KVM: PPC: Book3s: Retire H_PUT_TCE/etc real mode handlers

LoPAPR defines guest visible IOMMU with hypercalls to use it -
H_PUT_TCE/etc. Implemented first on POWER7 where hypercalls would trap
in the KVM in the real mode (with MMU off). The problem with the real mode
is some memory is not available and some API usage crashed the host but
enabling MMU was an expensive operation.

The problems with the real mode handlers are:
1. Occasionally these cannot complete the request so the code is
copied+modified to work in the virtual mode, very little is shared;
2. The real mode handlers have to be linked into vmlinux to work;
3. An exception in real mode immediately reboots the machine.

If the small DMA window is used, the real mode handlers bring better
performance. However since POWER8, there has always been a bigger DMA
window which VMs use to map the entire VM memory to avoid calling
H_PUT_TCE. Such 1:1 mapping happens once and uses H_PUT_TCE_INDIRECT
(a bulk version of H_PUT_TCE) which virtual mode handler is even closer
to its real mode version.

On POWER9 hypercalls trap straight to the virtual mode so the real mode
handlers never execute on POWER9 and later CPUs.

So with the current use of the DMA windows and MMU improvements in
POWER9 and later, there is no point in duplicating the code.
The 32bit passed through devices may slow down but we do not have many
of these in practice. For example, with this applied, a 1Gbit ethernet
adapter still demostrates above 800Mbit/s of actual throughput.

This removes the real mode handlers from KVM and related code from
the powernv platform.

This updates the list of implemented hcalls in KVM-HV as the realmode
handlers are removed.

This changes ABI - kvmppc_h_get_tce() moves to the KVM module and
kvmppc_find_table() is static now.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220506053755.3820702-1-aik@ozlabs.ru

arch/powerpc/include/asm/iommu.h

@@ -51,13 +51,11 @@ struct iommu_table_ops {
 	int (*xchg_no_kill)(struct iommu_table *tbl,
 			long index,
 			unsigned long *hpa,
-			enum dma_data_direction *direction,
-			bool realmode);
+			enum dma_data_direction *direction);
 
 	void (*tce_kill)(struct iommu_table *tbl,
 			unsigned long index,
-			unsigned long pages,
-			bool realmode);
+			unsigned long pages);
 
 	__be64 *(*useraddrptr)(struct iommu_table *tbl, long index, bool alloc);
 #endif

arch/powerpc/include/asm/kvm_ppc.h

@@ -177,8 +177,6 @@ extern void kvmppc_setup_partition_table(struct kvm *kvm);
 
 extern long kvm_vm_ioctl_create_spapr_tce(struct kvm *kvm,
 				struct kvm_create_spapr_tce_64 *args);
-extern struct kvmppc_spapr_tce_table *kvmppc_find_table(
-		struct kvm *kvm, unsigned long liobn);
 #define kvmppc_ioba_validate(stt, ioba, npages)                         \
 		(iommu_tce_check_ioba((stt)->page_shift, (stt)->offset, \
 				(stt)->size, (ioba), (npages)) ?        \

arch/powerpc/include/asm/mmu_context.h

@@ -34,15 +34,10 @@ extern void mm_iommu_init(struct mm_struct *mm);
 extern void mm_iommu_cleanup(struct mm_struct *mm);
 extern struct mm_iommu_table_group_mem_t *mm_iommu_lookup(struct mm_struct *mm,
 		unsigned long ua, unsigned long size);
-extern struct mm_iommu_table_group_mem_t *mm_iommu_lookup_rm(
-		struct mm_struct *mm, unsigned long ua, unsigned long size);
 extern struct mm_iommu_table_group_mem_t *mm_iommu_get(struct mm_struct *mm,
 		unsigned long ua, unsigned long entries);
 extern long mm_iommu_ua_to_hpa(struct mm_iommu_table_group_mem_t *mem,
 		unsigned long ua, unsigned int pageshift, unsigned long *hpa);
-extern long mm_iommu_ua_to_hpa_rm(struct mm_iommu_table_group_mem_t *mem,
-		unsigned long ua, unsigned int pageshift, unsigned long *hpa);
-extern void mm_iommu_ua_mark_dirty_rm(struct mm_struct *mm, unsigned long ua);
 extern bool mm_iommu_is_devmem(struct mm_struct *mm, unsigned long hpa,
 		unsigned int pageshift, unsigned long *size);
 extern long mm_iommu_mapped_inc(struct mm_iommu_table_group_mem_t *mem);

arch/powerpc/kernel/iommu.c

@@ -1065,7 +1065,7 @@ extern long iommu_tce_xchg_no_kill(struct mm_struct *mm,
 	long ret;
 	unsigned long size = 0;
 
-	ret = tbl->it_ops->xchg_no_kill(tbl, entry, hpa, direction, false);
+	ret = tbl->it_ops->xchg_no_kill(tbl, entry, hpa, direction);
 	if (!ret && ((*direction == DMA_FROM_DEVICE) ||
 			(*direction == DMA_BIDIRECTIONAL)) &&
 			!mm_iommu_is_devmem(mm, *hpa, tbl->it_page_shift,
@@ -1080,7 +1080,7 @@ void iommu_tce_kill(struct iommu_table *tbl,
 		unsigned long entry, unsigned long pages)
 {
 	if (tbl->it_ops->tce_kill)
-		tbl->it_ops->tce_kill(tbl, entry, pages, false);
+		tbl->it_ops->tce_kill(tbl, entry, pages);
 }
 EXPORT_SYMBOL_GPL(iommu_tce_kill);
 

arch/powerpc/kvm/Makefile

@@ -37,9 +37,6 @@ kvm-e500mc-objs := \
 	e500_emulate.o
 kvm-objs-$(CONFIG_KVM_E500MC) := $(kvm-e500mc-objs)
 
-kvm-book3s_64-builtin-objs-$(CONFIG_SPAPR_TCE_IOMMU) := \
-	book3s_64_vio_hv.o
-
 kvm-pr-y := \
 	fpu.o \
 	emulate.o \

arch/powerpc/kvm/book3s_64_vio.c

@@ -32,6 +32,18 @@
 #include <asm/tce.h>
 #include <asm/mmu_context.h>
 
+static struct kvmppc_spapr_tce_table *kvmppc_find_table(struct kvm *kvm,
+	unsigned long liobn)
+{
+	struct kvmppc_spapr_tce_table *stt;
+
+	list_for_each_entry_lockless(stt, &kvm->arch.spapr_tce_tables, list)
+		if (stt->liobn == liobn)
+			return stt;
+
+	return NULL;
+}
+
 static unsigned long kvmppc_tce_pages(unsigned long iommu_pages)
 {
 	return ALIGN(iommu_pages * sizeof(u64), PAGE_SIZE) / PAGE_SIZE;
@@ -753,3 +765,34 @@ long kvmppc_h_stuff_tce(struct kvm_vcpu *vcpu,
 	return ret;
 }
 EXPORT_SYMBOL_GPL(kvmppc_h_stuff_tce);
+
+long kvmppc_h_get_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
+		      unsigned long ioba)
+{
+	struct kvmppc_spapr_tce_table *stt;
+	long ret;
+	unsigned long idx;
+	struct page *page;
+	u64 *tbl;
+
+	stt = kvmppc_find_table(vcpu->kvm, liobn);
+	if (!stt)
+		return H_TOO_HARD;
+
+	ret = kvmppc_ioba_validate(stt, ioba, 1);
+	if (ret != H_SUCCESS)
+		return ret;
+
+	idx = (ioba >> stt->page_shift) - stt->offset;
+	page = stt->pages[idx / TCES_PER_PAGE];
+	if (!page) {
+		vcpu->arch.regs.gpr[4] = 0;
+		return H_SUCCESS;
+	}
+	tbl = (u64 *)page_address(page);
+
+	vcpu->arch.regs.gpr[4] = tbl[idx % TCES_PER_PAGE];
+
+	return H_SUCCESS;
+}
+EXPORT_SYMBOL_GPL(kvmppc_h_get_tce);

arch/powerpc/kvm/book3s_hv.c

@@ -1326,6 +1326,12 @@ static int kvmppc_hcall_impl_hv(unsigned long cmd)
 	case H_CONFER:
 	case H_REGISTER_VPA:
 	case H_SET_MODE:
+#ifdef CONFIG_SPAPR_TCE_IOMMU
+	case H_GET_TCE:
+	case H_PUT_TCE:
+	case H_PUT_TCE_INDIRECT:
+	case H_STUFF_TCE:
+#endif
 	case H_LOGICAL_CI_LOAD:
 	case H_LOGICAL_CI_STORE:
 #ifdef CONFIG_KVM_XICS

arch/powerpc/kvm/book3s_hv_rmhandlers.S

@@ -1792,13 +1792,8 @@ hcall_real_table:
 	.long	DOTSYM(kvmppc_h_clear_mod) - hcall_real_table
 	.long	DOTSYM(kvmppc_h_clear_ref) - hcall_real_table
 	.long	DOTSYM(kvmppc_h_protect) - hcall_real_table
-#ifdef CONFIG_SPAPR_TCE_IOMMU
-	.long	DOTSYM(kvmppc_h_get_tce) - hcall_real_table
-	.long	DOTSYM(kvmppc_rm_h_put_tce) - hcall_real_table
-#else
 	.long	0		/* 0x1c */
 	.long	0		/* 0x20 */
-#endif
 	.long	0		/* 0x24 - H_SET_SPRG0 */
 	.long	DOTSYM(kvmppc_h_set_dabr) - hcall_real_table
 	.long	DOTSYM(kvmppc_rm_h_page_init) - hcall_real_table
@@ -1876,13 +1871,8 @@ hcall_real_table:
 	.long	0		/* 0x12c */
 	.long	0		/* 0x130 */
 	.long	DOTSYM(kvmppc_h_set_xdabr) - hcall_real_table
-#ifdef CONFIG_SPAPR_TCE_IOMMU
-	.long	DOTSYM(kvmppc_rm_h_stuff_tce) - hcall_real_table
-	.long	DOTSYM(kvmppc_rm_h_put_tce_indirect) - hcall_real_table
-#else
 	.long	0		/* 0x138 */
 	.long	0		/* 0x13c */
-#endif
 	.long	0		/* 0x140 */
 	.long	0		/* 0x144 */
 	.long	0		/* 0x148 */

arch/powerpc/mm/book3s64/iommu_api.c

@@ -305,24 +305,6 @@ struct mm_iommu_table_group_mem_t *mm_iommu_lookup(struct mm_struct *mm,
 }
 EXPORT_SYMBOL_GPL(mm_iommu_lookup);
 
-struct mm_iommu_table_group_mem_t *mm_iommu_lookup_rm(struct mm_struct *mm,
-		unsigned long ua, unsigned long size)
-{
-	struct mm_iommu_table_group_mem_t *mem, *ret = NULL;
-
-	list_for_each_entry_lockless(mem, &mm->context.iommu_group_mem_list,
-			next) {
-		if ((mem->ua <= ua) &&
-				(ua + size <= mem->ua +
-				 (mem->entries << PAGE_SHIFT))) {
-			ret = mem;
-			break;
-		}
-	}
-
-	return ret;
-}
-
 struct mm_iommu_table_group_mem_t *mm_iommu_get(struct mm_struct *mm,
 		unsigned long ua, unsigned long entries)
 {
@@ -369,56 +351,6 @@ long mm_iommu_ua_to_hpa(struct mm_iommu_table_group_mem_t *mem,
 }
 EXPORT_SYMBOL_GPL(mm_iommu_ua_to_hpa);
 
-long mm_iommu_ua_to_hpa_rm(struct mm_iommu_table_group_mem_t *mem,
-		unsigned long ua, unsigned int pageshift, unsigned long *hpa)
-{
-	const long entry = (ua - mem->ua) >> PAGE_SHIFT;
-	unsigned long *pa;
-
-	if (entry >= mem->entries)
-		return -EFAULT;
-
-	if (pageshift > mem->pageshift)
-		return -EFAULT;
-
-	if (!mem->hpas) {
-		*hpa = mem->dev_hpa + (ua - mem->ua);
-		return 0;
-	}
-
-	pa = (void *) vmalloc_to_phys(&mem->hpas[entry]);
-	if (!pa)
-		return -EFAULT;
-
-	*hpa = (*pa & MM_IOMMU_TABLE_GROUP_PAGE_MASK) | (ua & ~PAGE_MASK);
-
-	return 0;
-}
-
-extern void mm_iommu_ua_mark_dirty_rm(struct mm_struct *mm, unsigned long ua)
-{
-	struct mm_iommu_table_group_mem_t *mem;
-	long entry;
-	void *va;
-	unsigned long *pa;
-
-	mem = mm_iommu_lookup_rm(mm, ua, PAGE_SIZE);
-	if (!mem)
-		return;
-
-	if (mem->dev_hpa != MM_IOMMU_TABLE_INVALID_HPA)
-		return;
-
-	entry = (ua - mem->ua) >> PAGE_SHIFT;
-	va = &mem->hpas[entry];
-
-	pa = (void *) vmalloc_to_phys(va);
-	if (!pa)
-		return;
-
-	*pa |= MM_IOMMU_TABLE_GROUP_PAGE_DIRTY;
-}
-
 bool mm_iommu_is_devmem(struct mm_struct *mm, unsigned long hpa,
 		unsigned int pageshift, unsigned long *size)
 {

arch/powerpc/platforms/powernv/pci-ioda-tce.c

@@ -145,8 +145,7 @@ int pnv_tce_build(struct iommu_table *tbl, long index, long npages,
 
 #ifdef CONFIG_IOMMU_API
 int pnv_tce_xchg(struct iommu_table *tbl, long index,
-		unsigned long *hpa, enum dma_data_direction *direction,
-		bool alloc)
+		unsigned long *hpa, enum dma_data_direction *direction)
 {
 	u64 proto_tce = iommu_direction_to_tce_perm(*direction);
 	unsigned long newtce = *hpa | proto_tce, oldtce;
@@ -164,7 +163,7 @@ int pnv_tce_xchg(struct iommu_table *tbl, long index,
 	}
 
 	if (!ptce) {
-		ptce = pnv_tce(tbl, false, idx, alloc);
+		ptce = pnv_tce(tbl, false, idx, true);
 		if (!ptce)
 			return -ENOMEM;
 	}

arch/powerpc/platforms/powernv/pci-ioda.c

@@ -1267,22 +1267,20 @@ static bool pnv_pci_ioda_iommu_bypass_supported(struct pci_dev *pdev,
 	return false;
 }
 
-static inline __be64 __iomem *pnv_ioda_get_inval_reg(struct pnv_phb *phb,
-						     bool real_mode)
+static inline __be64 __iomem *pnv_ioda_get_inval_reg(struct pnv_phb *phb)
 {
-	return real_mode ? (__be64 __iomem *)(phb->regs_phys + 0x210) :
-		(phb->regs + 0x210);
+	return phb->regs + 0x210;
 }
 
 static void pnv_pci_p7ioc_tce_invalidate(struct iommu_table *tbl,
-		unsigned long index, unsigned long npages, bool rm)
+		unsigned long index, unsigned long npages)
 {
 	struct iommu_table_group_link *tgl = list_first_entry_or_null(
 			&tbl->it_group_list, struct iommu_table_group_link,
 			next);
 	struct pnv_ioda_pe *pe = container_of(tgl->table_group,
 			struct pnv_ioda_pe, table_group);
-	__be64 __iomem *invalidate = pnv_ioda_get_inval_reg(pe->phb, rm);
+	__be64 __iomem *invalidate = pnv_ioda_get_inval_reg(pe->phb);
 	unsigned long start, end, inc;
 
 	start = __pa(((__be64 *)tbl->it_base) + index - tbl->it_offset);
@@ -1297,11 +1295,7 @@ static void pnv_pci_p7ioc_tce_invalidate(struct iommu_table *tbl,
 
         mb(); /* Ensure above stores are visible */
         while (start <= end) {
-		if (rm)
-			__raw_rm_writeq_be(start, invalidate);
-		else
-			__raw_writeq_be(start, invalidate);
-
+		__raw_writeq_be(start, invalidate);
                 start += inc;
         }
 
@@ -1320,7 +1314,7 @@ static int pnv_ioda1_tce_build(struct iommu_table *tbl, long index,
 			attrs);
 
 	if (!ret)
-		pnv_pci_p7ioc_tce_invalidate(tbl, index, npages, false);
+		pnv_pci_p7ioc_tce_invalidate(tbl, index, npages);
 
 	return ret;
 }
@@ -1328,10 +1322,9 @@ static int pnv_ioda1_tce_build(struct iommu_table *tbl, long index,
 #ifdef CONFIG_IOMMU_API
 /* Common for IODA1 and IODA2 */
 static int pnv_ioda_tce_xchg_no_kill(struct iommu_table *tbl, long index,
-		unsigned long *hpa, enum dma_data_direction *direction,
-		bool realmode)
+		unsigned long *hpa, enum dma_data_direction *direction)
 {
-	return pnv_tce_xchg(tbl, index, hpa, direction, !realmode);
+	return pnv_tce_xchg(tbl, index, hpa, direction);
 }
 #endif
 
@@ -1340,7 +1333,7 @@ static void pnv_ioda1_tce_free(struct iommu_table *tbl, long index,
 {
 	pnv_tce_free(tbl, index, npages);
 
-	pnv_pci_p7ioc_tce_invalidate(tbl, index, npages, false);
+	pnv_pci_p7ioc_tce_invalidate(tbl, index, npages);
 }
 
 static struct iommu_table_ops pnv_ioda1_iommu_ops = {
@@ -1361,18 +1354,18 @@ static struct iommu_table_ops pnv_ioda1_iommu_ops = {
 static inline void pnv_pci_phb3_tce_invalidate_pe(struct pnv_ioda_pe *pe)
 {
 	/* 01xb - invalidate TCEs that match the specified PE# */
-	__be64 __iomem *invalidate = pnv_ioda_get_inval_reg(pe->phb, false);
+	__be64 __iomem *invalidate = pnv_ioda_get_inval_reg(pe->phb);
 	unsigned long val = PHB3_TCE_KILL_INVAL_PE | (pe->pe_number & 0xFF);
 
 	mb(); /* Ensure above stores are visible */
 	__raw_writeq_be(val, invalidate);
 }
 
-static void pnv_pci_phb3_tce_invalidate(struct pnv_ioda_pe *pe, bool rm,
+static void pnv_pci_phb3_tce_invalidate(struct pnv_ioda_pe *pe,
 					unsigned shift, unsigned long index,
 					unsigned long npages)
 {
-	__be64 __iomem *invalidate = pnv_ioda_get_inval_reg(pe->phb, rm);
+	__be64 __iomem *invalidate = pnv_ioda_get_inval_reg(pe->phb);
 	unsigned long start, end, inc;
 
 	/* We'll invalidate DMA address in PE scope */
@@ -1387,10 +1380,7 @@ static void pnv_pci_phb3_tce_invalidate(struct pnv_ioda_pe *pe, bool rm,
 	mb();
 
 	while (start <= end) {
-		if (rm)
-			__raw_rm_writeq_be(start, invalidate);
-		else
-			__raw_writeq_be(start, invalidate);
+		__raw_writeq_be(start, invalidate);
 		start += inc;
 	}
 }
@@ -1407,7 +1397,7 @@ static inline void pnv_pci_ioda2_tce_invalidate_pe(struct pnv_ioda_pe *pe)
 }
 
 static void pnv_pci_ioda2_tce_invalidate(struct iommu_table *tbl,
-		unsigned long index, unsigned long npages, bool rm)
+		unsigned long index, unsigned long npages)
 {
 	struct iommu_table_group_link *tgl;
 
@@ -1418,7 +1408,7 @@ static void pnv_pci_ioda2_tce_invalidate(struct iommu_table *tbl,
 		unsigned int shift = tbl->it_page_shift;
 
 		if (phb->model == PNV_PHB_MODEL_PHB3 && phb->regs)
-			pnv_pci_phb3_tce_invalidate(pe, rm, shift,
+			pnv_pci_phb3_tce_invalidate(pe, shift,
 						    index, npages);
 		else
 			opal_pci_tce_kill(phb->opal_id,
@@ -1437,7 +1427,7 @@ static int pnv_ioda2_tce_build(struct iommu_table *tbl, long index,
 			attrs);
 
 	if (!ret)
-		pnv_pci_ioda2_tce_invalidate(tbl, index, npages, false);
+		pnv_pci_ioda2_tce_invalidate(tbl, index, npages);
 
 	return ret;
 }
@@ -1447,7 +1437,7 @@ static void pnv_ioda2_tce_free(struct iommu_table *tbl, long index,
 {
 	pnv_tce_free(tbl, index, npages);
 
-	pnv_pci_ioda2_tce_invalidate(tbl, index, npages, false);
+	pnv_pci_ioda2_tce_invalidate(tbl, index, npages);
 }
 
 static struct iommu_table_ops pnv_ioda2_iommu_ops = {
@@ -2738,7 +2728,7 @@ static void pnv_pci_ioda1_release_pe_dma(struct pnv_ioda_pe *pe)
 	if (rc != OPAL_SUCCESS)
 		return;
 
-	pnv_pci_p7ioc_tce_invalidate(tbl, tbl->it_offset, tbl->it_size, false);
+	pnv_pci_p7ioc_tce_invalidate(tbl, tbl->it_offset, tbl->it_size);
 	if (pe->table_group.group) {
 		iommu_group_put(pe->table_group.group);
 		WARN_ON(pe->table_group.group);

arch/powerpc/platforms/powernv/pci.h

@@ -311,8 +311,7 @@ extern int pnv_tce_build(struct iommu_table *tbl, long index, long npages,
 		unsigned long attrs);
 extern void pnv_tce_free(struct iommu_table *tbl, long index, long npages);
 extern int pnv_tce_xchg(struct iommu_table *tbl, long index,
-		unsigned long *hpa, enum dma_data_direction *direction,
-		bool alloc);
+		unsigned long *hpa, enum dma_data_direction *direction);
 extern __be64 *pnv_tce_useraddrptr(struct iommu_table *tbl, long index,
 		bool alloc);
 extern unsigned long pnv_tce_get(struct iommu_table *tbl, long index);

arch/powerpc/platforms/pseries/iommu.c

@@ -666,8 +666,7 @@ static void pci_dma_bus_setup_pSeries(struct pci_bus *bus)
 
 #ifdef CONFIG_IOMMU_API
 static int tce_exchange_pseries(struct iommu_table *tbl, long index, unsigned
-				long *tce, enum dma_data_direction *direction,
-				bool realmode)
+				long *tce, enum dma_data_direction *direction)
 {
 	long rc;
 	unsigned long ioba = (unsigned long) index << tbl->it_page_shift;