Skip to contentSkip to conversation

Various bugfixes related to verified boot

PATCH In review Simon Glass 3 participants 23 messages Feb 28, 2022 – Apr 7, 2022

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

cover Various bugfixes related to verified boot

This series collects together various misc patches that were needed when building mainline U-Boot against Chromium OS verified boot. Most of them fix minor bugs.

Simon Glass (18):
buildman: Update default config to build for sandbox
buildman: Fix up cfgutil
binman: Correct Chromium OS entry types
errno: Avoid including strings in SPL
abuf: Correct a corner case with abuf_realloc()
fdt: Correct condition for SEPARATE_BSS
fdt: sandbox: Avoid looking for an appended device tree
lzma: Tidy up the function prototype
cbfs: Add some more definititions
cros_ec: Complete the comment for cros_ec_read_batt_charge()
spi: Avoid checking console in SPL
disk: Correct the conditions for SPL
Add a default for TPL_TEXT_BASE
Make ASYMMETRIC_KEY_TYPE depend on FIT_SIGNATURE
stdint: Add a definition of UINT8_MAX
dm: core: Add a required struct declaration
dm: core: Tidy up comments in uclass headers
dm: blk: Expand iteration and add tests

 common/spl/Kconfig             |   1 +
 disk/Makefile                  |  10 +--
 drivers/block/blk-uclass.c     |  24 +++++++
 drivers/mtd/spi/spi-nor-core.c |   2 +-
 include/blk.h                  |  45 +++++++++++++
 include/cbfs.h                 |  43 +++++++++++++
 include/cros_ec.h              |   1 +
 include/dm/device-internal.h   |   1 +
 include/dm/uclass-internal.h   |  18 +++---
 include/dm/uclass.h            |  13 ++--
 include/errno.h                |   2 +-
 include/stdint.h               |   7 +++
 lib/abuf.c                     |   4 +-
 lib/crypto/Kconfig             |   1 +
 lib/fdtdec.c                   |   5 +-
 lib/lzma/LzmaTools.c           |   4 +-
 lib/lzma/LzmaTools.h           |  17 ++++-
 test/dm/blk.c                  | 111 +++++++++++++++++++++++++++++++++
 test/lib/abuf.c                |  29 +++++++++
 tools/binman/btool/futility.py |   4 +-
 tools/binman/etype/vblock.py   |   2 +-
 tools/buildman/bsettings.py    |   1 +
 tools/buildman/cfgutil.py      |   4 +-
 23 files changed, 315 insertions(+), 34 deletions(-)
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#1 lore raw

PATCH 1/18 buildman: Update default config to build for sandbox

At present the default .buildman file written by buildman does not specify a default toolchain. Add an 'other' line so this works correctly and sandbox builds run as expected.

---

 tools/buildman/bsettings.py | 1 +
 1 file changed, 1 insertion(+)
tools/buildman/bsettings.py +1 -0
index 0b7208da37..e634bbb279 100644
@@ -74,6 +74,7 @@ def CreateBuildmanConfigFile(config_fname):
     print('''[toolchain]
 # name = path
 # e.g. x86 = /opt/gcc-4.6.3-nolibc/x86_64-linux
+other = /
 
 [toolchain-prefix]
 # name = path to prefix
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#2 lore raw

Re: [PATCH 01/18] buildman: Update default config to build for sandbox

Quoted 5 lines · Simon Glass

At present the default .buildman file written by buildman does not specify a default toolchain. Add an 'other' line so this works correctly and sandbox builds run as expected.

For the series, applied to u-boot/master, thanks!

Signature · 1 line
Tom
#3 lore raw

PATCH 2/18 buildman: Fix up cfgutil

In reply to Simon Glass

The name of this function changed at the same time as this file was added. Fix the naming.

---

 tools/buildman/cfgutil.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
tools/buildman/cfgutil.py +2 -2
index 4eba50868f..166b7b2419 100644
@@ -123,7 +123,7 @@ def adjust_cfg_file(fname, adjust_cfg):
                  C=val to set the value of C (val must have quotes if C is
                      a string Kconfig)
     """
-    lines = tools.ReadFile(fname, binary=False).splitlines()
+    lines = tools.read_file(fname, binary=False).splitlines()
     out_lines = adjust_cfg_lines(lines, adjust_cfg)
     out = '\n'.join(out_lines) + '\n'
     tools.WriteFile(fname, out, binary=False)
@@ -219,7 +219,7 @@ def check_cfg_file(fname, adjust_cfg):
     Returns:
         str: None if OK, else an error string listing the problems
     """
-    lines = tools.ReadFile(fname, binary=False).splitlines()
+    lines = tools.read_file(fname, binary=False).splitlines()
     bad_cfgs = check_cfg_lines(lines, adjust_cfg)
     if bad_cfgs:
         out = [f'{cfg:20}  {line}' for cfg, line in bad_cfgs]
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#4 lore raw

PATCH 3/18 binman: Correct Chromium OS entry types

In reply to Simon Glass

The conversion to bintools broke the invocation of the utility, since the arguments are not correct. Fix it.

---

 tools/binman/btool/futility.py | 4 ++--
 tools/binman/etype/vblock.py   | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)
tools/binman/btool/futility.py +2 -2
index 614daaade4..8d00966a9d 100644
@@ -106,7 +106,7 @@ class Bintoolfutility(bintool.Bintool):
         Returns:
             str: Tool output
         """
-        args = ['gbb_utility'
+        args = ['gbb_utility',
             '-s',
             f'--hwid={hwid}',
             f'--rootkey={rootkey}',
@@ -139,7 +139,7 @@ class Bintoolfutility(bintool.Bintool):
             '--keyblock', keyblock,
             '--signprivate', signprivate,
             '--version', version,
-            '--fw', firmware,
+            '--fv', firmware,
             '--kernelkey', kernelkey,
             '--flags', flags
             ]
tools/binman/etype/vblock.py +1 -1
index a1de98290b..0074c577a2 100644
@@ -73,7 +73,7 @@ class Entry_vblock(Entry_collection):
             vblock=output_fname,
             keyblock=prefix + self.keyblock,
             signprivate=prefix + self.signprivate,
-            version=f'{self.version,}',
+            version=f'{self.version:d}',
             firmware=input_fname,
             kernelkey=prefix + self.kernelkey,
             flags=f'{self.preamble_flags}')
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#5 lore raw

PATCH 4/18 errno: Avoid including strings in SPL

In reply to Simon Glass

At present the header file defines this function in SPL but the file may not actually be built. This causes a build error if the option is enabled.

Fix the condition in the header file.

---

 include/errno.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/errno.h +1 -1
index 652ad67306..5a8816d0a1 100644
@@ -25,7 +25,7 @@ extern int errno __errno_asm_label;
  * Return:	string describing the error. If CONFIG_ERRNO_STR is not
  *		defined an empty string is returned.
  */
-#ifdef CONFIG_ERRNO_STR
+#if CONFIG_IS_ENABLED(ERRNO_STR)
 const char *errno_str(int errno);
 #else
 static const char error_message[] = "";
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#6 lore raw

PATCH 5/18 abuf: Correct a corner case with abuf_realloc()

In reply to Simon Glass

If the buffer is empty and not allocated, then abuf_realloc() tries to copy invalid data. This happens because an incorrect change to use memdup() was added after the original code was written.

---

 lib/abuf.c      |  4 +++-
 test/lib/abuf.c | 29 +++++++++++++++++++++++++++++
 2 files changed, 32 insertions(+), 1 deletion(-)
lib/abuf.c +3 -1
index 4b17e0b8c0..1635d58682 100644
@@ -51,9 +51,11 @@ bool abuf_realloc(struct abuf *abuf, size_t new_size)
 		/* not currently allocated and new size is larger. Alloc and
 		 * copy in data. The new space is not inited.
 		 */
-		ptr = memdup(abuf->data, new_size);
+		ptr = malloc(new_size);
 		if (!ptr)
 			return false;
+		if (abuf->size)
+			memcpy(ptr, abuf->data, abuf->size);
 		abuf->data = ptr;
 		abuf->size = new_size;
 		abuf->alloced = true;
test/lib/abuf.c +29 -0
index 086c9b2282..42ee4c1755 100644
@@ -126,6 +126,35 @@ static int lib_test_abuf_realloc(struct unit_test_state *uts)
 }
 LIB_TEST(lib_test_abuf_realloc, 0);
 
+/* Test abuf_realloc() on an non-allocated buffer of zero size */
+static int lib_test_abuf_realloc_size(struct unit_test_state *uts)
+{
+	struct abuf buf;
+	ulong start;
+
+	start = ut_check_free();
+
+	abuf_init(&buf);
+
+	/* Allocate some space */
+	ut_asserteq(true, abuf_realloc(&buf, TEST_DATA_LEN));
+	ut_assertnonnull(buf.data);
+	ut_asserteq(TEST_DATA_LEN, buf.size);
+	ut_asserteq(true, buf.alloced);
+
+	/* Free it */
+	ut_asserteq(true, abuf_realloc(&buf, 0));
+	ut_assertnull(buf.data);
+	ut_asserteq(0, buf.size);
+	ut_asserteq(false, buf.alloced);
+
+	/* Check for memory leaks */
+	ut_assertok(ut_check_delta(start));
+
+	return 0;
+}
+LIB_TEST(lib_test_abuf_realloc_size, 0);
+
 /* Test handling of buffers that are too large */
 static int lib_test_abuf_large(struct unit_test_state *uts)
 {
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#7 lore raw

PATCH 6/18 fdt: Correct condition for SEPARATE_BSS

In reply to Simon Glass

This may have different settings for SPL and TPL. Correct the condition.

---

 lib/fdtdec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
lib/fdtdec.c +1 -1
index 87aa677a4a..086f0c732a 100644
@@ -1227,7 +1227,7 @@ static void *fdt_find_separate(void)
 
 #ifdef CONFIG_SPL_BUILD
 	/* FDT is at end of BSS unless it is in a different memory region */
-	if (IS_ENABLED(CONFIG_SPL_SEPARATE_BSS))
+	if (CONFIG_IS_ENABLED(SEPARATE_BSS))
 		fdt_blob = (ulong *)&_image_binary_end;
 	else
 		fdt_blob = (ulong *)&__bss_end;
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#8 lore raw

PATCH 7/18 fdt: sandbox: Avoid looking for an appended device tree

In reply to Simon Glass

We don't use an appended tree for sandbox and the required symbols are not present. Add a condition to avoid a build error.

---

 lib/fdtdec.c | 3 +++
 1 file changed, 3 insertions(+)
lib/fdtdec.c +3 -0
index 086f0c732a..0c0ec034ec 100644
@@ -1225,6 +1225,9 @@ static void *fdt_find_separate(void)
 {
 	void *fdt_blob = NULL;
 
+	if (IS_ENABLED(CONFIG_SANDBOX))
+		return NULL;
+
 #ifdef CONFIG_SPL_BUILD
 	/* FDT is at end of BSS unless it is in a different memory region */
 	if (CONFIG_IS_ENABLED(SEPARATE_BSS))
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#9 lore raw

PATCH 8/18 lzma: Tidy up the function prototype

In reply to Simon Glass

This should use a const pointer for the input stream. Fix this and also add a proper comment.

---

 lib/lzma/LzmaTools.c |  4 ++--
 lib/lzma/LzmaTools.h | 17 +++++++++++++++--
 2 files changed, 17 insertions(+), 4 deletions(-)
lib/lzma/LzmaTools.c +2 -2
index 521258e623..af88900d31 100644
@@ -37,8 +37,8 @@
 static void *SzAlloc(void *p, size_t size) { return malloc(size); }
 static void SzFree(void *p, void *address) { free(address); }
 
-int lzmaBuffToBuffDecompress (unsigned char *outStream, SizeT *uncompressedSize,
-                  unsigned char *inStream,  SizeT  length)
+int lzmaBuffToBuffDecompress(unsigned char *outStream, SizeT *uncompressedSize,
+			     const unsigned char *inStream, SizeT length)
 {
     int res = SZ_ERROR_DATA;
     int i;
lib/lzma/LzmaTools.h +15 -2
index e52dfb8fac..2c46859a62 100644
@@ -13,6 +13,19 @@
 
 #include <lzma/LzmaTypes.h>
 
-extern int lzmaBuffToBuffDecompress (unsigned char *outStream, SizeT *uncompressedSize,
-			      unsigned char *inStream,  SizeT  length);
+/**
+ * lzmaBuffToBuffDecompress() - Decompress LZMA data
+ *
+ * @outStream: output buffer
+ * @uncompressedSize: On entry, the mnaximum uncompressed size of the data;
+ *	on exit, the actual uncompressed size after processing
+ * @inStream: Compressed bytes to decompress
+ * @length: Sizeof @inStream
+ * @return 0 if OK, SZ_ERROR_DATA if the data is in a format that cannot be
+ *	decompressed; SZ_ERROR_OUTPUT_EOF if *uncompressedSize is too small;
+ *	see also other SZ_ERROR... values
+ */
+int lzmaBuffToBuffDecompress(unsigned char *outStream, SizeT *uncompressedSize,
+			     const unsigned char *inStream, SizeT length);
+
 #endif
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#10 lore raw

PATCH 9/18 cbfs: Add some more definititions

In reply to Simon Glass

Add definitions for a payload or 'self', which is like an unpacked ELF file.

---

 include/cbfs.h | 43 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)
include/cbfs.h +43 -0
index 7449873b49..38efb1d2b0 100644
@@ -24,6 +24,8 @@ enum cbfs_filetype {
 	CBFS_TYPE_CBFSHEADER = 0x02,
 	CBFS_TYPE_STAGE = 0x10,
 	CBFS_TYPE_PAYLOAD = 0x20,
+	CBFS_TYPE_SELF = CBFS_TYPE_PAYLOAD,
+
 	CBFS_TYPE_FIT = 0x21,
 	CBFS_TYPE_OPTIONROM = 0x30,
 	CBFS_TYPE_BOOTSPLASH = 0x40,
@@ -120,6 +122,47 @@ struct cbfs_file_attr_hash {
 	u8  hash_data[];
 } __packed;
 
+/*** Component sub-headers ***/
+
+/* Following are component sub-headers for the "standard" component types */
+
+/**
+ * struct cbfs_stage - sub-header for stage components
+ *
+ * Stages are loaded by coreboot during the normal boot process
+ */
+struct cbfs_stage {
+	u32 compression;  /** Compression type */
+	u64 entry;  /** entry point */
+	u64 load;   /** Where to load in memory */
+	u32 len;          /** length of data to load */
+	u32 memlen;	   /** total length of object in memory */
+} __packed;
+
+/**
+ * struct cbfs_payload_segment - sub-header for payload components
+ *
+ * Payloads are loaded by coreboot at the end of the boot process
+ */
+struct cbfs_payload_segment {
+	u32 type;
+	u32 compression;
+	u32 offset;
+	u64 load_addr;
+	u32 len;
+	u32 mem_len;
+} __packed;
+
+struct cbfs_payload {
+	struct cbfs_payload_segment segments;
+};
+
+#define PAYLOAD_SEGMENT_CODE   0x45444F43
+#define PAYLOAD_SEGMENT_DATA   0x41544144
+#define PAYLOAD_SEGMENT_BSS    0x20535342
+#define PAYLOAD_SEGMENT_PARAMS 0x41524150
+#define PAYLOAD_SEGMENT_ENTRY  0x52544E45
+
 struct cbfs_cachenode {
 	struct cbfs_cachenode *next;
 	void *data;
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#11 lore raw

PATCH 10/18 cros_ec: Complete the comment for cros_ec_read_batt_charge()

In reply to Simon Glass

Add the missing 'Return' information.

---

 include/cros_ec.h | 1 +
 1 file changed, 1 insertion(+)
include/cros_ec.h +1 -0
index 2a7728622c..94c988a7d6 100644
@@ -648,6 +648,7 @@ int cros_ec_vstore_write(struct udevice *dev, int slot, const uint8_t *data,
  *
  * @dev: CROS-EC device
  * @chargep: Return battery-charge state as a percentage
+ * Return: 0 if OK, -ve on error
  */
 int cros_ec_read_batt_charge(struct udevice *dev, uint *chargep);
 
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#12 lore raw

PATCH 11/18 spi: Avoid checking console in SPL

In reply to Simon Glass

When SPI flash is used in SPL there is no console, so ctrlc() cannot be called. Add a condition to fix this.

---

 drivers/mtd/spi/spi-nor-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mtd/spi/spi-nor-core.c +1 -1
index a70fbda4bb..3b7c817c02 100644
@@ -929,7 +929,7 @@ static int spi_nor_erase(struct mtd_info *mtd, struct erase_info *instr)
 
 	while (len) {
 		WATCHDOG_RESET();
-		if (ctrlc()) {
+		if (!IS_ENABLED(CONFIG_SPL_BUILD) && ctrlc()) {
 			addr_known = false;
 			ret = -EINTR;
 			goto erase_err;
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#13 lore raw

PATCH 12/18 disk: Correct the conditions for SPL

In reply to Simon Glass

These filesystems may have different settings for SPL and TPL. Use the correct Makefile variable to handle this.

---

 disk/Makefile | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
disk/Makefile +5 -5
index 6ce5a687b3..83f5af6069 100644
@@ -6,8 +6,8 @@
 #ccflags-y += -DET_DEBUG -DDEBUG
 
 obj-$(CONFIG_PARTITIONS)	+= part.o
-obj-$(CONFIG_$(SPL_)MAC_PARTITION)   += part_mac.o
-obj-$(CONFIG_$(SPL_)DOS_PARTITION)   += part_dos.o
-obj-$(CONFIG_$(SPL_)ISO_PARTITION)   += part_iso.o
-obj-$(CONFIG_$(SPL_)AMIGA_PARTITION) += part_amiga.o
-obj-$(CONFIG_$(SPL_)EFI_PARTITION)   += part_efi.o
+obj-$(CONFIG_$(SPL_TPL_)MAC_PARTITION)   += part_mac.o
+obj-$(CONFIG_$(SPL_TPL_)DOS_PARTITION)   += part_dos.o
+obj-$(CONFIG_$(SPL_TPL_)ISO_PARTITION)   += part_iso.o
+obj-$(CONFIG_$(SPL_TPL_)AMIGA_PARTITION) += part_amiga.o
+obj-$(CONFIG_$(SPL_TPL_)EFI_PARTITION)   += part_efi.o
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#14 lore raw

PATCH 13/18 Add a default for TPL_TEXT_BASE

In reply to Simon Glass

If this value is not provided it causes a hang in the build. Add a default value to avoid this.

---

 common/spl/Kconfig | 1 +
 1 file changed, 1 insertion(+)
common/spl/Kconfig +1 -0
index 9418d37b2e..b59215fe4f 100644
@@ -1412,6 +1412,7 @@ config TPL_POWER
 
 config TPL_TEXT_BASE
 	hex "Base address for the .text section of the TPL stage"
+	default 0
 	help
 	  The base address for the .text section of the TPL stage.
 
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#15 lore raw

PATCH 14/18 Make ASYMMETRIC_KEY_TYPE depend on FIT_SIGNATURE

In reply to Simon Glass

Add this dependency to avoid a build error if FIT_SIGNATURE is not enabled.

---

 lib/crypto/Kconfig | 1 +
 1 file changed, 1 insertion(+)
lib/crypto/Kconfig +1 -0
index 6369bafac0..5c6e964862 100644
@@ -1,5 +1,6 @@
 menuconfig ASYMMETRIC_KEY_TYPE
 	bool "Asymmetric (public-key cryptographic) key Support"
+	depends on FIT_SIGNATURE
 	help
 	  This option provides support for a key type that holds the data for
 	  the asymmetric keys used for public key cryptographic operations such
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#16 lore raw

Re: [PATCH 14/18] Make ASYMMETRIC_KEY_TYPE depend on FIT_SIGNATURE

Simon,

Quoted 2 lines · Simon Glass

Add this dependency to avoid a build error if FIT_SIGNATURE is not enabled.

I doubt it.
With qemu_arm64_defconfig & ASYMMETRIC_KEY_TYPE & !FIT_SIGNATURE, I don't see any build error on 2022.04-rc2.

-Takahiro Akashi

Quoted 20 lines

---

 lib/crypto/Kconfig | 1 +
 1 file changed, 1 insertion(+)
lib/crypto/Kconfig +1 -0
index 6369bafac0..5c6e964862 100644
@@ -1,5 +1,6 @@
 menuconfig ASYMMETRIC_KEY_TYPE
 	bool "Asymmetric (public-key cryptographic) key Support"
+	depends on FIT_SIGNATURE
 	help
 	  This option provides support for a key type that holds the data for
 	  the asymmetric keys used for public key cryptographic operations such
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#17 lore raw

Re: [PATCH 14/18] Make ASYMMETRIC_KEY_TYPE depend on FIT_SIGNATURE

Hi Takahiro,

Quoted 10 lines · AKASHI Takahiro

Simon,

Quoted 2 lines · Simon Glass

Add this dependency to avoid a build error if FIT_SIGNATURE is not enabled.

I doubt it.
With qemu_arm64_defconfig & ASYMMETRIC_KEY_TYPE & !FIT_SIGNATURE, I don't see any build error on 2022.04-rc2.

This is what I see when building sandbox with Chromium OS verified boot (where FIT_SIGNATURE is not enabled).

/usr/bin/ld: lib/crypto/public_key.o: in function `public_key_verify_signature':
lib/crypto/public_key.c:113: undefined reference to image_get_padding_algo' /usr/bin/ld: lib/crypto/public_key.c:124: undefined reference to image_get_checksum_algo'
/usr/bin/ld: lib/crypto/public_key.c:126: undefined reference to `image_get_crypto_algo'
/usr/bin/ld: lib/crypto/public_key.c:136: undefined reference to `rsa_verify_with_pkey'
collect2: error: ld returned 1 exit status
make[1]: *** [Makefile:1901: u-boot] Error 1

Regards,
Simon
[..]

#18 lore raw

Re: [PATCH 14/18] Make ASYMMETRIC_KEY_TYPE depend on FIT_SIGNATURE

Hi Simon,

Quoted 18 lines · Simon Glass

Hi Takahiro,

Quoted 10 lines · AKASHI Takahiro

Simon,

Quoted 2 lines · Simon Glass

Add this dependency to avoid a build error if FIT_SIGNATURE is not enabled.

I doubt it.
With qemu_arm64_defconfig & ASYMMETRIC_KEY_TYPE & !FIT_SIGNATURE, I don't see any build error on 2022.04-rc2.

This is what I see when building sandbox with Chromium OS verified boot (where FIT_SIGNATURE is not enabled).

Quoted 6 lines

/usr/bin/ld: lib/crypto/public_key.o: in function `public_key_verify_signature':
lib/crypto/public_key.c:113: undefined reference to image_get_padding_algo' /usr/bin/ld: lib/crypto/public_key.c:124: undefined reference to image_get_checksum_algo'
/usr/bin/ld: lib/crypto/public_key.c:126: undefined reference to `image_get_crypto_algo'

After my commit b983cc2da0ba ("lib: rsa: decouple rsa from FIT image verification"), those symbols are independently compiled in with CONFIG_IMAGE_SIGN_INFO.

Quoted 2 lines

/usr/bin/ld: lib/crypto/public_key.c:136: undefined reference to `rsa_verify_with_pkey'

Likewise, please select CONFIG_RSA_VERIFY_WITH_PKEY specifically instead of CONFIG_FIT_SIGNATURE.

-Takahiro Akashi

Quoted 7 lines

collect2: error: ld returned 1 exit status
make[1]: *** [Makefile:1901: u-boot] Error 1

Regards,
Simon
[..]

#19 lore raw

PATCH 15/18 stdint: Add a definition of UINT8_MAX

In reply to Simon Glass

This is normally defined in stdint.h but is not used in U-Boot. When libraries (such as Chromium OS vboot) are built against U-Boot they may expect this value to be available. Add it to avoid build errors in this case.

---

 include/stdint.h | 7 +++++++
 1 file changed, 7 insertions(+)
include/stdint.h +7 -0
index 2e126d14bd..dea83c8226 100644
@@ -5,3 +5,10 @@
  *
  * U-Boot uses linux types (linux/types.h) so does not make use of stdint.h
  */
+
+#ifndef __UB_STDINT_H
+#define __UB_STDINT_H
+
+#define UINT8_MAX	0xff
+
+#endif
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#20 lore raw

PATCH 16/18 dm: core: Add a required struct declaration

In reply to Simon Glass

This file uses struct driver, so declare it at the top in case the header-inclusion order is not as expected.

---

 include/dm/device-internal.h | 1 +
 1 file changed, 1 insertion(+)
include/dm/device-internal.h +1 -0
index 02002acb78..46ae0e1f1f 100644
@@ -14,6 +14,7 @@
 #include <dm/ofnode.h>
 
 struct device_node;
+struct driver_info;
 struct udevice;
 
 /*
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#21 lore raw

PATCH 17/18 dm: core: Tidy up comments in uclass headers

In reply to Simon Glass

Improve some of the comments in these files, which don't follow the correct style.

---

 include/dm/uclass-internal.h | 18 +++++++++---------
 include/dm/uclass.h          | 13 ++++++-------
 2 files changed, 15 insertions(+), 16 deletions(-)
include/dm/uclass-internal.h +9 -9
index e71b86a973..daf856c03c 100644
@@ -55,7 +55,7 @@
  *
  * @_name: Name of the uclass. This must be a valid C identifier, used by the
  *	linker_list
- * @returns struct uclass * for the device
+ * Return: struct uclass * for the device
  */
 #define DM_UCLASS_REF(_name)						\
 	ll_entry_ref(struct uclass, _name, uclass)
@@ -120,7 +120,7 @@ int dev_get_uclass_index(struct udevice *dev, struct uclass **ucp);
  * uclass_find_device() - Return n-th child of uclass
  * @id:		Id number of the uclass
  * @index:	Position of the child in uclass's list
- * #devp:	Returns pointer to device, or NULL on error
+ * @devp:	Returns pointer to device, or NULL on error
  *
  * The device is not prepared for use - this is an internal function.
  * The function uclass_get_device_tail() can be used to probe the device.
@@ -133,7 +133,7 @@ int uclass_find_device(enum uclass_id id, int index, struct udevice **devp);
 /**
  * uclass_find_first_device() - Return the first device in a uclass
  * @id:		Id number of the uclass
- * #devp:	Returns pointer to device, or NULL on error
+ * @devp:	Returns pointer to device, or NULL on error
  *
  * The device is not prepared for use - this is an internal function.
  * The function uclass_get_device_tail() can be used to probe the device.
@@ -239,7 +239,7 @@ int uclass_find_device_by_phandle(enum uclass_id id, struct udevice *parent,
  * Connect the device into uclass's list of devices.
  *
  * @dev:	Pointer to the device
- * #return 0 on success, -ve on error
+ * Return: 0 on success, -ve on error
  */
 int uclass_bind_device(struct udevice *dev);
 
@@ -250,7 +250,7 @@ int uclass_bind_device(struct udevice *dev);
  * Call any handled needed before uclass_unbind_device() is called
  *
  * @dev:	Pointer to the device
- * #return 0 on success, -ve on error
+ * Return: 0 on success, -ve on error
  */
 int uclass_pre_unbind_device(struct udevice *dev);
 
@@ -260,7 +260,7 @@ int uclass_pre_unbind_device(struct udevice *dev);
  * Disconnect the device from uclass's list of devices.
  *
  * @dev:	Pointer to the device
- * #return 0 on success, -ve on error
+ * Return: 0 on success, -ve on error
  */
 int uclass_unbind_device(struct udevice *dev);
 
@@ -277,7 +277,7 @@ static inline int uclass_unbind_device(struct udevice *dev) { return 0; }
  * uclass' child_pre_probe() method.
  *
  * @dev:	Pointer to the device
- * #return 0 on success, -ve on error
+ * Return: 0 on success, -ve on error
  */
 int uclass_pre_probe_device(struct udevice *dev);
 
@@ -288,7 +288,7 @@ int uclass_pre_probe_device(struct udevice *dev);
  * uclass.
  *
  * @dev:	Pointer to the device
- * #return 0 on success, -ve on error
+ * Return: 0 on success, -ve on error
  */
 int uclass_post_probe_device(struct udevice *dev);
 
@@ -298,7 +298,7 @@ int uclass_post_probe_device(struct udevice *dev);
  * Perform any pre-processing of a device that is about to be removed.
  *
  * @dev:	Pointer to the device
- * #return 0 on success, -ve on error
+ * Return: 0 on success, -ve on error
  */
 #if CONFIG_IS_ENABLED(DM_DEVICE_REMOVE)
 int uclass_pre_remove_device(struct udevice *dev);
include/dm/uclass.h +6 -7
index 7f33c34214..aafe652288 100644
@@ -435,7 +435,7 @@ int uclass_probe_all(enum uclass_id id);
 int uclass_id_count(enum uclass_id id);
 
 /**
- * uclass_id_foreach_dev() - Helper function to iteration through devices
+ * uclass_id_foreach_dev() - iterate through devices of a given uclass ID
  *
  * This creates a for() loop which works through the available devices in
  * a uclass ID in order from start to end.
@@ -452,20 +452,20 @@ int uclass_id_count(enum uclass_id id);
 		list_for_each_entry(pos, &uc->dev_head, uclass_node)
 
 /**
- * uclass_foreach_dev() - Helper function to iteration through devices
+ * uclass_foreach_dev() - iterate through devices of a given uclass
  *
  * This creates a for() loop which works through the available devices in
  * a uclass in order from start to end.
  *
  * @pos: struct udevice * to hold the current device. Set to NULL when there
  * are no more devices.
- * @uc: uclass to scan
+ * @uc: uclass to scan (`struct uclass *`)
  */
 #define uclass_foreach_dev(pos, uc)	\
 	list_for_each_entry(pos, &uc->dev_head, uclass_node)
 
 /**
- * uclass_foreach_dev_safe() - Helper function to safely iteration through devs
+ * uclass_foreach_dev_safe() - safely iterate through devices of a given uclass
  *
  * This creates a for() loop which works through the available devices in
  * a uclass in order from start to end. Inside the loop, it is safe to remove
@@ -474,14 +474,13 @@ int uclass_id_count(enum uclass_id id);
  * @pos: struct udevice * to hold the current device. Set to NULL when there
  * are no more devices.
  * @next: struct udevice * to hold the next next
- * @uc: uclass to scan
+ * @uc: uclass to scan (`struct uclass *`)
  */
 #define uclass_foreach_dev_safe(pos, next, uc)	\
 	list_for_each_entry_safe(pos, next, &uc->dev_head, uclass_node)
 
 /**
- * uclass_foreach_dev_probe() - Helper function to iteration through devices
- * of given uclass
+ * uclass_foreach_dev_probe() - iterate through devices of a given uclass ID
  *
  * This creates a for() loop which works through the available devices in
  * a uclass in order from start to end. Devices are probed if necessary,
Signature · 1 line
2.35.1.574.g5d30c73bfb-goog
#22 lore raw

PATCH 18/18 dm: blk: Expand iteration and add tests

In reply to Simon Glass

Add some functions which support iteration before probing. Also add tests for the functions.

---

 drivers/block/blk-uclass.c |  24 ++++++++
 include/blk.h              |  45 +++++++++++++++
 test/dm/blk.c              | 111 +++++++++++++++++++++++++++++++++++++
 3 files changed, 180 insertions(+)
drivers/block/blk-uclass.c +24 -0
index bee1cd6f0d..84502bb47c 100644
@@ -552,6 +552,30 @@ static int blk_flags_check(struct udevice *dev, enum blk_flag_t req_flags)
 	return flags & req_flags ? 0 : 1;
 }
 
+int blk_find_first(enum blk_flag_t flags, struct udevice **devp)
+{
+	int ret;
+
+	for (ret = uclass_find_first_device(UCLASS_BLK, devp);
+	     *devp && !blk_flags_check(*devp, flags);
+	     ret = uclass_find_next_device(devp))
+		return 0;
+
+	return -ENODEV;
+}
+
+int blk_find_next(enum blk_flag_t flags, struct udevice **devp)
+{
+	int ret;
+
+	for (ret = uclass_find_next_device(devp);
+	     *devp && !blk_flags_check(*devp, flags);
+	     ret = uclass_find_next_device(devp))
+		return 0;
+
+	return -ENODEV;
+}
+
 int blk_first_device_err(enum blk_flag_t flags, struct udevice **devp)
 {
 	int ret;
include/blk.h +45 -0

51 lines not on this page · raw message

test/dm/blk.c +111 -0

114 lines not on this page · raw message

Signature · 1 line
2.35.1.574.g5d30c73bfb-goog