Merge branch 'for-6.19/cxl-addr-xlat' into cxl-for-next

Enable unit testing for XOR address translation of SPA to DPA and vice versa.

Author: davejiang

drivers/cxl/acpi.c

@@ -11,25 +11,36 @@
 #include "cxlpci.h"
 #include "cxl.h"
 
-struct cxl_cxims_data {
-	int nr_maps;
-	u64 xormaps[] __counted_by(nr_maps);
-};
-
 static const guid_t acpi_cxl_qtg_id_guid =
 	GUID_INIT(0xF365F9A6, 0xA7DE, 0x4071,
 		  0xA6, 0x6A, 0xB4, 0x0C, 0x0B, 0x4F, 0x8E, 0x52);
 
-static u64 cxl_apply_xor_maps(struct cxl_root_decoder *cxlrd, u64 addr)
+#define HBIW_TO_NR_MAPS_SIZE (CXL_DECODER_MAX_INTERLEAVE + 1)
+static const int hbiw_to_nr_maps[HBIW_TO_NR_MAPS_SIZE] = {
+	[1] = 0, [2] = 1, [3] = 0, [4] = 2, [6] = 1, [8] = 3, [12] = 2, [16] = 4
+};
+
+static const int valid_hbiw[] = { 1, 2, 3, 4, 6, 8, 12, 16 };
+
+u64 cxl_do_xormap_calc(struct cxl_cxims_data *cximsd, u64 addr, int hbiw)
 {
-	struct cxl_cxims_data *cximsd = cxlrd->platform_data;
-	int hbiw = cxlrd->cxlsd.nr_targets;
+	int nr_maps_to_apply = -1;
 	u64 val;
 	int pos;
 
-	/* No xormaps for host bridge interleave ways of 1 or 3 */
-	if (hbiw == 1 || hbiw == 3)
-		return addr;
+	/*
+	 * Strictly validate hbiw since this function is used for testing and
+	 * that nullifies any expectation of trusted parameters from the CXL
+	 * Region Driver.
+	 */
+	for (int i = 0; i < ARRAY_SIZE(valid_hbiw); i++) {
+		if (valid_hbiw[i] == hbiw) {
+			nr_maps_to_apply = hbiw_to_nr_maps[hbiw];
+			break;
+		}
+	}
+	if (nr_maps_to_apply == -1 || nr_maps_to_apply > cximsd->nr_maps)
+		return ULLONG_MAX;
 
 	/*
 	 * In regions using XOR interleave arithmetic the CXL HPA may not
@@ -60,6 +71,14 @@ static u64 cxl_apply_xor_maps(struct cxl_root_decoder *cxlrd, u64 addr)
 
 	return addr;
 }
+EXPORT_SYMBOL_FOR_MODULES(cxl_do_xormap_calc, "cxl_translate");
+
+static u64 cxl_apply_xor_maps(struct cxl_root_decoder *cxlrd, u64 addr)
+{
+	struct cxl_cxims_data *cximsd = cxlrd->platform_data;
+
+	return cxl_do_xormap_calc(cximsd, addr, cxlrd->cxlsd.nr_targets);
+}
 
 struct cxl_cxims_context {
 	struct device *dev;

drivers/cxl/core/region.c

@@ -2953,28 +2953,119 @@ static bool has_spa_to_hpa(struct cxl_root_decoder *cxlrd)
 	return cxlrd->ops && cxlrd->ops->spa_to_hpa;
 }
 
-u64 cxl_dpa_to_hpa(struct cxl_region *cxlr, const struct cxl_memdev *cxlmd,
-		   u64 dpa)
+#define CXL_POS_ZERO 0
+/**
+ * cxl_validate_translation_params
+ * @eiw: encoded interleave ways
+ * @eig: encoded interleave granularity
+ * @pos: position in interleave
+ *
+ * Callers pass CXL_POS_ZERO when no position parameter needs validating.
+ *
+ * Returns: 0 on success, -EINVAL on first invalid parameter
+ */
+int cxl_validate_translation_params(u8 eiw, u16 eig, int pos)
 {
-	struct cxl_root_decoder *cxlrd = to_cxl_root_decoder(cxlr->dev.parent);
-	u64 dpa_offset, hpa_offset, bits_upper, mask_upper, hpa;
-	struct cxl_region_params *p = &cxlr->params;
-	struct cxl_endpoint_decoder *cxled = NULL;
-	u16 eig = 0;
-	u8 eiw = 0;
-	int pos;
+	int ways, gran;
 
-	for (int i = 0; i < p->nr_targets; i++) {
-		cxled = p->targets[i];
-		if (cxlmd == cxled_to_memdev(cxled))
-			break;
+	if (eiw_to_ways(eiw, &ways)) {
+		pr_debug("%s: invalid eiw=%u\n", __func__, eiw);
+		return -EINVAL;
+	}
+	if (eig_to_granularity(eig, &gran)) {
+		pr_debug("%s: invalid eig=%u\n", __func__, eig);
+		return -EINVAL;
 	}
-	if (!cxled || cxlmd != cxled_to_memdev(cxled))
+	if (pos < 0 || pos >= ways) {
+		pr_debug("%s: invalid pos=%d for ways=%u\n", __func__, pos,
+			 ways);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_FOR_MODULES(cxl_validate_translation_params, "cxl_translate");
+
+u64 cxl_calculate_dpa_offset(u64 hpa_offset, u8 eiw, u16 eig)
+{
+	u64 dpa_offset, bits_lower, bits_upper, temp;
+	int ret;
+
+	ret = cxl_validate_translation_params(eiw, eig, CXL_POS_ZERO);
+	if (ret)
 		return ULLONG_MAX;
 
-	pos = cxled->pos;
-	ways_to_eiw(p->interleave_ways, &eiw);
-	granularity_to_eig(p->interleave_granularity, &eig);
+	/*
+	 * DPA offset: CXL Spec 3.2 Section 8.2.4.20.13
+	 * Lower bits [IG+7:0] pass through unchanged
+	 * (eiw < 8)
+	 *	Per spec: DPAOffset[51:IG+8] = (HPAOffset[51:IG+IW+8] >> IW)
+	 *	Clear the position bits to isolate upper section, then
+	 *	reverse the left shift by eiw that occurred during DPA->HPA
+	 * (eiw >= 8)
+	 *	Per spec: DPAOffset[51:IG+8] = HPAOffset[51:IG+IW] / 3
+	 *	Extract upper bits from the correct bit range and divide by 3
+	 *	to recover the original DPA upper bits
+	 */
+	bits_lower = hpa_offset & GENMASK_ULL(eig + 7, 0);
+	if (eiw < 8) {
+		temp = hpa_offset &= ~GENMASK_ULL(eig + eiw + 8 - 1, 0);
+		dpa_offset = temp >> eiw;
+	} else {
+		bits_upper = div64_u64(hpa_offset >> (eig + eiw), 3);
+		dpa_offset = bits_upper << (eig + 8);
+	}
+	dpa_offset |= bits_lower;
+
+	return dpa_offset;
+}
+EXPORT_SYMBOL_FOR_MODULES(cxl_calculate_dpa_offset, "cxl_translate");
+
+int cxl_calculate_position(u64 hpa_offset, u8 eiw, u16 eig)
+{
+	unsigned int ways = 0;
+	u64 shifted, rem;
+	int pos, ret;
+
+	ret = cxl_validate_translation_params(eiw, eig, CXL_POS_ZERO);
+	if (ret)
+		return ret;
+
+	if (!eiw)
+		/* position is 0 if no interleaving */
+		return 0;
+
+	/*
+	 * Interleave position: CXL Spec 3.2 Section 8.2.4.20.13
+	 * eiw < 8
+	 *	Position is in the IW bits at HPA_OFFSET[IG+8+IW-1:IG+8].
+	 *	Per spec "remove IW bits starting with bit position IG+8"
+	 * eiw >= 8
+	 *	Position is not explicitly stored in HPA_OFFSET bits. It is
+	 *	derived from the modulo operation of the upper bits using
+	 *	the total number of interleave ways.
+	 */
+	if (eiw < 8) {
+		pos = (hpa_offset >> (eig + 8)) & GENMASK(eiw - 1, 0);
+	} else {
+		shifted = hpa_offset >> (eig + 8);
+		eiw_to_ways(eiw, &ways);
+		div64_u64_rem(shifted, ways, &rem);
+		pos = rem;
+	}
+
+	return pos;
+}
+EXPORT_SYMBOL_FOR_MODULES(cxl_calculate_position, "cxl_translate");
+
+u64 cxl_calculate_hpa_offset(u64 dpa_offset, int pos, u8 eiw, u16 eig)
+{
+	u64 mask_upper, hpa_offset, bits_upper;
+	int ret;
+
+	ret = cxl_validate_translation_params(eiw, eig, pos);
+	if (ret)
+		return ULLONG_MAX;
 
 	/*
 	 * The device position in the region interleave set was removed
@@ -2986,9 +3077,6 @@ u64 cxl_dpa_to_hpa(struct cxl_region *cxlr, const struct cxl_memdev *cxlmd,
 	 * 8.2.4.19.13 Implementation Note: Device Decode Logic
 	 */
 
-	/* Remove the dpa base */
-	dpa_offset = dpa - cxl_dpa_resource_start(cxled);
-
 	mask_upper = GENMASK_ULL(51, eig + 8);
 
 	if (eiw < 8) {
@@ -3003,6 +3091,37 @@ u64 cxl_dpa_to_hpa(struct cxl_region *cxlr, const struct cxl_memdev *cxlmd,
 	/* The lower bits remain unchanged */
 	hpa_offset |= dpa_offset & GENMASK_ULL(eig + 7, 0);
 
+	return hpa_offset;
+}
+EXPORT_SYMBOL_FOR_MODULES(cxl_calculate_hpa_offset, "cxl_translate");
+
+u64 cxl_dpa_to_hpa(struct cxl_region *cxlr, const struct cxl_memdev *cxlmd,
+		   u64 dpa)
+{
+	struct cxl_root_decoder *cxlrd = to_cxl_root_decoder(cxlr->dev.parent);
+	struct cxl_region_params *p = &cxlr->params;
+	struct cxl_endpoint_decoder *cxled = NULL;
+	u64 dpa_offset, hpa_offset, hpa;
+	u16 eig = 0;
+	u8 eiw = 0;
+	int pos;
+
+	for (int i = 0; i < p->nr_targets; i++) {
+		if (cxlmd == cxled_to_memdev(p->targets[i])) {
+			cxled = p->targets[i];
+			break;
+		}
+	}
+	if (!cxled)
+		return ULLONG_MAX;
+
+	pos = cxled->pos;
+	ways_to_eiw(p->interleave_ways, &eiw);
+	granularity_to_eig(p->interleave_granularity, &eig);
+
+	dpa_offset = dpa - cxl_dpa_resource_start(cxled);
+	hpa_offset = cxl_calculate_hpa_offset(dpa_offset, pos, eiw, eig);
+
 	/* Apply the hpa_offset to the region base address */
 	hpa = hpa_offset + p->res->start + p->cache_size;
 
@@ -3035,8 +3154,6 @@ static int region_offset_to_dpa_result(struct cxl_region *cxlr, u64 offset,
 	struct cxl_root_decoder *cxlrd = to_cxl_root_decoder(cxlr->dev.parent);
 	struct cxl_endpoint_decoder *cxled;
 	u64 hpa, hpa_offset, dpa_offset;
-	u64 bits_upper, bits_lower;
-	u64 shifted, rem, temp;
 	u16 eig = 0;
 	u8 eiw = 0;
 	int pos;
@@ -3058,50 +3175,15 @@ static int region_offset_to_dpa_result(struct cxl_region *cxlr, u64 offset,
 	} else {
 		hpa_offset = offset;
 	}
-	/*
-	 * Interleave position: CXL Spec 3.2 Section 8.2.4.20.13
-	 * eiw < 8
-	 *	Position is in the IW bits at HPA_OFFSET[IG+8+IW-1:IG+8].
-	 *	Per spec "remove IW bits starting with bit position IG+8"
-	 * eiw >= 8
-	 *	Position is not explicitly stored in HPA_OFFSET bits. It is
-	 *	derived from the modulo operation of the upper bits using
-	 *	the total number of interleave ways.
-	 */
-	if (eiw < 8) {
-		pos = (hpa_offset >> (eig + 8)) & GENMASK(eiw - 1, 0);
-	} else {
-		shifted = hpa_offset >> (eig + 8);
-		div64_u64_rem(shifted, p->interleave_ways, &rem);
-		pos = rem;
-	}
+
+	pos = cxl_calculate_position(hpa_offset, eiw, eig);
 	if (pos < 0 || pos >= p->nr_targets) {
 		dev_dbg(&cxlr->dev, "Invalid position %d for %d targets\n",
 			pos, p->nr_targets);
 		return -ENXIO;
 	}
 
-	/*
-	 * DPA offset: CXL Spec 3.2 Section 8.2.4.20.13
-	 * Lower bits [IG+7:0] pass through unchanged
-	 * (eiw < 8)
-	 *	Per spec: DPAOffset[51:IG+8] = (HPAOffset[51:IG+IW+8] >> IW)
-	 *	Clear the position bits to isolate upper section, then
-	 *	reverse the left shift by eiw that occurred during DPA->HPA
-	 * (eiw >= 8)
-	 *	Per spec: DPAOffset[51:IG+8] = HPAOffset[51:IG+IW] / 3
-	 *	Extract upper bits from the correct bit range and divide by 3
-	 *	to recover the original DPA upper bits
-	 */
-	bits_lower = hpa_offset & GENMASK_ULL(eig + 7, 0);
-	if (eiw < 8) {
-		temp = hpa_offset &= ~((u64)GENMASK(eig + eiw + 8 - 1, 0));
-		dpa_offset = temp >> eiw;
-	} else {
-		bits_upper = div64_u64(hpa_offset >> (eig + eiw), 3);
-		dpa_offset = bits_upper << (eig + 8);
-	}
-	dpa_offset |= bits_lower;
+	dpa_offset = cxl_calculate_dpa_offset(hpa_offset, eiw, eig);
 
 	/* Look-up and return the result: a memdev and a DPA */
 	for (int i = 0; i < p->nr_targets; i++) {

drivers/cxl/cxl.h

@@ -746,6 +746,25 @@ static inline bool is_cxl_root(struct cxl_port *port)
 	return port->uport_dev == port->dev.parent;
 }
 
+/* Address translation functions exported to cxl_translate test module only */
+int cxl_validate_translation_params(u8 eiw, u16 eig, int pos);
+u64 cxl_calculate_hpa_offset(u64 dpa_offset, int pos, u8 eiw, u16 eig);
+u64 cxl_calculate_dpa_offset(u64 hpa_offset, u8 eiw, u16 eig);
+int cxl_calculate_position(u64 hpa_offset, u8 eiw, u16 eig);
+struct cxl_cxims_data {
+	int nr_maps;
+	u64 xormaps[] __counted_by(nr_maps);
+};
+
+#if IS_ENABLED(CONFIG_CXL_ACPI)
+u64 cxl_do_xormap_calc(struct cxl_cxims_data *cximsd, u64 addr, int hbiw);
+#else
+static inline u64 cxl_do_xormap_calc(struct cxl_cxims_data *cximsd, u64 addr, int hbiw)
+{
+	return ULLONG_MAX;
+}
+#endif
+
 int cxl_num_decoders_committed(struct cxl_port *port);
 bool is_cxl_port(const struct device *dev);
 struct cxl_port *to_cxl_port(const struct device *dev);

tools/testing/cxl/test/Kbuild

@@ -4,6 +4,7 @@ ccflags-y := -I$(srctree)/drivers/cxl/ -I$(srctree)/drivers/cxl/core
 obj-m += cxl_test.o
 obj-m += cxl_mock.o
 obj-m += cxl_mock_mem.o
+obj-m += cxl_translate.o
 
 cxl_test-y := cxl.o
 cxl_mock-y := mock.o