Merge tag 'cxl-fixes-6.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl

Pull Compute Express Link (CXL) fixes from Dave Jiang:

- Recognize all ZONE_DEVICE users as physaddr consumers

- Fix format string for extended_linear_cache_size_show()

- Fix target list setup for multiple decoders sharing the same
downstream port

- Restore HBIW check before derefernce platform data

- Fix potential infinite loop in __cxl_dpa_reserve()

- Check for invalid addresses returned from translation functions on
error

* tag 'cxl-fixes-6.19-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl:
cxl: Check for invalid addresses returned from translation functions on errors
cxl/hdm: Fix potential infinite loop in __cxl_dpa_reserve()
cxl/acpi: Restore HBIW check before dereferencing platform_data
cxl/port: Fix target list setup for multiple decoders sharing the same dport
cxl/region: fix format string for resource_size_t
x86/kaslr: Recognize all ZONE_DEVICE users as physaddr consumers

arch/x86/mm/kaslr.c

···
 
 	/*
 	 * Adapt physical memory region size based on available memory,
-	 * except when CONFIG_PCI_P2PDMA is enabled. P2PDMA exposes the
-	 * device BAR space assuming the direct map space is large enough
-	 * for creating a ZONE_DEVICE mapping in the direct map corresponding
-	 * to the physical BAR address.
+	 * except when CONFIG_ZONE_DEVICE is enabled. ZONE_DEVICE wants to map
+	 * any physical address into the direct-map. KASLR wants to reliably
+	 * steal some physical address bits. Those design choices are in direct
+	 * conflict.
 	 */
-	if (!IS_ENABLED(CONFIG_PCI_P2PDMA) && (memory_tb < kaslr_regions[0].size_tb))
+	if (!IS_ENABLED(CONFIG_ZONE_DEVICE) && (memory_tb < kaslr_regions[0].size_tb))
 		kaslr_regions[0].size_tb = memory_tb;
 
 	/*

drivers/cxl/acpi.c

···
 
 static u64 cxl_apply_xor_maps(struct cxl_root_decoder *cxlrd, u64 addr)
 {
-	struct cxl_cxims_data *cximsd = cxlrd->platform_data;
+	int hbiw = cxlrd->cxlsd.nr_targets;
+	struct cxl_cxims_data *cximsd;
 
-	return cxl_do_xormap_calc(cximsd, addr, cxlrd->cxlsd.nr_targets);
+	/* No xormaps for host bridge interleave ways of 1 or 3 */
+	if (hbiw == 1 || hbiw == 3)
+		return addr;
+
+	cximsd = cxlrd->platform_data;
+
+	return cxl_do_xormap_calc(cximsd, addr, hbiw);
 }
 
 struct cxl_cxims_context {

drivers/cxl/core/hdm.c

···
 	 * is not set.
 	 */
 	if (cxled->part < 0)
-		for (int i = 0; cxlds->nr_partitions; i++)
+		for (int i = 0; i < cxlds->nr_partitions; i++)
 			if (resource_contains(&cxlds->part[i].res, res)) {
 				cxled->part = i;
 				break;
···
 
 resource_size_t cxl_dpa_resource_start(struct cxl_endpoint_decoder *cxled)
 {
-	resource_size_t base = -1;
+	resource_size_t base = RESOURCE_SIZE_MAX;
 
 	lockdep_assert_held(&cxl_rwsem.dpa);
 	if (cxled->dpa_res)

drivers/cxl/core/port.c

···
 			cxlsd->target[i] = dport;
 			dev_dbg(dev, "dport%d found in target list, index %d\n",
 				dport->port_id, i);
-			return 1;
+			return 0;
 		}
 	}
 

drivers/cxl/core/region.c

···
 	ACQUIRE(rwsem_read_intr, rwsem)(&cxl_rwsem.region);
 	if ((rc = ACQUIRE_ERR(rwsem_read_intr, &rwsem)))
 		return rc;
-	return sysfs_emit(buf, "%#llx\n", p->cache_size);
+	return sysfs_emit(buf, "%pap\n", &p->cache_size);
 }
 static DEVICE_ATTR_RO(extended_linear_cache_size);
 
···
 	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;
+	u64 base, dpa_offset, hpa_offset, hpa;
 	u16 eig = 0;
 	u8 eiw = 0;
 	int pos;
···
 	ways_to_eiw(p->interleave_ways, &eiw);
 	granularity_to_eig(p->interleave_granularity, &eig);
 
-	dpa_offset = dpa - cxl_dpa_resource_start(cxled);
+	base = cxl_dpa_resource_start(cxled);
+	if (base == RESOURCE_SIZE_MAX)
+		return ULLONG_MAX;
+
+	dpa_offset = dpa - base;
 	hpa_offset = cxl_calculate_hpa_offset(dpa_offset, pos, eiw, eig);
+	if (hpa_offset == ULLONG_MAX)
+		return ULLONG_MAX;
 
 	/* Apply the hpa_offset to the region base address */
 	hpa = hpa_offset + p->res->start + p->cache_size;
···
 	/* Root decoder translation overrides typical modulo decode */
 	if (cxlrd->ops.hpa_to_spa)
 		hpa = cxlrd->ops.hpa_to_spa(cxlrd, hpa);
+
+	if (hpa == ULLONG_MAX)
+		return ULLONG_MAX;
 
 	if (!cxl_resource_contains_addr(p->res, hpa)) {
 		dev_dbg(&cxlr->dev,
···
 	struct cxl_region_params *p = &cxlr->params;
 	struct cxl_root_decoder *cxlrd = to_cxl_root_decoder(cxlr->dev.parent);
 	struct cxl_endpoint_decoder *cxled;
-	u64 hpa, hpa_offset, dpa_offset;
+	u64 hpa_offset = offset;
+	u64 dpa, dpa_offset;
 	u16 eig = 0;
 	u8 eiw = 0;
 	int pos;
···
 	 * CXL HPA is assumed to equal SPA.
 	 */
 	if (cxlrd->ops.spa_to_hpa) {
-		hpa = cxlrd->ops.spa_to_hpa(cxlrd, p->res->start + offset);
-		hpa_offset = hpa - p->res->start;
-	} else {
-		hpa_offset = offset;
+		hpa_offset = cxlrd->ops.spa_to_hpa(cxlrd, p->res->start + offset);
+		if (hpa_offset == ULLONG_MAX) {
+			dev_dbg(&cxlr->dev, "HPA not found for %pr offset %#llx\n",
+				p->res, offset);
+			return -ENXIO;
+		}
+		hpa_offset -= p->res->start;
 	}
 
 	pos = cxl_calculate_position(hpa_offset, eiw, eig);
···
 		cxled = p->targets[i];
 		if (cxled->pos != pos)
 			continue;
+
+		dpa = cxl_dpa_resource_start(cxled);
+		if (dpa != RESOURCE_SIZE_MAX)
+			dpa += dpa_offset;
+
 		result->cxlmd = cxled_to_memdev(cxled);
-		result->dpa = cxl_dpa_resource_start(cxled) + dpa_offset;
+		result->dpa = dpa;
 
 		return 0;
 	}

drivers/pci/Kconfig

···
 	  P2P DMA transactions must be between devices behind the same root
 	  port.
 
-	  Enabling this option will reduce the entropy of x86 KASLR memory
-	  regions. For example - on a 46 bit system, the entropy goes down
-	  from 16 bits to 15 bits. The actual reduction in entropy depends
-	  on the physical address bits, on processor features, kernel config
-	  (5 level page table) and physical memory present on the system.
-
 	  If unsure, say N.
 
 config PCI_LABEL

mm/Kconfig

···
 	  Device memory hotplug support allows for establishing pmem,
 	  or other device driver discovered memory regions, in the
 	  memmap. This allows pfn_to_page() lookups of otherwise
-	  "device-physical" addresses which is needed for using a DAX
-	  mapping in an O_DIRECT operation, among other things.
+	  "device-physical" addresses which is needed for DAX, PCI_P2PDMA, and
+	  DEVICE_PRIVATE features among others.
 
-	  If FS_DAX is enabled, then say Y.
+	  Enabling this option will reduce the entropy of x86 KASLR memory
+	  regions. For example - on a 46 bit system, the entropy goes down
+	  from 16 bits to 15 bits. The actual reduction in entropy depends
+	  on the physical address bits, on processor features, kernel config
+	  (5 level page table) and physical memory present on the system.
 
 #
 # Helpers to mirror range of the CPU page tables of a process into device page

tools/testing/cxl/test/cxl_translate.c

···
 
 	/* Calculate base HPA offset from DPA and position */
 	hpa_offset = cxl_calculate_hpa_offset(dpa_offset, pos, r_eiw, r_eig);
+	if (hpa_offset == ULLONG_MAX)
+		return ULLONG_MAX;
 
 	if (math == XOR_MATH) {
 		cximsd->nr_maps = hbiw_to_nr_maps[hb_ways];
···
 		pos = get_random_u32() % ways;
 		dpa = get_random_u64() >> 12;
 
+		reverse_dpa = ULLONG_MAX;
+		reverse_pos = -1;
+
 		hpa = cxl_calculate_hpa_offset(dpa, pos, eiw, eig);
-		reverse_dpa = cxl_calculate_dpa_offset(hpa, eiw, eig);
-		reverse_pos = cxl_calculate_position(hpa, eiw, eig);
+		if (hpa != ULLONG_MAX) {
+			reverse_dpa = cxl_calculate_dpa_offset(hpa, eiw, eig);
+			reverse_pos = cxl_calculate_position(hpa, eiw, eig);
+			if (reverse_dpa == dpa && reverse_pos == pos)
+				continue;
+		}
 
-		if (reverse_dpa != dpa || reverse_pos != pos) {
-			pr_err("test random iter %d FAIL hpa=%llu, dpa=%llu reverse_dpa=%llu, pos=%d reverse_pos=%d eiw=%u eig=%u\n",
-			       i, hpa, dpa, reverse_dpa, pos, reverse_pos, eiw,
-			       eig);
+		pr_err("test random iter %d FAIL hpa=%llu, dpa=%llu reverse_dpa=%llu, pos=%d reverse_pos=%d eiw=%u eig=%u\n",
+		       i, hpa, dpa, reverse_dpa, pos, reverse_pos, eiw, eig);
 
-			if (failures++ > 10) {
-				pr_err("test random too many failures, stop\n");
-				break;
-			}
+		if (failures++ > 10) {
+			pr_err("test random too many failures, stop\n");
+			break;
 		}
 	}
 	pr_info("..... test random: PASS %d FAIL %d\n", i - failures, failures);