE(pe->val[0], 0); - WRITE_ONCE(pe->val[1], 0); - WRITE_ONCE(pe->val[2], 0); - WRITE_ONCE(pe->val[3], 0); - WRITE_ONCE(pe->val[4], 0); - WRITE_ONCE(pe->val[5], 0); - WRITE_ONCE(pe->val[6], 0); - WRITE_ONCE(pe->val[7], 0); -} - -static inline void pasid_clear_entry_with_fpd(struct pasid_entry *pe) -{ - WRITE_ONCE(pe->val[0], PASID_PTE_FPD); - WRITE_ONCE(pe->val[1], 0); - WRITE_ONCE(pe->val[2], 0); - WRITE_ONCE(pe->val[3], 0); - WRITE_ONCE(pe->val[4], 0); - WRITE_ONCE(pe->val[5], 0); - WRITE_ONCE(pe->val[6], 0); - WRITE_ONCE(pe->val[7], 0); -} - static void intel_pasid_clear_entry(struct device *dev, u32 pasid, bool fault_ignore) { @@ -212,174 +188,6 @@ intel_pasid_clear_entry(struct device *dev, u32 pasid, bool fault_ignore) pasid_clear_entry(pe); } -static inline void pasid_set_bits(u64 *ptr, u64 mask, u64 bits) -{ - u64 old; - - old = READ_ONCE(*ptr); - WRITE_ONCE(*ptr, (old & ~mask) | bits); -} - -static inline u64 pasid_get_bits(u64 *ptr) -{ - return READ_ONCE(*ptr); -} - -/* - * Setup the DID(Domain Identifier) field (Bit 64~79) of scalable mode - * PASID entry. - */ -static inline void -pasid_set_domain_id(struct pasid_entry *pe, u64 value) -{ - pasid_set_bits(&pe->val[1], GENMASK_ULL(15, 0), value); -} - -/* - * Get domain ID value of a scalable mode PASID entry. - */ -static inline u16 -pasid_get_domain_id(struct pasid_entry *pe) -{ - return (u16)(READ_ONCE(pe->val[1]) & GENMASK_ULL(15, 0)); -} - -/* - * Setup the SLPTPTR(Second Level Page Table Pointer) field (Bit 12~63) - * of a scalable mode PASID entry. - */ -static inline void -pasid_set_slptr(struct pasid_entry *pe, u64 value) -{ - pasid_set_bits(&pe->val[0], VTD_PAGE_MASK, value); -} - -/* - * Setup the AW(Address Width) field (Bit 2~4) of a scalable mode PASID - * entry. - */ -static inline void -pasid_set_address_width(struct pasid_entry *pe, u64 value) -{ - pasid_set_bits(&pe->val[0], GENMASK_ULL(4, 2), value << 2); -} - -/* - * Setup the PGTT(PASID Granular Translation Type) field (Bit 6~8) - * of a scalable mode PASID entry. - */ -static inline void -pasid_set_translation_type(struct pasid_entry *pe, u64 value) -{ - pasid_set_bits(&pe->val[0], GENMASK_ULL(8, 6), value << 6); -} - -/* - * Enable fault processing by clearing the FPD(Fault Processing - * Disable) field (Bit 1) of a scalable mode PASID entry. - */ -static inline void pasid_set_fault_enable(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[0], 1 << 1, 0); -} - -/* - * Enable second level A/D bits by setting the SLADE (Second Level - * Access Dirty Enable) field (Bit 9) of a scalable mode PASID - * entry. - */ -static inline void pasid_set_ssade(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[0], 1 << 9, 1 << 9); -} - -/* - * Disable second level A/D bits by clearing the SLADE (Second Level - * Access Dirty Enable) field (Bit 9) of a scalable mode PASID - * entry. - */ -static inline void pasid_clear_ssade(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[0], 1 << 9, 0); -} - -/* - * Checks if second level A/D bits specifically the SLADE (Second Level - * Access Dirty Enable) field (Bit 9) of a scalable mode PASID - * entry is set. - */ -static inline bool pasid_get_ssade(struct pasid_entry *pe) -{ - return pasid_get_bits(&pe->val[0]) & (1 << 9); -} - -/* - * Setup the WPE(Write Protect Enable) field (Bit 132) of a - * scalable mode PASID entry. - */ -static inline void pasid_set_wpe(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[2], 1 << 4, 1 << 4); -} - -/* - * Setup the P(Present) field (Bit 0) of a scalable mode PASID - * entry. - */ -static inline void pasid_set_present(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[0], 1 << 0, 1); -} - -/* - * Setup Page Walk Snoop bit (Bit 87) of a scalable mode PASID - * entry. - */ -static inline void pasid_set_page_snoop(struct pasid_entry *pe, bool value) -{ - pasid_set_bits(&pe->val[1], 1 << 23, value << 23); -} - -/* - * Setup No Execute Enable bit (Bit 133) of a scalable mode PASID - * entry. It is required when XD bit of the first level page table - * entry is about to be set. - */ -static inline void pasid_set_nxe(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[2], 1 << 5, 1 << 5); -} - -/* - * Setup the Page Snoop (PGSNP) field (Bit 88) of a scalable mode - * PASID entry. - */ -static inline void -pasid_set_pgsnp(struct pasid_entry *pe) -{ - pasid_set_bits(&pe->val[1], 1ULL << 24, 1ULL << 24); -} - -/* - * Setup the First Level Page table Pointer field (Bit 140~191) - * of a scalable mode PASID entry. - */ -static inline void -pasid_set_flptr(struct pasid_entry *pe, u64 value) -{ - pasid_set_bits(&pe->val[2], VTD_PAGE_MASK, value); -} - -/* - * Setup the First Level Paging Mode field (Bit 130~131) of a - * scalable mode PASID entry. - */ -static inline void -pasid_set_flpm(struct pasid_entry *pe, u64 value) -{ - pasid_set_bits(&pe->val[2], GENMASK_ULL(3, 2), value << 2); -} - static void pasid_cache_invalidation_with_pasid(struct intel_iommu *iommu, u16 did, u32 pasid) @@ -541,9 +349,9 @@ int intel_pasid_setup_first_level(struct intel_iommu *iommu, * Skip top levels of page tables for iommu which has less agaw * than default. Unnecessary for PT mode. */ -static inline int iommu_skip_agaw(struct dmar_domain *domain, - struct intel_iommu *iommu, - struct dma_pte **pgd) +static int iommu_skip_agaw(struct dmar_domain *domain, + struct intel_iommu *iommu, + struct dma_pte **pgd) { int agaw; diff --git a/drivers/iommu/intel/pasid.h b/drivers/iommu/intel/pasid.h index 16e559b03f64..c5a6e03b7b9e 100644 --- a/drivers/iommu/intel/pasid.h +++ b/drivers/iommu/intel/pasid.h @@ -86,6 +86,216 @@ static inline u16 pasid_pte_get_pgtt(struct pasid_entry *pte) return (u16)((READ_ONCE(pte->val[0]) >> 6) & 0x7); } +static inline void pasid_clear_entry(struct pasid_entry *pe) +{ + WRITE_ONCE(pe->val[0], 0); + WRITE_ONCE(pe->val[1], 0); + WRITE_ONCE(pe->val[2], 0); + WRITE_ONCE(pe->val[3], 0); + WRITE_ONCE(pe->val[4], 0); + WRITE_ONCE(pe->val[5], 0); + WRITE_ONCE(pe->val[6], 0); + WRITE_ONCE(pe->val[7], 0); +} + +static inline void pasid_clear_entry_with_fpd(struct pasid_entry *pe) +{ + WRITE_ONCE(pe->val[0], PASID_PTE_FPD); + WRITE_ONCE(pe->val[1], 0); + WRITE_ONCE(pe->val[2], 0); + WRITE_ONCE(pe->val[3], 0); + WRITE_ONCE(pe->val[4], 0); + WRITE_ONCE(pe->val[5], 0); + WRITE_ONCE(pe->val[6], 0); + WRITE_ONCE(pe->val[7], 0); +} + +static inline void pasid_set_bits(u64 *ptr, u64 mask, u64 bits) +{ + u64 old; + + old = READ_ONCE(*ptr); + WRITE_ONCE(*ptr, (old & ~mask) | bits); +} + +static inline u64 pasid_get_bits(u64 *ptr) +{ + return READ_ONCE(*ptr); +} + +/* + * Setup the DID(Domain Identifier) field (Bit 64~79) of scalable mode + * PASID entry. + */ +static inline void +pasid_set_domain_id(struct pasid_entry *pe, u64 value) +{ + pasid_set_bits(&pe->val[1], GENMASK_ULL(15, 0), value); +} + +/* + * Get domain ID value of a scalable mode PASID entry. + */ +static inline u16 +pasid_get_domain_id(struct pasid_entry *pe) +{ + return (u16)(READ_ONCE(pe->val[1]) & GENMASK_ULL(15, 0)); +} + +/* + * Setup the SLPTPTR(Second Level Page Table Pointer) field (Bit 12~63) + * of a scalable mode PASID entry. + */ +static inline void +pasid_set_slptr(struct pasid_entry *pe, u64 value) +{ + pasid_set_bits(&pe->val[0], VTD_PAGE_MASK, value); +} + +/* + * Setup the AW(Address Width) field (Bit 2~4) of a scalable mode PASID + * entry. + */ +static inline void +pasid_set_address_width(struct pasid_entry *pe, u64 value) +{ + pasid_set_bits(&pe->val[0], GENMASK_ULL(4, 2), value << 2); +} + +/* + * Setup the PGTT(PASID Granular Translation Type) field (Bit 6~8) + * of a scalable mode PASID entry. + */ +static inline void +pasid_set_translation_type(struct pasid_entry *pe, u64 value) +{ + pasid_set_bits(&pe->val[0], GENMASK_ULL(8, 6), value << 6); +} + +/* + * Enable fault processing by clearing the FPD(Fault Processing + * Disable) field (Bit 1) of a scalable mode PASID entry. + */ +static inline void pasid_set_fault_enable(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[0], 1 << 1, 0); +} + +/* + * Enable second level A/D bits by setting the SLADE (Second Level + * Access Dirty Enable) field (Bit 9) of a scalable mode PASID + * entry. + */ +static inline void pasid_set_ssade(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[0], 1 << 9, 1 << 9); +} + +/* + * Disable second level A/D bits by clearing the SLADE (Second Level + * Access Dirty Enable) field (Bit 9) of a scalable mode PASID + * entry. + */ +static inline void pasid_clear_ssade(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[0], 1 << 9, 0); +} + +/* + * Checks if second level A/D bits specifically the SLADE (Second Level + * Access Dirty Enable) field (Bit 9) of a scalable mode PASID + * entry is set. + */ +static inline bool pasid_get_ssade(struct pasid_entry *pe) +{ + return pasid_get_bits(&pe->val[0]) & (1 << 9); +} + +/* + * Setup the SRE(Supervisor Request Enable) field (Bit 128) of a + * scalable mode PASID entry. + */ +static inline void pasid_set_sre(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[2], 1 << 0, 1); +} + +/* + * Setup the WPE(Write Protect Enable) field (Bit 132) of a + * scalable mode PASID entry. + */ +static inline void pasid_set_wpe(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[2], 1 << 4, 1 << 4); +} + +/* + * Setup the P(Present) field (Bit 0) of a scalable mode PASID + * entry. + */ +static inline void pasid_set_present(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[0], 1 << 0, 1); +} + +/* + * Setup Page Walk Snoop bit (Bit 87) of a scalable mode PASID + * entry. + */ +static inline void pasid_set_page_snoop(struct pasid_entry *pe, bool value) +{ + pasid_set_bits(&pe->val[1], 1 << 23, value << 23); +} + +/* + * Setup No Execute Enable bit (Bit 133) of a scalable mode PASID + * entry. It is required when XD bit of the first level page table + * entry is about to be set. + */ +static inline void pasid_set_nxe(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[2], 1 << 5, 1 << 5); +} + +/* + * Setup the Page Snoop (PGSNP) field (Bit 88) of a scalable mode + * PASID entry. + */ +static inline void +pasid_set_pgsnp(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[1], 1ULL << 24, 1ULL << 24); +} + +/* + * Setup the First Level Page table Pointer field (Bit 140~191) + * of a scalable mode PASID entry. + */ +static inline void +pasid_set_flptr(struct pasid_entry *pe, u64 value) +{ + pasid_set_bits(&pe->val[2], VTD_PAGE_MASK, value); +} + +/* + * Setup the First Level Paging Mode field (Bit 130~131) of a + * scalable mode PASID entry. + */ +static inline void +pasid_set_flpm(struct pasid_entry *pe, u64 value) +{ + pasid_set_bits(&pe->val[2], GENMASK_ULL(3, 2), value << 2); +} + +/* + * Setup the Extended Access Flag Enable (EAFE) field (Bit 135) + * of a scalable mode PASID entry. + */ +static inline void pasid_set_eafe(struct pasid_entry *pe) +{ + pasid_set_bits(&pe->val[2], 1 << 7, 1 << 7); +} + extern unsigned int intel_pasid_max_id; int intel_pasid_alloc_table(struct device *dev); void intel_pasid_free_table(struct device *dev); -- Gitee From 297a71c01c0f6725d31f03a1188a3d33a8154b5b Mon Sep 17 00:00:00 2001 From: gaorui Date: Wed, 11 Jun 2025 20:02:33 +0800 Subject: [PATCH 14/18] iommu: Add static iommu_ops->release_domain ANBZ: #21698 commit 0061ffe289e19caabeea8103e69cb0f1896e34d8 upstream. The current device_release callback for individual iommu drivers does the following: 1) Silent IOMMU DMA translation: It detaches any existing domain from the device and puts it into a blocking state (some drivers might use the identity state). 2) Resource release: It releases resources allocated during the device_probe callback and restores the device to its pre-probe state. Step 1 is challenging for individual iommu drivers because each must check if a domain is already attached to the device. Additionally, if a deferred attach never occurred, the device_release should avoid modifying hardware configuration regardless of the reason for its call. To simplify this process, introduce a static release_domain within the iommu_ops structure. It can be either a blocking or identity domain depending on the iommu hardware. The iommu core will decide whether to attach this domain before the device_release callback, eliminating the need for repetitive code in various drivers. Consequently, the device_release callback can focus solely on the opposite operations of device_probe, including releasing all resources allocated during that callback. Co-developed-by: Jason Gunthorpe Signed-off-by: Jason Gunthorpe Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20240305013305.204605-2-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/iommu.c | 19 +++++++++++++++---- include/linux/iommu.h | 1 + 2 files changed, 16 insertions(+), 4 deletions(-) diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c index b85e62bf16d3..0dbe4d6a452f 100644 --- a/drivers/iommu/iommu.c +++ b/drivers/iommu/iommu.c @@ -458,13 +458,24 @@ static void iommu_deinit_device(struct device *dev) /* * release_device() must stop using any attached domain on the device. - * If there are still other devices in the group they are not effected + * If there are still other devices in the group, they are not affected * by this callback. * - * The IOMMU driver must set the device to either an identity or - * blocking translation and stop using any domain pointer, as it is - * going to be freed. + * If the iommu driver provides release_domain, the core code ensures + * that domain is attached prior to calling release_device. Drivers can + * use this to enforce a translation on the idle iommu. Typically, the + * global static blocked_domain is a good choice. + * + * Otherwise, the iommu driver must set the device to either an identity + * or a blocking translation in release_device() and stop using any + * domain pointer, as it is going to be freed. + * + * Regardless, if a delayed attach never occurred, then the release + * should still avoid touching any hardware configuration either. */ + if (!dev->iommu->attach_deferred && ops->release_domain) + ops->release_domain->ops->attach_dev(ops->release_domain, dev); + if (ops->release_device) ops->release_device(dev); diff --git a/include/linux/iommu.h b/include/linux/iommu.h index e98ec027dce0..1e9c3b68ff00 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -357,6 +357,7 @@ struct iommu_ops { struct module *owner; struct iommu_domain *identity_domain; struct iommu_domain *blocked_domain; + struct iommu_domain *release_domain; struct iommu_domain *default_domain; CK_KABI_RESERVE(1) -- Gitee From dfd79a0ec9bcd9328dcb73834b50a37f7c5c300c Mon Sep 17 00:00:00 2001 From: gaorui Date: Wed, 11 Jun 2025 20:07:25 +0800 Subject: [PATCH 15/18] iommu/vt-d: Fix NULL domain on device release ANBZ: #21698 commit 81e921fd321614c2ad8ac333b041aae1da7a1c6d upstream. In the kdump kernel, the IOMMU operates in deferred_attach mode. In this mode, info->domain may not yet be assigned by the time the release_device function is called. It leads to the following crash in the crash kernel: BUG: kernel NULL pointer dereference, address: 000000000000003c ... RIP: 0010:do_raw_spin_lock+0xa/0xa0 ... _raw_spin_lock_irqsave+0x1b/0x30 intel_iommu_release_device+0x96/0x170 iommu_deinit_device+0x39/0xf0 __iommu_group_remove_device+0xa0/0xd0 iommu_bus_notifier+0x55/0xb0 notifier_call_chain+0x5a/0xd0 blocking_notifier_call_chain+0x41/0x60 bus_notify+0x34/0x50 device_del+0x269/0x3d0 pci_remove_bus_device+0x77/0x100 p2sb_bar+0xae/0x1d0 ... i801_probe+0x423/0x740 Use the release_domain mechanism to fix it. The scalable mode context entry which is not part of release domain should be cleared in release_device(). Fixes: 586081d3f6b1 ("iommu/vt-d: Remove DEFER_DEVICE_DOMAIN_INFO") Reported-by: Eric Badger Closes: https://lore.kernel.org/r/20240113181713.1817855-1-ebadger@purestorage.com Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20240305013305.204605-3-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 31 ++++-------------- drivers/iommu/intel/pasid.c | 64 +++++++++++++++++++++++++++++++++++++ drivers/iommu/intel/pasid.h | 1 + 3 files changed, 71 insertions(+), 25 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 5e92543a390a..8bada5c94742 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -3818,30 +3818,6 @@ static void domain_context_clear(struct device_domain_info *info) &domain_context_clear_one_cb, info); } -static void dmar_remove_one_dev_info(struct device *dev) -{ - struct device_domain_info *info = dev_iommu_priv_get(dev); - struct dmar_domain *domain = info->domain; - struct intel_iommu *iommu = info->iommu; - unsigned long flags; - - if (!dev_is_real_dma_subdevice(info->dev)) { - if (dev_is_pci(info->dev) && sm_supported(iommu)) - intel_pasid_tear_down_entry(iommu, info->dev, - IOMMU_NO_PASID, false); - - iommu_disable_pci_caps(info); - domain_context_clear(info); - } - - spin_lock_irqsave(&domain->lock, flags); - list_del(&info->link); - spin_unlock_irqrestore(&domain->lock, flags); - - domain_detach_iommu(domain, iommu); - info->domain = NULL; -} - /* * Clear the page table pointer in context or pasid table entries so that * all DMA requests without PASID from the device are blocked. If the page @@ -4343,8 +4319,12 @@ static struct iommu_device *intel_iommu_probe_device(struct device *dev) static void intel_iommu_release_device(struct device *dev) { struct device_domain_info *info = dev_iommu_priv_get(dev); + struct intel_iommu *iommu = info->iommu; + + if (sm_supported(iommu) && !dev_is_real_dma_subdevice(dev) && + !context_copied(iommu, info->bus, info->devfn)) + intel_pasid_teardown_sm_context(dev); - dmar_remove_one_dev_info(dev); intel_pasid_free_table(dev); kfree(info); set_dma_ops(dev, NULL); @@ -4799,6 +4779,7 @@ static const struct iommu_dirty_ops intel_dirty_ops = { const struct iommu_ops intel_iommu_ops = { .blocked_domain = &blocking_domain, + .release_domain = &blocking_domain, .capable = intel_iommu_capable, .hw_info = intel_iommu_hw_info, .domain_alloc = intel_iommu_domain_alloc, diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index 066f6837a07c..006fa77d585d 100644 --- a/drivers/iommu/intel/pasid.c +++ b/drivers/iommu/intel/pasid.c @@ -575,3 +575,67 @@ void intel_pasid_setup_page_snoop_control(struct intel_iommu *iommu, if (!cap_caching_mode(iommu->cap)) devtlb_invalidation_with_pasid(iommu, dev, pasid); } + +/* + * Interfaces to setup or teardown a pasid table to the scalable-mode + * context table entry: + */ + +static void device_pasid_table_teardown(struct device *dev, u8 bus, u8 devfn) +{ + struct device_domain_info *info = dev_iommu_priv_get(dev); + struct intel_iommu *iommu = info->iommu; + struct context_entry *context; + + spin_lock(&iommu->lock); + context = iommu_context_addr(iommu, bus, devfn, false); + if (!context) { + spin_unlock(&iommu->lock); + return; + } + + context_clear_entry(context); + __iommu_flush_cache(iommu, context, sizeof(*context)); + spin_unlock(&iommu->lock); + + /* + * Cache invalidation for changes to a scalable-mode context table + * entry. + * + * Section 6.5.3.3 of the VT-d spec: + * - Device-selective context-cache invalidation; + * - Domain-selective PASID-cache invalidation to affected domains + * (can be skipped if all PASID entries were not-present); + * - Domain-selective IOTLB invalidation to affected domains; + * - Global Device-TLB invalidation to affected functions. + * + * The iommu has been parked in the blocking state. All domains have + * been detached from the device or PASID. The PASID and IOTLB caches + * have been invalidated during the domain detach path. + */ + iommu->flush.flush_context(iommu, 0, PCI_DEVID(bus, devfn), + DMA_CCMD_MASK_NOBIT, DMA_CCMD_DEVICE_INVL); + devtlb_invalidation_with_pasid(iommu, dev, IOMMU_NO_PASID); +} + +static int pci_pasid_table_teardown(struct pci_dev *pdev, u16 alias, void *data) +{ + struct device *dev = data; + + if (dev == &pdev->dev) + device_pasid_table_teardown(dev, PCI_BUS_NUM(alias), alias & 0xff); + + return 0; +} + +void intel_pasid_teardown_sm_context(struct device *dev) +{ + struct device_domain_info *info = dev_iommu_priv_get(dev); + + if (!dev_is_pci(dev)) { + device_pasid_table_teardown(dev, info->bus, info->devfn); + return; + } + + pci_for_each_dma_alias(to_pci_dev(dev), pci_pasid_table_teardown, dev); +} diff --git a/drivers/iommu/intel/pasid.h b/drivers/iommu/intel/pasid.h index c5a6e03b7b9e..8ab35081ac89 100644 --- a/drivers/iommu/intel/pasid.h +++ b/drivers/iommu/intel/pasid.h @@ -317,4 +317,5 @@ void intel_pasid_tear_down_entry(struct intel_iommu *iommu, bool fault_ignore); void intel_pasid_setup_page_snoop_control(struct intel_iommu *iommu, struct device *dev, u32 pasid); +void intel_pasid_teardown_sm_context(struct device *dev); #endif /* __INTEL_PASID_H */ -- Gitee From a1be71f86189fb2c04a9f23fbb74ace18c5696d2 Mon Sep 17 00:00:00 2001 From: gaorui Date: Wed, 11 Jun 2025 20:24:17 +0800 Subject: [PATCH 16/18] iommu/vt-d: Setup scalable mode context entry in probe path ANBZ: #21698 commit 301f1a80487fd2f51012533792583d4425e8b8c0 upstream. In contrast to legacy mode, the DMA translation table is configured in the PASID table entry instead of the context entry for scalable mode. For this reason, it is more appropriate to set up the scalable mode context entry in the device_probe callback and direct it to the appropriate PASID table. The iommu domain attach/detach operations only affect the PASID table entry. Therefore, there is no need to modify the context entry when configuring the translation type and page table. The only exception is the kdump case, where context entry setup is postponed until the device driver invokes the first DMA interface. Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20240305013305.204605-4-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 12 ++++ drivers/iommu/intel/pasid.c | 138 ++++++++++++++++++++++++++++++++++++ drivers/iommu/intel/pasid.h | 1 + 3 files changed, 151 insertions(+) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 8bada5c94742..69c41317f8e9 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -4003,6 +4003,10 @@ static int prepare_domain_attach_device(struct iommu_domain *domain, dmar_domain->agaw--; } + if (sm_supported(iommu) && !dev_is_real_dma_subdevice(dev) && + context_copied(iommu, info->bus, info->devfn)) + return intel_pasid_setup_sm_context(dev); + return 0; } @@ -4311,6 +4315,14 @@ static struct iommu_device *intel_iommu_probe_device(struct device *dev) kfree(info); return ERR_PTR(ret); } + + if (!context_copied(iommu, info->bus, info->devfn)) { + ret = intel_pasid_setup_sm_context(dev); + if (ret) { + intel_pasid_free_table(dev); + return ERR_PTR(ret); + } + } } return &iommu->iommu; diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index 006fa77d585d..d236c140234b 100644 --- a/drivers/iommu/intel/pasid.c +++ b/drivers/iommu/intel/pasid.c @@ -639,3 +639,141 @@ void intel_pasid_teardown_sm_context(struct device *dev) pci_for_each_dma_alias(to_pci_dev(dev), pci_pasid_table_teardown, dev); } + +/* + * Get the PASID directory size for scalable mode context entry. + * Value of X in the PDTS field of a scalable mode context entry + * indicates PASID directory with 2^(X + 7) entries. + */ +static unsigned long context_get_sm_pds(struct pasid_table *table) +{ + unsigned long pds, max_pde; + + max_pde = table->max_pasid >> PASID_PDE_SHIFT; + pds = find_first_bit(&max_pde, MAX_NR_PASID_BITS); + if (pds < 7) + return 0; + + return pds - 7; +} + +static int context_entry_set_pasid_table(struct context_entry *context, + struct device *dev) +{ + struct device_domain_info *info = dev_iommu_priv_get(dev); + struct pasid_table *table = info->pasid_table; + struct intel_iommu *iommu = info->iommu; + unsigned long pds; + + context_clear_entry(context); + + pds = context_get_sm_pds(table); + context->lo = (u64)virt_to_phys(table->table) | context_pdts(pds); + context_set_sm_rid2pasid(context, IOMMU_NO_PASID); + + if (info->ats_supported) + context_set_sm_dte(context); + if (info->pri_supported) + context_set_sm_pre(context); + if (info->pasid_supported) + context_set_pasid(context); + + context_set_fault_enable(context); + context_set_present(context); + __iommu_flush_cache(iommu, context, sizeof(*context)); + + return 0; +} + +static int device_pasid_table_setup(struct device *dev, u8 bus, u8 devfn) +{ + struct device_domain_info *info = dev_iommu_priv_get(dev); + struct intel_iommu *iommu = info->iommu; + struct context_entry *context; + + spin_lock(&iommu->lock); + context = iommu_context_addr(iommu, bus, devfn, true); + if (!context) { + spin_unlock(&iommu->lock); + return -ENOMEM; + } + + if (context_present(context) && !context_copied(iommu, bus, devfn)) { + spin_unlock(&iommu->lock); + return 0; + } + + if (context_copied(iommu, bus, devfn)) { + context_clear_entry(context); + __iommu_flush_cache(iommu, context, sizeof(*context)); + + /* + * For kdump cases, old valid entries may be cached due to + * the in-flight DMA and copied pgtable, but there is no + * unmapping behaviour for them, thus we need explicit cache + * flushes for all affected domain IDs and PASIDs used in + * the copied PASID table. Given that we have no idea about + * which domain IDs and PASIDs were used in the copied tables, + * upgrade them to global PASID and IOTLB cache invalidation. + */ + iommu->flush.flush_context(iommu, 0, + PCI_DEVID(bus, devfn), + DMA_CCMD_MASK_NOBIT, + DMA_CCMD_DEVICE_INVL); + qi_flush_pasid_cache(iommu, 0, QI_PC_GLOBAL, 0); + iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH); + devtlb_invalidation_with_pasid(iommu, dev, IOMMU_NO_PASID); + + /* + * At this point, the device is supposed to finish reset at + * its driver probe stage, so no in-flight DMA will exist, + * and we don't need to worry anymore hereafter. + */ + clear_context_copied(iommu, bus, devfn); + } + + context_entry_set_pasid_table(context, dev); + spin_unlock(&iommu->lock); + + /* + * It's a non-present to present mapping. If hardware doesn't cache + * non-present entry we don't need to flush the caches. If it does + * cache non-present entries, then it does so in the special + * domain #0, which we have to flush: + */ + if (cap_caching_mode(iommu->cap)) { + iommu->flush.flush_context(iommu, 0, + PCI_DEVID(bus, devfn), + DMA_CCMD_MASK_NOBIT, + DMA_CCMD_DEVICE_INVL); + iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_DSI_FLUSH); + } + + return 0; +} + +static int pci_pasid_table_setup(struct pci_dev *pdev, u16 alias, void *data) +{ + struct device *dev = data; + + if (dev != &pdev->dev) + return 0; + + return device_pasid_table_setup(dev, PCI_BUS_NUM(alias), alias & 0xff); +} + +/* + * Set the device's PASID table to its context table entry. + * + * The PASID table is set to the context entries of both device itself + * and its alias requester ID for DMA. + */ +int intel_pasid_setup_sm_context(struct device *dev) +{ + struct device_domain_info *info = dev_iommu_priv_get(dev); + + if (!dev_is_pci(dev)) + return device_pasid_table_setup(dev, info->bus, info->devfn); + + return pci_for_each_dma_alias(to_pci_dev(dev), pci_pasid_table_setup, dev); +} diff --git a/drivers/iommu/intel/pasid.h b/drivers/iommu/intel/pasid.h index 8ab35081ac89..7cc889d6cba9 100644 --- a/drivers/iommu/intel/pasid.h +++ b/drivers/iommu/intel/pasid.h @@ -317,5 +317,6 @@ void intel_pasid_tear_down_entry(struct intel_iommu *iommu, bool fault_ignore); void intel_pasid_setup_page_snoop_control(struct intel_iommu *iommu, struct device *dev, u32 pasid); +int intel_pasid_setup_sm_context(struct device *dev); void intel_pasid_teardown_sm_context(struct device *dev); #endif /* __INTEL_PASID_H */ -- Gitee From 9fe2433d488000571e2ef2ce85faf9c29200db4a Mon Sep 17 00:00:00 2001 From: gaorui Date: Wed, 11 Jun 2025 20:31:13 +0800 Subject: [PATCH 17/18] iommu/vt-d: Remove scalable mode context entry setup from attach_dev ANBZ: #21698 commit a016e53843ed8bb9cccb832768e2be9856b0a913 upstream. The scalable mode context entry is now setup in the probe_device path, eliminating the need to configure it in the attach_dev path. Removes the redundant code from the attach_dev path to avoid dead code. Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20240305013305.204605-5-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 156 ++++++++++-------------------------- 1 file changed, 44 insertions(+), 112 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 69c41317f8e9..00823fd033ce 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -1742,34 +1742,17 @@ static void domain_exit(struct dmar_domain *domain) kfree(domain); } -/* - * Get the PASID directory size for scalable mode context entry. - * Value of X in the PDTS field of a scalable mode context entry - * indicates PASID directory with 2^(X + 7) entries. - */ -static unsigned long context_get_sm_pds(struct pasid_table *table) -{ - unsigned long pds, max_pde; - - max_pde = table->max_pasid >> PASID_PDE_SHIFT; - pds = find_first_bit(&max_pde, MAX_NR_PASID_BITS); - if (pds < 7) - return 0; - - return pds - 7; -} - static int domain_context_mapping_one(struct dmar_domain *domain, struct intel_iommu *iommu, - struct pasid_table *table, u8 bus, u8 devfn) { struct device_domain_info *info = domain_lookup_dev_info(domain, iommu, bus, devfn); u16 did = domain_id_iommu(domain, iommu); int translation = CONTEXT_TT_MULTI_LEVEL; + struct dma_pte *pgd = domain->pgd; struct context_entry *context; - int ret; + int agaw, ret; if (hw_pass_through && domain_type_is_si(domain)) translation = CONTEXT_TT_PASS_THROUGH; @@ -1812,65 +1795,37 @@ static int domain_context_mapping_one(struct dmar_domain *domain, } context_clear_entry(context); + context_set_domain_id(context, did); - if (sm_supported(iommu)) { - unsigned long pds; - - /* Setup the PASID DIR pointer: */ - pds = context_get_sm_pds(table); - context->lo = (u64)virt_to_phys(table->table) | - context_pdts(pds); - - /* Setup the RID_PASID field: */ - context_set_sm_rid2pasid(context, IOMMU_NO_PASID); - + if (translation != CONTEXT_TT_PASS_THROUGH) { /* - * Setup the Device-TLB enable bit and Page request - * Enable bit: + * Skip top levels of page tables for iommu which has + * less agaw than default. Unnecessary for PT mode. */ - if (info && info->ats_supported) - context_set_sm_dte(context); - if (info && info->pri_supported) - context_set_sm_pre(context); - if (info && info->pasid_supported) - context_set_pasid(context); - } else { - struct dma_pte *pgd = domain->pgd; - int agaw; - - context_set_domain_id(context, did); - - if (translation != CONTEXT_TT_PASS_THROUGH) { - /* - * Skip top levels of page tables for iommu which has - * less agaw than default. Unnecessary for PT mode. - */ - for (agaw = domain->agaw; agaw > iommu->agaw; agaw--) { - ret = -ENOMEM; - pgd = phys_to_virt(dma_pte_addr(pgd)); - if (!dma_pte_present(pgd)) - goto out_unlock; - } - - if (info && info->ats_supported) - translation = CONTEXT_TT_DEV_IOTLB; - else - translation = CONTEXT_TT_MULTI_LEVEL; - - context_set_address_root(context, virt_to_phys(pgd)); - context_set_address_width(context, agaw); - } else { - /* - * In pass through mode, AW must be programmed to - * indicate the largest AGAW value supported by - * hardware. And ASR is ignored by hardware. - */ - context_set_address_width(context, iommu->msagaw); + for (agaw = domain->agaw; agaw > iommu->agaw; agaw--) { + ret = -ENOMEM; + pgd = phys_to_virt(dma_pte_addr(pgd)); + if (!dma_pte_present(pgd)) + goto out_unlock; } - context_set_translation_type(context, translation); + if (info && info->ats_supported) + translation = CONTEXT_TT_DEV_IOTLB; + else + translation = CONTEXT_TT_MULTI_LEVEL; + + context_set_address_root(context, virt_to_phys(pgd)); + context_set_address_width(context, agaw); + } else { + /* + * In pass through mode, AW must be programmed to + * indicate the largest AGAW value supported by + * hardware. And ASR is ignored by hardware. + */ + context_set_address_width(context, iommu->msagaw); } + context_set_translation_type(context, translation); context_set_fault_enable(context); context_set_present(context); if (!ecap_coherent(iommu->ecap)) @@ -1900,43 +1855,29 @@ static int domain_context_mapping_one(struct dmar_domain *domain, return ret; } -struct domain_context_mapping_data { - struct dmar_domain *domain; - struct intel_iommu *iommu; - struct pasid_table *table; -}; - static int domain_context_mapping_cb(struct pci_dev *pdev, u16 alias, void *opaque) { - struct domain_context_mapping_data *data = opaque; + struct device_domain_info *info = dev_iommu_priv_get(&pdev->dev); + struct intel_iommu *iommu = info->iommu; + struct dmar_domain *domain = opaque; - return domain_context_mapping_one(data->domain, data->iommu, - data->table, PCI_BUS_NUM(alias), - alias & 0xff); + return domain_context_mapping_one(domain, iommu, + PCI_BUS_NUM(alias), alias & 0xff); } static int domain_context_mapping(struct dmar_domain *domain, struct device *dev) { struct device_domain_info *info = dev_iommu_priv_get(dev); - struct domain_context_mapping_data data; struct intel_iommu *iommu = info->iommu; u8 bus = info->bus, devfn = info->devfn; - struct pasid_table *table; - - table = intel_pasid_get_table(dev); if (!dev_is_pci(dev)) - return domain_context_mapping_one(domain, iommu, table, - bus, devfn); - - data.domain = domain; - data.iommu = iommu; - data.table = table; + return domain_context_mapping_one(domain, iommu, bus, devfn); return pci_for_each_dma_alias(to_pci_dev(dev), - &domain_context_mapping_cb, &data); + domain_context_mapping_cb, domain); } /* Returns a number of VTD pages, but aligned to MM page size */ @@ -2288,28 +2229,19 @@ static int dmar_domain_attach_device(struct dmar_domain *domain, list_add(&info->link, &domain->devices); spin_unlock_irqrestore(&domain->lock, flags); - /* PASID table is mandatory for a PCI device in scalable mode. */ - if (sm_supported(iommu) && !dev_is_real_dma_subdevice(dev)) { - /* Setup the PASID entry for requests without PASID: */ - if (hw_pass_through && domain_type_is_si(domain)) - ret = intel_pasid_setup_pass_through(iommu, - dev, IOMMU_NO_PASID); - else if (domain->use_first_level) - ret = domain_setup_first_level(iommu, domain, dev, - IOMMU_NO_PASID); - else - ret = intel_pasid_setup_second_level(iommu, domain, - dev, IOMMU_NO_PASID); - if (ret) { - dev_err(dev, "Setup RID2PASID failed\n"); - device_block_translation(dev); - return ret; - } - } + if (dev_is_real_dma_subdevice(dev)) + return 0; + + if (!sm_supported(iommu)) + ret = domain_context_mapping(domain, dev); + else if (hw_pass_through && domain_type_is_si(domain)) + ret = intel_pasid_setup_pass_through(iommu, dev, IOMMU_NO_PASID); + else if (domain->use_first_level) + ret = domain_setup_first_level(iommu, domain, dev, IOMMU_NO_PASID); + else + ret = intel_pasid_setup_second_level(iommu, domain, dev, IOMMU_NO_PASID); - ret = domain_context_mapping(domain, dev); if (ret) { - dev_err(dev, "Domain context map failed\n"); device_block_translation(dev); return ret; } -- Gitee From 4fa9faa4ce2c13f6d7b8179cd1e424ec954c434c Mon Sep 17 00:00:00 2001 From: gaorui Date: Wed, 11 Jun 2025 20:35:08 +0800 Subject: [PATCH 18/18] iommu/vt-d: Remove scalabe mode in domain_context_clear_one() ANBZ: #21698 commit 80ca79f398bff2708ee7a19d5904804415680aad upstream. domain_context_clear_one() only handles the context entry teardown in legacy mode. Remove the scalable mode check in it to avoid dead code. Remove an unnecessary check in the code as well. Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/20240305013305.204605-6-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel Signed-off-by: gaorui --- drivers/iommu/intel/iommu.c | 15 +-------------- 1 file changed, 1 insertion(+), 14 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 00823fd033ce..176589778783 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -2059,9 +2059,6 @@ static void domain_context_clear_one(struct device_domain_info *info, u8 bus, u8 struct context_entry *context; u16 did_old; - if (!iommu) - return; - spin_lock(&iommu->lock); context = iommu_context_addr(iommu, bus, devfn, 0); if (!context) { @@ -2069,14 +2066,7 @@ static void domain_context_clear_one(struct device_domain_info *info, u8 bus, u8 return; } - if (sm_supported(iommu)) { - if (hw_pass_through && domain_type_is_si(info->domain)) - did_old = FLPT_DEFAULT_DID; - else - did_old = domain_id_iommu(info->domain, iommu); - } else { - did_old = context_domain_id(context); - } + did_old = context_domain_id(context); context_clear_entry(context); __iommu_flush_cache(iommu, context, sizeof(*context)); @@ -2087,9 +2077,6 @@ static void domain_context_clear_one(struct device_domain_info *info, u8 bus, u8 DMA_CCMD_MASK_NOBIT, DMA_CCMD_DEVICE_INVL); - if (sm_supported(iommu)) - qi_flush_pasid_cache(iommu, did_old, QI_PC_ALL_PASIDS, 0); - iommu->flush.flush_iotlb(iommu, did_old, 0, -- Gitee