tests: Add lib/lzma-test test case

Files used by this test are in: tests/data/lib/lzma-test/
file.bin - files with uncompressed data
file.lzma.bin - files with LZMA-compressed data from file.bin

How to prepare compressed file:
  util/cbfs-compression-tool compress file.bin /tmp/file.lzma.bin lzma
  dd if=/tmp/file.lzma.bin of=file.lzma.bin skip=8 ibs=1

Signed-off-by: Jakub Czapiga <jacz@semihalf.com>
Change-Id: Id75e0b41991382d4c391b031862106de58eacdf7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57555
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Fagerburg <pfagerburg@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>

tests/Makefile.inc

@@ -141,7 +141,8 @@ $$($(1)-config-file): $(TEST_KCONFIG_AUTOHEADER)
 
 $($(1)-objs): TEST_CFLAGS += -I$$(dir $$($(1)-config-file)) \
 	-D__$$(shell echo $$($(1)-stage) | tr '[:lower:]' '[:upper:]')__ \
-	-D__TEST_NAME__=\"$(subst /,_,$(1))\"
+	-D__TEST_NAME__=\"$(subst /,_,$(1))\" \
+	-D__TEST_DATA_DIR__=\"$(testsrc)/data\"
 
 # Give us a way to distinguish between coreboot source files and test files in code.
 $($(1)-srcobjs): TEST_CFLAGS += -D__TEST_SRCOBJ__

tests/data/lib/lzma-test/data.2.bin

@@ -0,0 +1,104 @@
+coreboot README
+===============
+
+coreboot is a Free Software project aimed at replacing the proprietary BIOS
+(firmware) found in most computers.  coreboot performs a little bit of
+hardware initialization and then executes additional boot logic, called a
+payload.
+
+With the separation of hardware initialization and later boot logic,
+coreboot can scale from specialized applications that run directly
+firmware, run operating systems in flash, load custom
+bootloaders, or implement firmware standards, like PC BIOS services or
+UEFI. This allows for systems to only include the features necessary
+in the target application, reducing the amount of code and flash space
+required.
+
+coreboot was formerly known as LinuxBIOS.
+
+
+Payloads
+--------
+
+After the basic initialization of the hardware has been performed, any
+desired "payload" can be started by coreboot.
+
+See <https://www.coreboot.org/Payloads> for a list of supported payloads.
+
+
+Supported Hardware
+------------------
+
+coreboot supports a wide range of chipsets, devices, and mainboards.
+
+For details please consult:
+
+ * <https://www.coreboot.org/Supported_Motherboards>
+
+
+Build Requirements
+------------------
+
+ * make
+ * gcc / g++
+   Because Linux distribution compilers tend to use lots of patches. coreboot
+   does lots of "unusual" things in its build system, some of which break due
+   to those patches, sometimes by gcc aborting, sometimes - and that's worse -
+   by generating broken object code.
+   Two options: use our toolchain (eg. make crosstools-i386) or enable the
+   `ANY_TOOLCHAIN` Kconfig option if you're feeling lucky (no support in this
+   case).
+ * iasl (for targets with ACPI support)
+ * pkg-config
+ * libssl-dev (openssl)
+
+Optional:
+
+ * doxygen (for generating/viewing documentation)
+ * gdb (for better debugging facilities on some targets)
+ * ncurses (for `make menuconfig` and `make nconfig`)
+ * flex and bison (for regenerating parsers)
+
+
+Building coreboot
+-----------------
+
+Please consult <https://www.coreboot.org/Build_HOWTO> for details.
+
+
+Testing coreboot Without Modifying Your Hardware
+------------------------------------------------
+
+If you want to test coreboot without any risks before you really decide
+to use it on your hardware, you can use the QEMU system emulator to run
+coreboot virtually in QEMU.
+
+Please see <https://www.coreboot.org/QEMU> for details.
+
+
+Website and Mailing List
+------------------------
+
+Further details on the project, a FAQ, many HOWTOs, news, development
+guidelines and more can be found on the coreboot website:
+
+  <https://www.coreboot.org>
+
+You can contact us directly on the coreboot mailing list:
+
+  <https://www.coreboot.org/Mailinglist>
+
+
+Copyright and License
+---------------------
+
+The copyright on coreboot is owned by quite a large number of individual
+developers and companies. Please check the individual source files for details.
+
+coreboot is licensed under the terms of the GNU General Public License (GPL).
+Some files are licensed under the "GPL (version 2, or any later version)",
+and some files are licensed under the "GPL, version 2". For some parts, which
+were derived from other projects, other (GPL-compatible) licenses may apply.
+Please check the individual source files for details.
+
+This makes the resulting coreboot images licensed under the GPL, version 2.

tests/data/lib/lzma-test/data.3.bin

@@ -0,0 +1,764 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#include <stdlib.h>
+#include <types.h>
+#include <string.h>
+#include <tests/test.h>
+#include <imd.h>
+#include <imd_private.h>
+#include <cbmem.h>
+#include <commonlib/bsd/helpers.h>
+#include <lib.h>
+
+/* Auxiliary functions and definitions. */
+
+#define LG_ROOT_SIZE align_up_pow2(sizeof(struct imd_root_pointer) +\
+	 sizeof(struct imd_root) + 3 * sizeof(struct imd_entry))
+#define LG_ENTRY_ALIGN (2 * sizeof(int32_t))
+#define LG_ENTRY_SIZE (2 * sizeof(int32_t))
+#define LG_ENTRY_ID 0xA001
+
+#define SM_ROOT_SIZE LG_ROOT_SIZE
+#define SM_ENTRY_ALIGN sizeof(uint32_t)
+#define SM_ENTRY_SIZE sizeof(uint32_t)
+#define SM_ENTRY_ID 0xB001
+
+#define INVALID_REGION_ID 0xC001
+
+static uint32_t align_up_pow2(uint32_t x)
+{
+	return (1 << log2_ceil(x));
+}
+
+static size_t max_entries(size_t root_size)
+{
+	return (root_size - sizeof(struct imd_root_pointer) - sizeof(struct imd_root))
+			/ sizeof(struct imd_entry);
+}
+
+/*
+ * Mainly, we should check that imd_handle_init() aligns upper_limit properly
+ * for various inputs. Upper limit is the _exclusive_ address, so we expect
+ * ALIGN_DOWN.
+ */
+static void test_imd_handle_init(void **state)
+{
+	int i;
+	void *base;
+	struct imd imd;
+	uintptr_t test_inputs[] = {
+			0,                   /* Lowest possible address */
+			0xA000,              /* Fits in 16 bits, should not get rounded down*/
+			0xDEAA,              /* Fits in 16 bits */
+			0xB0B0B000,          /* Fits in 32 bits, should not get rounded down */
+			0xF0F0F0F0,          /* Fits in 32 bits */
+			((1ULL << 32) + 4),  /* Just above 32-bit limit */
+			0x6666777788889000,  /* Fits in 64 bits, should not get rounded down */
+			((1ULL << 60) - 100) /* Very large address, fitting in 64 bits */
+	};
+
+	for (i = 0; i < ARRAY_SIZE(test_inputs); i++) {
+		base = (void *)test_inputs[i];
+
+		imd_handle_init(&imd, (void *)base);
+
+		assert_int_equal(imd.lg.limit % LIMIT_ALIGN, 0);
+		assert_int_equal(imd.lg.limit, ALIGN_DOWN(test_inputs[i], LIMIT_ALIGN));
+		assert_ptr_equal(imd.lg.r, NULL);
+
+		/* Small allocations not initialized */
+		assert_ptr_equal(imd.sm.limit, NULL);
+		assert_ptr_equal(imd.sm.r, NULL);
+	}
+}
+
+static void test_imd_handle_init_partial_recovery(void **state)
+{
+	void *base;
+	struct imd imd = {0};
+	const struct imd_entry *entry;
+
+	imd_handle_init_partial_recovery(&imd);
+	assert_null(imd.lg.limit);
+	assert_null(imd.sm.limit);
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+	imd_handle_init_partial_recovery(&imd);
+
+	assert_non_null(imd.lg.r);
+	assert_null(imd.sm.limit);
+
+	assert_int_equal(0, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+	entry = imd_entry_add(&imd, SMALL_REGION_ID, LG_ENTRY_SIZE);
+	assert_non_null(entry);
+
+	imd_handle_init_partial_recovery(&imd);
+
+	assert_non_null(imd.lg.r);
+	assert_non_null(imd.sm.limit);
+	assert_ptr_equal(imd.lg.r + entry->start_offset + LG_ENTRY_SIZE, imd.sm.limit);
+	assert_non_null(imd.sm.r);
+
+	free(base);
+}
+
+static void test_imd_create_empty(void **state)
+{
+	struct imd imd = {0};
+	void *base;
+	struct imd_root *r;
+	struct imd_entry *e;
+
+	/* Expect imd_create_empty to fail, since imd handle is not initialized */
+	assert_int_equal(-1, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+	base = malloc(sizeof(struct imd_root_pointer) + sizeof(struct imd_root));
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	/* Try incorrect sizes */
+	assert_int_equal(-1, imd_create_empty(&imd,
+					sizeof(struct imd_root_pointer),
+					LG_ENTRY_ALIGN));
+	assert_int_equal(-1, imd_create_empty(&imd, LG_ROOT_SIZE, 2 * LG_ROOT_SIZE));
+
+	/* Working case */
+	assert_int_equal(0, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+
+	/* Only large allocation initialized with one entry for the root region */
+	r = (struct imd_root *) (imd.lg.r);
+	assert_non_null(r);
+
+	e = &r->entries[r->num_entries - 1];
+
+	assert_int_equal(max_entries(LG_ROOT_SIZE), r->max_entries);
+	assert_int_equal(1, r->num_entries);
+	assert_int_equal(0, r->flags);
+	assert_int_equal(LG_ENTRY_ALIGN, r->entry_align);
+	assert_int_equal(0, r->max_offset);
+	assert_ptr_equal(e, &r->entries);
+
+	assert_int_equal(IMD_ENTRY_MAGIC, e->magic);
+	assert_int_equal(0, e->start_offset);
+	assert_int_equal(LG_ROOT_SIZE, e->size);
+	assert_int_equal(CBMEM_ID_IMD_ROOT, e->id);
+
+	free(base);
+}
+
+static void test_imd_create_tiered_empty(void **state)
+{
+	void *base;
+	size_t sm_region_size, lg_region_wrong_size;
+	struct imd imd = {0};
+	struct imd_root *r;
+	struct imd_entry *fst_lg_entry, *snd_lg_entry, *sm_entry;
+
+	/* Uninitialized imd handle */
+	assert_int_equal(-1, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+						     LG_ROOT_SIZE, SM_ENTRY_ALIGN));
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	/* Too small root_size for small region */
+	assert_int_equal(-1, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+			 sizeof(int32_t), 2 * sizeof(int32_t)));
+
+	/* Fail when large region doesn't have capacity for more than 1 entry */
+	lg_region_wrong_size = sizeof(struct imd_root_pointer) + sizeof(struct imd_root) +
+			       sizeof(struct imd_entry);
+	expect_assert_failure(
+		imd_create_tiered_empty(&imd, lg_region_wrong_size, LG_ENTRY_ALIGN,
+					SM_ROOT_SIZE, SM_ENTRY_ALIGN)
+	);
+
+	assert_int_equal(0, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+						    SM_ROOT_SIZE, SM_ENTRY_ALIGN));
+
+	r = imd.lg.r;
+
+	/* One entry for root_region and one for small allocations */
+	assert_int_equal(2, r->num_entries);
+
+	fst_lg_entry = &r->entries[0];
+	assert_int_equal(IMD_ENTRY_MAGIC, fst_lg_entry->magic);
+	assert_int_equal(0, fst_lg_entry->start_offset);
+	assert_int_equal(LG_ROOT_SIZE, fst_lg_entry->size);
+	assert_int_equal(CBMEM_ID_IMD_ROOT, fst_lg_entry->id);
+
+	/* Calculated like in imd_create_tiered_empty */
+	sm_region_size = max_entries(SM_ROOT_SIZE) * SM_ENTRY_ALIGN;
+	sm_region_size += SM_ROOT_SIZE;
+	sm_region_size = ALIGN_UP(sm_region_size, LG_ENTRY_ALIGN);
+
+	snd_lg_entry = &r->entries[1];
+	assert_int_equal(IMD_ENTRY_MAGIC, snd_lg_entry->magic);
+	assert_int_equal(-sm_region_size, snd_lg_entry->start_offset);
+	assert_int_equal(CBMEM_ID_IMD_SMALL, snd_lg_entry->id);
+
+	assert_int_equal(sm_region_size, snd_lg_entry->size);
+
+	r = imd.sm.r;
+	assert_int_equal(1, r->num_entries);
+
+	sm_entry = &r->entries[0];
+	assert_int_equal(IMD_ENTRY_MAGIC, sm_entry->magic);
+	assert_int_equal(0, sm_entry->start_offset);
+	assert_int_equal(SM_ROOT_SIZE, sm_entry->size);
+	assert_int_equal(CBMEM_ID_IMD_ROOT, sm_entry->id);
+
+	free(base);
+}
+
+/* Tests for imdr_recover. */
+static void test_imd_recover(void **state)
+{
+	int32_t offset_copy, max_offset_copy;
+	uint32_t rp_magic_copy, num_entries_copy;
+	uint32_t e_align_copy, e_magic_copy, e_id_copy;
+	uint32_t size_copy, diff;
+	void *base;
+	struct imd imd = {0};
+	struct imd_root_pointer *rp;
+	struct imd_root *r;
+	struct imd_entry *lg_root_entry, *sm_root_entry,  *ptr;
+	const struct imd_entry *lg_entry;
+
+	/* Fail when the limit for lg was not set. */
+	imd.lg.limit = (uintptr_t) NULL;
+	assert_int_equal(-1, imd_recover(&imd));
+
+	/* Set the limit for lg. */
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	/* Fail when the root pointer is not valid. */
+	rp = (void *)imd.lg.limit - sizeof(struct imd_root_pointer);
+	assert_non_null(rp);
+	assert_int_equal(IMD_ROOT_PTR_MAGIC, rp->magic);
+
+	rp_magic_copy = rp->magic;
+	rp->magic = 0;
+	assert_int_equal(-1, imd_recover(&imd));
+	rp->magic = rp_magic_copy;
+
+	/* Set the root pointer. */
+	assert_int_equal(0, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+						    SM_ROOT_SIZE, SM_ENTRY_ALIGN));
+	assert_int_equal(2, ((struct imd_root *)imd.lg.r)->num_entries);
+	assert_int_equal(1, ((struct imd_root *)imd.sm.r)->num_entries);
+
+	/* Fail if the number of entries exceeds the maximum number of entries. */
+	r = imd.lg.r;
+	num_entries_copy = r->num_entries;
+	r->num_entries = r->max_entries + 1;
+	assert_int_equal(-1, imd_recover(&imd));
+	r->num_entries = num_entries_copy;
+
+	/* Fail if entry align is not a power of 2.  */
+	e_align_copy = r->entry_align;
+	r->entry_align++;
+	assert_int_equal(-1, imd_recover(&imd));
+	r->entry_align = e_align_copy;
+
+	/* Fail when an entry is not valid. */
+	lg_root_entry = &r->entries[0];
+	e_magic_copy = lg_root_entry->magic;
+	lg_root_entry->magic = 0;
+	assert_int_equal(-1, imd_recover(&imd));
+	lg_root_entry->magic = e_magic_copy;
+
+	/* Add new entries: large and small. */
+	lg_entry = imd_entry_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE);
+	assert_non_null(lg_entry);
+	assert_int_equal(3, r->num_entries);
+
+	assert_non_null(imd_entry_add(&imd, SM_ENTRY_ID, SM_ENTRY_SIZE));
+	assert_int_equal(2, ((struct imd_root *)imd.sm.r)->num_entries);
+
+	/* Fail when start_addr is lower than low_limit. */
+	r = imd.lg.r;
+	max_offset_copy = r->max_offset;
+	r->max_offset = lg_entry->start_offset + sizeof(int32_t);
+	assert_int_equal(-1, imd_recover(&imd));
+	r->max_offset = max_offset_copy;
+
+	/* Fail when start_addr is at least imdr->limit. */
+	offset_copy = lg_entry->start_offset;
+	ptr = (struct imd_entry *)lg_entry;
+	ptr->start_offset = (void *)imd.lg.limit - (void *)r;
+	assert_int_equal(-1, imd_recover(&imd));
+	ptr->start_offset = offset_copy;
+
+	/* Fail when (start_addr + e->size) is higher than imdr->limit. */
+	size_copy = lg_entry->size;
+	diff = (void *)imd.lg.limit - ((void *)r + lg_entry->start_offset);
+	ptr->size = diff + 1;
+	assert_int_equal(-1, imd_recover(&imd));
+	ptr->size = size_copy;
+
+	/* Succeed if small region is not present. */
+	sm_root_entry = &r->entries[1];
+	e_id_copy = sm_root_entry->id;
+	sm_root_entry->id = 0;
+	assert_int_equal(0, imd_recover(&imd));
+	sm_root_entry->id = e_id_copy;
+
+	assert_int_equal(0, imd_recover(&imd));
+
+	free(base);
+}
+
+static void test_imd_limit_size(void **state)
+{
+	void *base;
+	struct imd imd = {0};
+	size_t root_size, max_size;
+
+	max_size = align_up_pow2(sizeof(struct imd_root_pointer)
+			+ sizeof(struct imd_root) + 3 * sizeof(struct imd_entry));
+
+	assert_int_equal(-1, imd_limit_size(&imd, max_size));
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	root_size = align_up_pow2(sizeof(struct imd_root_pointer)
+			+ sizeof(struct imd_root) + 2 * sizeof(struct imd_entry));
+	imd.lg.r = (void *)imd.lg.limit - root_size;
+
+	imd_create_empty(&imd, root_size, LG_ENTRY_ALIGN);
+	assert_int_equal(-1, imd_limit_size(&imd, root_size - 1));
+	assert_int_equal(0, imd_limit_size(&imd, max_size));
+
+	/* Cannot create such a big entry */
+	assert_null(imd_entry_add(&imd, LG_ENTRY_ID, max_size - root_size + 1));
+
+	free(base);
+}
+
+static void test_imd_lockdown(void **state)
+{
+	struct imd imd = {0};
+	struct imd_root *r_lg, *r_sm;
+
+	assert_int_equal(-1, imd_lockdown(&imd));
+
+	imd.lg.r = malloc(sizeof(struct imd_root));
+	if (imd.lg.r == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	r_lg = (struct imd_root *) (imd.lg.r);
+
+	assert_int_equal(0, imd_lockdown(&imd));
+	assert_true(r_lg->flags & IMD_FLAG_LOCKED);
+
+	imd.sm.r = malloc(sizeof(struct imd_root));
+	if (imd.sm.r == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	r_sm = (struct imd_root *) (imd.sm.r);
+
+	assert_int_equal(0, imd_lockdown(&imd));
+	assert_true(r_sm->flags & IMD_FLAG_LOCKED);
+
+	free(imd.lg.r);
+	free(imd.sm.r);
+}
+
+static void test_imd_region_used(void **state)
+{
+	struct imd imd = {0};
+	struct imd_entry *first_entry, *new_entry;
+	struct imd_root *r;
+	size_t size;
+	void *imd_base;
+	void *base;
+
+	assert_int_equal(-1, imd_region_used(&imd, &base, &size));
+
+	imd_base = malloc(LIMIT_ALIGN);
+	if (imd_base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)imd_base));
+
+	assert_int_equal(-1, imd_region_used(&imd, &base, &size));
+	assert_int_equal(0, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+	assert_int_equal(0, imd_region_used(&imd, &base, &size));
+
+	r = (struct imd_root *)imd.lg.r;
+	first_entry = &r->entries[r->num_entries - 1];
+
+	assert_int_equal(r + first_entry->start_offset, (uintptr_t)base);
+	assert_int_equal(first_entry->size, size);
+
+	assert_non_null(imd_entry_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE));
+	assert_int_equal(2, r->num_entries);
+
+	assert_int_equal(0, imd_region_used(&imd, &base, &size));
+
+	new_entry = &r->entries[r->num_entries - 1];
+
+	assert_true((void *)r + new_entry->start_offset == base);
+	assert_int_equal(first_entry->size + new_entry->size, size);
+
+	free(imd_base);
+}
+
+static void test_imd_entry_add(void **state)
+{
+	int i;
+	struct imd imd = {0};
+	size_t entry_size = 0;
+	size_t used_size;
+	ssize_t entry_offset;
+	void *base;
+	struct imd_root *r, *sm_r, *lg_r;
+	struct imd_entry *first_entry, *new_entry;
+	uint32_t num_entries_copy;
+	int32_t max_offset_copy;
+
+	/* No small region case. */
+	assert_null(imd_entry_add(&imd, LG_ENTRY_ID, entry_size));
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	assert_int_equal(0, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+
+	r = (struct imd_root *)imd.lg.r;
+	first_entry = &r->entries[r->num_entries - 1];
+
+	/* Cannot add an entry when root is locked. */
+	r->flags = IMD_FLAG_LOCKED;
+	assert_null(imd_entry_add(&imd, LG_ENTRY_ID, entry_size));
+	r->flags = 0;
+
+	/* Fail when the maximum number of entries has been reached. */
+	num_entries_copy = r->num_entries;
+	r->num_entries = r->max_entries;
+	assert_null(imd_entry_add(&imd, LG_ENTRY_ID, entry_size));
+	r->num_entries = num_entries_copy;
+
+	/* Fail when entry size is 0 */
+	assert_null(imd_entry_add(&imd, LG_ENTRY_ID, 0));
+
+	/* Fail when entry size (after alignment) overflows imd total size. */
+	entry_size = 2049;
+	max_offset_copy = r->max_offset;
+	r->max_offset = -entry_size;
+	assert_null(imd_entry_add(&imd, LG_ENTRY_ID, entry_size));
+	r->max_offset = max_offset_copy;
+
+	/* Finally succeed. */
+	entry_size = 2 * sizeof(int32_t);
+	assert_non_null(imd_entry_add(&imd, LG_ENTRY_ID, entry_size));
+	assert_int_equal(2, r->num_entries);
+
+	new_entry = &r->entries[r->num_entries - 1];
+	assert_int_equal(sizeof(struct imd_entry), (void *)new_entry - (void *)first_entry);
+
+	assert_int_equal(IMD_ENTRY_MAGIC, new_entry->magic);
+	assert_int_equal(LG_ENTRY_ID, new_entry->id);
+	assert_int_equal(entry_size, new_entry->size);
+
+	used_size = ALIGN_UP(entry_size, r->entry_align);
+	entry_offset = first_entry->start_offset - used_size;
+	assert_int_equal(entry_offset, new_entry->start_offset);
+
+	/* Use small region case. */
+	imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN, SM_ROOT_SIZE,
+				SM_ENTRY_ALIGN);
+
+	lg_r = imd.lg.r;
+	sm_r = imd.sm.r;
+
+	/* All five new entries should be added to small allocations */
+	for (i = 0; i < 5; i++) {
+		assert_non_null(imd_entry_add(&imd, SM_ENTRY_ID, SM_ENTRY_SIZE));
+		assert_int_equal(i+2, sm_r->num_entries);
+		assert_int_equal(2, lg_r->num_entries);
+	}
+
+	/* But next should fall back on large region */
+	assert_non_null(imd_entry_add(&imd, SM_ENTRY_ID, SM_ENTRY_SIZE));
+	assert_int_equal(6, sm_r->num_entries);
+	assert_int_equal(3, lg_r->num_entries);
+
+	/*
+	 * Small allocation is created when occupies less than 1/4 of available
+	 * small region. Verify this.
+	 */
+	imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN, SM_ROOT_SIZE,
+				SM_ENTRY_ALIGN);
+
+	assert_non_null(imd_entry_add(&imd, SM_ENTRY_ID, -sm_r->max_offset / 4 + 1));
+	assert_int_equal(1, sm_r->num_entries);
+	assert_int_equal(3, lg_r->num_entries);
+
+	/* Next two should go into small region */
+	assert_non_null(imd_entry_add(&imd, SM_ENTRY_ID, -sm_r->max_offset / 4));
+	assert_int_equal(2, sm_r->num_entries);
+	assert_int_equal(3, lg_r->num_entries);
+
+	/* (1/4 * 3/4) */
+	assert_non_null(imd_entry_add(&imd, SM_ENTRY_ID, -sm_r->max_offset / 16 * 3));
+	assert_int_equal(3, sm_r->num_entries);
+	assert_int_equal(3, lg_r->num_entries);
+
+	free(base);
+}
+
+static void test_imd_entry_find(void **state)
+{
+	struct imd imd = {0};
+	void *base;
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	assert_int_equal(0, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+						    SM_ROOT_SIZE, SM_ENTRY_ALIGN));
+
+	assert_non_null(imd_entry_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE));
+
+	assert_non_null(imd_entry_find(&imd, LG_ENTRY_ID));
+	assert_non_null(imd_entry_find(&imd, SMALL_REGION_ID));
+
+	/* Try invalid id, should fail */
+	assert_null(imd_entry_find(&imd, INVALID_REGION_ID));
+
+	free(base);
+}
+
+static void test_imd_entry_find_or_add(void **state)
+{
+	struct imd imd = {0};
+	const struct imd_entry *entry;
+	struct imd_root *r;
+	void *base;
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	assert_null(imd_entry_find_or_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE));
+
+	assert_int_equal(0, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+	entry = imd_entry_find_or_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE);
+	assert_non_null(entry);
+
+	r = (struct imd_root *)imd.lg.r;
+
+	assert_int_equal(entry->id, LG_ENTRY_ID);
+	assert_int_equal(2, r->num_entries);
+	assert_non_null(imd_entry_find_or_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE));
+	assert_int_equal(2, r->num_entries);
+
+	free(base);
+}
+
+static void test_imd_entry_size(void **state)
+{
+	struct imd_entry entry = { .size =  LG_ENTRY_SIZE };
+
+	assert_int_equal(LG_ENTRY_SIZE, imd_entry_size(&entry));
+
+	entry.size = 0;
+	assert_int_equal(0, imd_entry_size(&entry));
+}
+
+static void test_imd_entry_at(void **state)
+{
+	struct imd imd = {0};
+	struct imd_root *r;
+	struct imd_entry *e = NULL;
+	const struct imd_entry *entry;
+	void *base;
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	assert_int_equal(0, imd_create_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN));
+
+	/* Fail when entry is NULL */
+	assert_null(imd_entry_at(&imd, e));
+
+	entry = imd_entry_add(&imd, LG_ENTRY_ID, LG_ENTRY_SIZE);
+	assert_non_null(entry);
+
+	r = (struct imd_root *)imd.lg.r;
+	assert_ptr_equal((void *)r + entry->start_offset, imd_entry_at(&imd, entry));
+
+	free(base);
+}
+
+static void test_imd_entry_id(void **state)
+{
+	struct imd_entry entry = { .id =  LG_ENTRY_ID };
+
+	assert_int_equal(LG_ENTRY_ID, imd_entry_id(&entry));
+}
+
+static void test_imd_entry_remove(void **state)
+{
+	void *base;
+	struct imd imd = {0};
+	struct imd_root *r;
+	const struct imd_entry *fst_lg_entry, *snd_lg_entry, *fst_sm_entry;
+	const struct imd_entry *e = NULL;
+
+	/* Uninitialized handle */
+	assert_int_equal(-1, imd_entry_remove(&imd, e));
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	assert_int_equal(0, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+						    SM_ROOT_SIZE, SM_ENTRY_ALIGN));
+
+	r = imd.lg.r;
+	assert_int_equal(2, r->num_entries);
+	fst_lg_entry = &r->entries[0];
+	snd_lg_entry = &r->entries[1];
+
+	/* Only last entry can be removed */
+	assert_int_equal(-1, imd_entry_remove(&imd, fst_lg_entry));
+	r->flags = IMD_FLAG_LOCKED;
+	assert_int_equal(-1, imd_entry_remove(&imd, snd_lg_entry));
+	r->flags = 0;
+
+	r = imd.sm.r;
+	assert_int_equal(1, r->num_entries);
+	fst_sm_entry = &r->entries[0];
+
+	/* Fail trying to remove root entry */
+	assert_int_equal(-1, imd_entry_remove(&imd, fst_sm_entry));
+	assert_int_equal(1, r->num_entries);
+
+	r = imd.lg.r;
+	assert_int_equal(0, imd_entry_remove(&imd, snd_lg_entry));
+	assert_int_equal(1, r->num_entries);
+
+	/* Fail trying to remove root entry */
+	assert_int_equal(-1, imd_entry_remove(&imd, fst_lg_entry));
+	assert_int_equal(1, r->num_entries);
+
+	free(base);
+}
+
+static void test_imd_cursor_init(void **state)
+{
+	struct imd imd = {0};
+	struct imd_cursor cursor;
+
+	assert_int_equal(-1, imd_cursor_init(NULL, NULL));
+	assert_int_equal(-1, imd_cursor_init(NULL, &cursor));
+	assert_int_equal(-1, imd_cursor_init(&imd, NULL));
+	assert_int_equal(0, imd_cursor_init(&imd, &cursor));
+
+	assert_ptr_equal(cursor.imdr[0], &imd.lg);
+	assert_ptr_equal(cursor.imdr[1], &imd.sm);
+}
+
+static void test_imd_cursor_next(void **state)
+{
+	void *base;
+	struct imd imd = {0};
+	struct imd_cursor cursor;
+	struct imd_root *r;
+	const struct imd_entry *entry;
+	struct imd_entry *fst_lg_entry, *snd_lg_entry, *fst_sm_entry;
+	assert_int_equal(0, imd_cursor_init(&imd, &cursor));
+
+	cursor.current_imdr = 3;
+	cursor.current_entry = 0;
+	assert_null(imd_cursor_next(&cursor));
+
+	cursor.current_imdr = 0;
+	assert_null(imd_cursor_next(&cursor));
+
+	base = malloc(LIMIT_ALIGN);
+	if (base == NULL)
+		fail_msg("Cannot allocate enough memory - fail test");
+	imd_handle_init(&imd, (void *)(LIMIT_ALIGN + (uintptr_t)base));
+
+	assert_int_equal(0, imd_create_tiered_empty(&imd, LG_ROOT_SIZE, LG_ENTRY_ALIGN,
+						    SM_ROOT_SIZE, SM_ENTRY_ALIGN));
+
+	r = imd.lg.r;
+	entry = imd_cursor_next(&cursor);
+	assert_non_null(entry);
+
+	fst_lg_entry = &r->entries[0];
+	assert_int_equal(fst_lg_entry->id, entry->id);
+	assert_ptr_equal(fst_lg_entry, entry);
+
+	entry = imd_cursor_next(&cursor);
+	assert_non_null(entry);
+
+	snd_lg_entry = &r->entries[1];
+	assert_int_equal(snd_lg_entry->id, entry->id);
+	assert_ptr_equal(snd_lg_entry, entry);
+
+	entry = imd_cursor_next(&cursor);
+	assert_non_null(entry);
+
+	r = imd.sm.r;
+	fst_sm_entry = &r->entries[0];
+	assert_int_equal(fst_sm_entry->id, entry->id);
+	assert_ptr_equal(fst_sm_entry, entry);
+
+	entry = imd_cursor_next(&cursor);
+	assert_null(entry);
+}
+
+int main(void)
+{
+	const struct CMUnitTest tests[] = {
+		cmocka_unit_test(test_imd_handle_init),
+		cmocka_unit_test(test_imd_handle_init_partial_recovery),
+		cmocka_unit_test(test_imd_create_empty),
+		cmocka_unit_test(test_imd_create_tiered_empty),
+		cmocka_unit_test(test_imd_recover),
+		cmocka_unit_test(test_imd_limit_size),
+		cmocka_unit_test(test_imd_lockdown),
+		cmocka_unit_test(test_imd_region_used),
+		cmocka_unit_test(test_imd_entry_add),
+		cmocka_unit_test(test_imd_entry_find),
+		cmocka_unit_test(test_imd_entry_find_or_add),
+		cmocka_unit_test(test_imd_entry_size),
+		cmocka_unit_test(test_imd_entry_at),
+		cmocka_unit_test(test_imd_entry_id),
+		cmocka_unit_test(test_imd_entry_remove),
+		cmocka_unit_test(test_imd_cursor_init),
+		cmocka_unit_test(test_imd_cursor_next),
+	};
+
+	return cb_run_group_tests(tests, NULL, NULL);
+}
+

tests/lib/Makefile.inc

@@ -38,6 +38,7 @@ tests-y += cbfs-no-verification-no-sha512-test
 tests-y += cbfs-no-verification-has-sha512-test
 tests-y += cbfs-lookup-no-mcache-test
 tests-y += cbfs-lookup-has-mcache-test
+tests-y += lzma-test
 
 string-test-srcs += tests/lib/string-test.c
 string-test-srcs += src/lib/string.c
@@ -231,3 +232,8 @@ cbfs-lookup-no-mcache-test-config += CONFIG_ARCH_X86=0 \
 
 $(call copy-test,cbfs-lookup-no-mcache-test,cbfs-lookup-has-mcache-test)
 cbfs-lookup-has-mcache-test-config += CONFIG_NO_CBFS_MCACHE=0
+
+lzma-test-srcs += tests/lib/lzma-test.c
+lzma-test-srcs += tests/stubs/console.c
+lzma-test-srcs += src/lib/lzma.c
+lzma-test-srcs += src/lib/lzmadecode.c

tests/lib/lzma-test.c

@@ -0,0 +1,190 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#include <fcntl.h>
+#include <lib.h>
+#include <lib/lzmadecode.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <tests/test.h>
+#include <unistd.h>
+
+
+struct lzma_test_state {
+	char *raw_filename;
+	size_t raw_file_sz;
+	char *comp_filename;
+	size_t comp_file_sz;
+};
+
+static int get_file_size(const char *fname)
+{
+	struct stat st;
+	if (stat(fname, &st) == -1)
+		return -1;
+	return st.st_size;
+}
+
+static int teardown_ulzman_file(void **state)
+{
+	struct lzma_test_state *s = *state;
+
+	test_free(s->raw_filename);
+	test_free(s->comp_filename);
+	test_free(s);
+
+	return 0;
+}
+
+/* Set data file with prestate */
+static int setup_ulzman_file(void **state)
+{
+	int ret = 0;
+	const char *fname_base = *state;
+	const char path_prefix[] = __TEST_DATA_DIR__ "/lib/lzma-test/%s%s";
+	const char raw_file_suffix[] = ".bin";
+	const char comp_file_suffix[] = ".lzma.bin";
+	struct lzma_test_state *s = test_malloc(sizeof(*s));
+	memset(s, 0, sizeof(*s));
+
+	if (!s)
+		return 1;
+
+	const size_t raw_filename_size = strlen(path_prefix) + strlen(fname_base)
+		+ ARRAY_SIZE(raw_file_suffix);
+	s->raw_filename = test_malloc(raw_filename_size);
+
+	const size_t comp_filename_size = strlen(path_prefix) + strlen(fname_base)
+		+ ARRAY_SIZE(comp_file_suffix);
+	s->comp_filename = test_malloc(comp_filename_size);
+
+	if (!s->raw_filename || !s->comp_filename) {
+		print_error("File path allocation error\n");
+		ret = 2;
+		goto error;
+	}
+
+	snprintf(s->raw_filename, raw_filename_size, path_prefix, fname_base, raw_file_suffix);
+	snprintf(s->comp_filename, comp_filename_size, path_prefix, fname_base,
+			comp_file_suffix);
+
+	s->raw_file_sz = get_file_size(s->raw_filename);
+	s->comp_file_sz = get_file_size(s->comp_filename);
+
+	if (s->raw_file_sz == -1) {
+		print_error("Unable to open file: %s\n", s->raw_filename);
+		ret = 3;
+		goto error;
+	}
+
+	if (s->comp_file_sz == -1) {
+		print_error("Unable to open file: %s\n", s->comp_filename);
+		ret = 3;
+		goto error;
+	}
+
+	*state = s;
+	return 0;
+error:
+	teardown_ulzman_file((void **)&s);
+	return ret;
+}
+
+static int read_file(const char *fname, uint8_t *buf, size_t sz)
+{
+	int f = open(fname, O_RDONLY);
+	int read_sz = 0;
+
+	if (f == -1)
+		return -1;
+
+	read_sz = read(f, buf, sz);
+
+	close(f);
+	return read_sz;
+}
+
+static void test_ulzman_correct_file(void **state)
+{
+	struct lzma_test_state *s = *state;
+	uint8_t *raw_buf = test_malloc(s->raw_file_sz);
+	uint8_t *decomp_buf = test_malloc(s->raw_file_sz);
+	uint8_t *comp_buf = test_malloc(s->comp_file_sz);
+
+	assert_non_null(raw_buf);
+	assert_non_null(decomp_buf);
+	assert_non_null(comp_buf);
+	assert_int_equal(s->raw_file_sz, read_file(s->raw_filename, raw_buf, s->raw_file_sz));
+	assert_int_equal(s->comp_file_sz,
+			read_file(s->comp_filename, comp_buf, s->comp_file_sz));
+
+	assert_int_equal(s->raw_file_sz,
+			ulzman(comp_buf, s->comp_file_sz, decomp_buf, s->raw_file_sz));
+	assert_memory_equal(raw_buf, decomp_buf, s->raw_file_sz);
+
+	test_free(raw_buf);
+	test_free(decomp_buf);
+	test_free(comp_buf);
+}
+
+static void test_ulzman_input_too_small(void **state)
+{
+	uint8_t in_buf[32];
+	uint8_t out_buf[32];
+
+	assert_int_equal(0, ulzman(in_buf, LZMA_PROPERTIES_SIZE, out_buf, sizeof(out_buf)));
+}
+
+static void test_ulzman_zero_buffer(void **state)
+{
+	uint8_t in_buf[LZMA_PROPERTIES_SIZE + 1 * KiB];
+	uint8_t out_buf[2 * KiB];
+
+	memset(in_buf, 0, sizeof(in_buf));
+	memset(out_buf, 0, sizeof(out_buf));
+
+	assert_int_equal(0, ulzman(in_buf, sizeof(in_buf), out_buf, sizeof(out_buf)));
+}
+
+#define ULZMAN_CORRECT_FILE_TEST(_file_prefix)                                                 \
+{                                                                                      \
+	.name = "test_ulzman_correct_file(" _file_prefix ")",                          \
+	.test_func = test_ulzman_correct_file,                                         \
+	.setup_func = setup_ulzman_file,                                               \
+	.teardown_func = teardown_ulzman_file,                                         \
+	.initial_state = (_file_prefix)                                                \
+}
+
+int main(void)
+{
+	const struct CMUnitTest tests[] = {
+		/* "data.N" in macros below refers to files:
+		   - __TEST_DATA_DIR__ /lib/lzma-test/data.N.bin
+		   - __TEST_DATA_DIR__ /lib/lzma-test/data.N.bin.lzma
+		   Files data.N.bin suffix are raw data, and data.N.lzma.bin are its
+		   LZMA-compressed form. Both are required to exist.
+		 */
+
+		/* util/cbfs-compression-tool compressed by itself.
+		   To test compression of executable files like payloads. */
+		ULZMAN_CORRECT_FILE_TEST("data.1"),
+
+		/* README.md compressed by util/cbfs-compression-tool. */
+		ULZMAN_CORRECT_FILE_TEST("data.2"),
+
+		/* tests/lib/imd-test.c compressed by util/cbfs-compression-tool
+		   Structured text file. */
+		ULZMAN_CORRECT_FILE_TEST("data.3"),
+
+		/* libcmocka.so.0.7.0 compressed by util/cbfs-compression-tool
+		   Another binary file, shared object. */
+		ULZMAN_CORRECT_FILE_TEST("data.4"),
+
+		cmocka_unit_test(test_ulzman_input_too_small),
+
+		cmocka_unit_test(test_ulzman_zero_buffer),
+	};
+
+	return cb_run_group_tests(tests, NULL, NULL);
+}