[PATCH 07/18] powerpc/fadump: consider reserved ranges while reserving memory

[Hari Bathini]

Commit 0962e8004e97 ("powerpc/prom: Scan reserved-ranges node for
memory reservations") enabled support to parse reserved-ranges DT
node and reserve kernel memory falling in these ranges for F/W
purposes. Ensure memory in these ranges is not overlapped with
memory reserved for FADump.

Also, use a smaller offset, instead of the size of the memory to
be reserved, by which to skip memory before making another attempt
at reserving memory, after the previous attempt to reserve memory
for FADump failed due to memory holes and/or reserved ranges, to
reduce the likelihood of memory reservation failure.

Signed-off-by: Hari Bathini <hbathini at linux.ibm.com>
---
 arch/powerpc/kernel/fadump.c          |  137 ++++++++++++++++++++++++++++++++-
 arch/powerpc/kernel/fadump_internal.h |   11 +++
 2 files changed, 145 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/kernel/fadump.c b/arch/powerpc/kernel/fadump.c
index 6ff05046..a4510c91 100644
--- a/arch/powerpc/kernel/fadump.c
+++ b/arch/powerpc/kernel/fadump.c
@@ -53,6 +53,9 @@ int crash_memory_ranges_size;
 int crash_mem_ranges;
 int max_crash_mem_ranges;
 
+struct fadump_memory_range reserved_ranges[MAX_RESERVED_RANGES];
+int reserved_ranges_cnt;
+
 #ifdef CONFIG_CMA
 static struct cma *fadump_cma;
 
@@ -116,12 +119,116 @@ int __init fadump_cma_init(void)
 static int __init fadump_cma_init(void) { return 1; }
 #endif /* CONFIG_CMA */
 
+/*
+ * Sort the reserved ranges in-place and merge adjacent ranges
+ * to minimize the reserved ranges count.
+ */
+static void __init sort_and_merge_reserved_ranges(void)
+{
+	unsigned long long base, size;
+	struct fadump_memory_range tmp_range;
+	int i, j, idx;
+
+	if (!reserved_ranges_cnt)
+		return;
+
+	/* Sort the reserved ranges */
+	for (i = 0; i < reserved_ranges_cnt; i++) {
+		idx = i;
+		for (j = i + 1; j < reserved_ranges_cnt; j++) {
+			if (reserved_ranges[idx].base > reserved_ranges[j].base)
+				idx = j;
+		}
+		if (idx != i) {
+			tmp_range = reserved_ranges[idx];
+			reserved_ranges[idx] = reserved_ranges[i];
+			reserved_ranges[i] = tmp_range;
+		}
+	}
+
+	/* Merge adjacent reserved ranges */
+	idx = 0;
+	for (i = 1; i < reserved_ranges_cnt; i++) {
+		base = reserved_ranges[i-1].base;
+		size = reserved_ranges[i-1].size;
+		if (reserved_ranges[i].base == (base + size))
+			reserved_ranges[idx].size += reserved_ranges[i].size;
+		else {
+			idx++;
+			if (i == idx)
+				continue;
+
+			reserved_ranges[idx] = reserved_ranges[i];
+		}
+	}
+	reserved_ranges_cnt = idx + 1;
+}
+
+static int __init add_reserved_range(unsigned long base,
+				     unsigned long size)
+{
+	int i;
+
+	if (reserved_ranges_cnt == MAX_RESERVED_RANGES) {
+		/* Compact reserved ranges and try again. */
+		sort_and_merge_reserved_ranges();
+		if (reserved_ranges_cnt == MAX_RESERVED_RANGES)
+			return 0;
+	}
+
+	i = reserved_ranges_cnt++;
+	reserved_ranges[i].base = base;
+	reserved_ranges[i].size = size;
+	return 1;
+}
+
+/*
+ * Scan reserved-ranges to consider them while reserving/releasing
+ * memory for FADump.
+ */
+static void __init early_init_dt_scan_reserved_ranges(unsigned long node)
+{
+	int len, ret;
+	unsigned long i;
+	const __be32 *prop;
+
+	/* reserved-ranges already scanned */
+	if (reserved_ranges_cnt != 0)
+		return;
+
+	prop = of_get_flat_dt_prop(node, "reserved-ranges", &len);
+
+	if (!prop)
+		return;
+
+	/*
+	 * Each reserved range is an (address,size) pair, 2 cells each,
+	 * totalling 4 cells per range.
+	 */
+	for (i = 0; i < len / (sizeof(*prop) * 4); i++) {
+		u64 base, size;
+
+		base = of_read_number(prop + (i * 4) + 0, 2);
+		size = of_read_number(prop + (i * 4) + 2, 2);
+
+		if (size) {
+			ret = add_reserved_range(base, size);
+			if (ret == 0)
+				pr_warn("some reserved ranges are ignored!\n");
+		}
+	}
+}
+
 /* Scan the Firmware Assisted dump configuration details. */
 int __init early_init_dt_scan_fw_dump(unsigned long node, const char *uname,
 				      int depth, void *data)
 {
-	if (depth != 1)
+	if (depth != 1) {
+		if (depth == 0)
+			early_init_dt_scan_reserved_ranges(node);
+
 		return 0;
+	}
 
 	if (strcmp(uname, "rtas") == 0)
 		return pseries_dt_scan_fadump(&fw_dump, node);
@@ -363,6 +470,26 @@ static int __init fadump_get_rmr_regions(void)
 	return ret;
 }
 
+static bool overlaps_with_reserved_ranges(ulong base, ulong end)
+{
+	int i, ret = 0;
+
+	for (i = 0; i < reserved_ranges_cnt; i++) {
+		ulong rbase = (ulong)reserved_ranges[i].base;
+		ulong rend = rbase + (ulong)reserved_ranges[i].size;
+
+		if (end <= rbase)
+			break;
+
+		if ((end > rbase) &&  (base < rend)) {
+			ret = 1;
+			break;
+		}
+	}
+
+	return ret;
+}
+
 /* Preserve everything above the base address */
 static void __init fadump_reserve_crash_area(unsigned long base)
 {
@@ -399,6 +526,9 @@ int __init fadump_reserve_mem(void)
 		goto error_out;
 	}
 
+	/* Compact reserved ranges */
+	sort_and_merge_reserved_ranges();
+
 	/*
 	 * Initialize boot memory size
 	 * If dump is active then we have already calculated the size during
@@ -474,10 +604,11 @@ int __init fadump_reserve_mem(void)
 		 */
 		while (base <= (memory_boundary - size)) {
 			if (memblock_is_region_memory(base, size) &&
-			    !memblock_is_region_reserved(base, size))
+			    !memblock_is_region_reserved(base, size) &&
+			    !overlaps_with_reserved_ranges(base, (base+size)))
 				break;
 
-			base += size;
+			base += OFFSET_SIZE;
 		}
 
 		if ((base > (memory_boundary - size)) ||
diff --git a/arch/powerpc/kernel/fadump_internal.h b/arch/powerpc/kernel/fadump_internal.h
index 8ad98db..ff764d4 100644
--- a/arch/powerpc/kernel/fadump_internal.h
+++ b/arch/powerpc/kernel/fadump_internal.h
@@ -101,6 +101,17 @@ struct fadump_memory_range {
 	unsigned long long	size;
 };
 
+/*
+ * Amount of memory (1024MB) to skip before making another attempt at
+ * reserving memory (after the previous attempt to reserve memory for
+ * FADump failed due to memory holes and/or reserved ranges) to reduce
+ * the likelihood of memory reservation failure.
+ */
+#define OFFSET_SIZE			0x40000000U
+
+/* Maximum no. of reserved ranges supported for processing. */
+#define MAX_RESERVED_RANGES		128
+
 /* Maximum no. of real memory regions supported by the kernel */
 #define MAX_REAL_MEM_REGIONS		8