Skip to contentSkip to conversation

import x509/pkcs7 parsers from linux

PATCH New AKASHI Takahiro 3 participants 39 messages Oct 11, 2019 – Oct 23, 2019

Subsystems (7)
lore
Share & tools
lore
#0 lore raw

cover import x509/pkcs7 parsers from linux

Asn1 parsers of x509 certificates and pkcs7 messages are required to implement image authentication and variable authentication as part of UEFI secure boot feature.

As we discussed before in the thread[1], most people insisted that we should re-use corresponding source code from Linux repository for this purpose.

Here is my attempt to import all the necessary files from Linux; Those will eventually be part of UEFI secure boot implementation, but I'd like to get early feedback from other peoples before submitting the whole patchset so that they will be better formatted for merging.

My approach here is
* files from the latest Linux
* modify files as little as possible
* mark/protect unavoidable changes with "#if(n)def __UBOOT__" so that future fixes/differences in Linux repository will easily be applied to U-Boot.

Please note that checkpatch.pl will complain with a bunch of warnings/errors but I intentionally left them unchanged for the sake of better maintainability I said above.

Any comments will be appreciated.
-Takahiro Akashi

[1] https://lists.denx.de/pipermail/u-boot/2019-April/366423.html

Changes in v1 (Oct 11, 2019) from RFC
* change the kernel code base from v5.0 to v5.3
* comment off x509_check_for_self_signed() which is not useful for UEFI secure boot (patch#9)
* improve usages of "#if(n)def __UBOOT__* to minimize differences between U-Boot and linux kernel

AKASHI Takahiro (11):
linux_compat: add kmemdup()
include: time.h: define time64_t
include: kernel.h: include printk.h
cmd: add asn1_compiler
Makefile: add build script for asn1 parsers
lib: add asn1 decoder
lib: add oid registry utility
lib: crypto: add public key utility
lib: crypto: add x509 parser
lib: crypto: add pkcs7 message parser
lib: crypto: add rsa public key parser

 cmd/Kconfig                       |    3 +
 include/crypto/internal/rsa.h     |   57 +
 include/crypto/pkcs7.h            |   47 +
 include/crypto/public_key.h       |   90 ++
 include/keys/asymmetric-type.h    |   88 ++
 include/linux/asn1.h              |   65 ++
 include/linux/asn1_ber_bytecode.h |   89 ++
 include/linux/asn1_decoder.h      |   20 +
 include/linux/compat.h            |    4 +-
 include/linux/kernel.h            |    2 +
 include/linux/oid_registry.h      |  117 +++
 include/linux/time.h              |   24 +
 lib/Kconfig                       |   12 +
 lib/Makefile                      |   18 +
 lib/asn1_decoder.c                |  527 ++++++++++
 lib/build_OID_registry            |  203 ++++
 lib/crypto/Kconfig                |   38 +
 lib/crypto/Makefile               |   46 +
 lib/crypto/asymmetric_type.c      |  668 ++++++++++++
 lib/crypto/pkcs7.asn1             |  135 +++
 lib/crypto/pkcs7_parser.c         |  693 +++++++++++++
 lib/crypto/pkcs7_parser.h         |   65 ++
 lib/crypto/public_key.c           |  376 +++++++
 lib/crypto/rsa_helper.c           |  198 ++++
 lib/crypto/rsapubkey.asn1         |    4 +
 lib/crypto/x509.asn1              |   60 ++
 lib/crypto/x509_akid.asn1         |   35 +
 lib/crypto/x509_cert_parser.c     |  697 +++++++++++++
 lib/crypto/x509_parser.h          |   57 +
 lib/crypto/x509_public_key.c      |  292 ++++++
 lib/linux_compat.c                |   11 +
 lib/oid_registry.c                |  179 ++++
 scripts/Makefile                  |    3 +
 scripts/Makefile.build            |    2 +-
 scripts/asn1_compiler.c           | 1611 +++++++++++++++++++++++++++++
 35 files changed, 6533 insertions(+), 3 deletions(-)
 create mode 100644 include/crypto/internal/rsa.h
 create mode 100644 include/crypto/pkcs7.h
 create mode 100644 include/crypto/public_key.h
 create mode 100644 include/keys/asymmetric-type.h
 create mode 100644 include/linux/asn1.h
 create mode 100644 include/linux/asn1_ber_bytecode.h
 create mode 100644 include/linux/asn1_decoder.h
 create mode 100644 include/linux/oid_registry.h
 create mode 100644 lib/asn1_decoder.c
 create mode 100755 lib/build_OID_registry
 create mode 100644 lib/crypto/Kconfig
 create mode 100644 lib/crypto/Makefile
 create mode 100644 lib/crypto/asymmetric_type.c
 create mode 100644 lib/crypto/pkcs7.asn1
 create mode 100644 lib/crypto/pkcs7_parser.c
 create mode 100644 lib/crypto/pkcs7_parser.h
 create mode 100644 lib/crypto/public_key.c
 create mode 100644 lib/crypto/rsa_helper.c
 create mode 100644 lib/crypto/rsapubkey.asn1
 create mode 100644 lib/crypto/x509.asn1
 create mode 100644 lib/crypto/x509_akid.asn1
 create mode 100644 lib/crypto/x509_cert_parser.c
 create mode 100644 lib/crypto/x509_parser.h
 create mode 100644 lib/crypto/x509_public_key.c
 create mode 100644 lib/oid_registry.c
 create mode 100644 scripts/asn1_compiler.c
Signature · 1 line
2.21.0
#1 lore raw

PATCH 1/11 linux_compat: add kmemdup()

Adding kmemdup() will help improve portability from linux kernel code (in my case, lib/crypto/x509_cert_parser.c and pkcs7_parser.c).

---
 include/linux/compat.h |  4 ++--
 lib/linux_compat.c     | 11 +++++++++++
 2 files changed, 13 insertions(+), 2 deletions(-)
include/linux/compat.h +2 -2
index d0f51baab407..d23ef50454ce 100644
@@ -117,6 +117,8 @@ static inline void kmem_cache_destroy(struct kmem_cache *cachep)
 	free(cachep);
 }
 
+void *kmemdup(const void *src, size_t size, int flags);
+
 #define DECLARE_WAITQUEUE(...)	do { } while (0)
 #define add_wait_queue(...)	do { } while (0)
 #define remove_wait_queue(...)	do { } while (0)
@@ -346,8 +348,6 @@ struct writeback_control {
 	unsigned for_sync:1;		/* sync(2) WB_SYNC_ALL writeback */
 };
 
-void *kmemdup(const void *src, size_t len, gfp_t gfp);
-
 typedef int irqreturn_t;
 
 struct timer_list {};
lib/linux_compat.c +11 -0
index 6373b4451eb3..dd1e5b3c2087 100644
@@ -40,3 +40,14 @@ void *kmem_cache_alloc(struct kmem_cache *obj, int flag)
 {
 	return malloc_cache_aligned(obj->sz);
 }
+
+void *kmemdup(const void *src, size_t size, int flags)
+{
+	void *p;
+
+	p = kmalloc(size, flags);
+	if (p)
+		memcpy(p, src, size);
+
+	return p;
+}
Signature · 1 line
2.21.0
#2 lore raw

PATCH 1/11 linux_compat: add kmemdup()

Quoted 14 lines · AKASHI Takahiro

Adding kmemdup() will help improve portability from linux kernel code (in my case, lib/crypto/x509_cert_parser.c and pkcs7_parser.c).

---
  include/linux/compat.h |  4 ++--
  lib/linux_compat.c     | 11 +++++++++++
  2 files changed, 13 insertions(+), 2 deletions(-)
include/linux/compat.h +0 -0
index d0f51baab407..d23ef50454ce 100644

include/linux/compat.h static inline void kmem_cache_destroy(struct kmem_cache *cachep) · line 120

Quoted 4 lines
include/linux/compat.h +1 -0
@@ -117,6 +117,8 @@ static inline void kmem_cache_destroy(struct kmem_cache *cachep)
  	free(cachep);
  }
 
+void *kmemdup(const void *src, size_t size, int flags);

kememdup() is already implemented in fs/ubifs/ubifs.c and used both in drivers/mtd/ and in fs/ubifs/.

Why would you want to use a different signature than both the Linux kernel and current U-Boot?

Quoted 18 lines
include/linux/compat.h +1 -2
+
  #define DECLARE_WAITQUEUE(...)	do { } while (0)
  #define add_wait_queue(...)	do { } while (0)
  #define remove_wait_queue(...)	do { } while (0)
@@ -346,8 +348,6 @@ struct writeback_control {
  	unsigned for_sync:1;		/* sync(2) WB_SYNC_ALL writeback */
  };
 
-void *kmemdup(const void *src, size_t len, gfp_t gfp);
-
  typedef int irqreturn_t;
 
  struct timer_list {};
lib/linux_compat.c +0 -0
index 6373b4451eb3..dd1e5b3c2087 100644

lib/linux_compat.c void *kmem_cache_alloc(struct kmem_cache *obj, int flag) · line 45

Quoted 6 lines
lib/linux_compat.c +3 -0
@@ -40,3 +40,14 @@ void *kmem_cache_alloc(struct kmem_cache *obj, int flag)
  {
  	return malloc_cache_aligned(obj->sz);
  }
+
+void *kmemdup(const void *src, size_t size, int flags)
+{

You should not duplicate the implementation in the UBI filesystem. Instead move the existing implementation to a place that can be linked in your future work.

Best regards

Heinrich

Quoted 9 lines
lib/linux_compat.c +8 -0
+	void *p;
+
+	p = kmalloc(size, flags);
+	if (p)
+		memcpy(p, src, size);
+
+	return p;
+}
#3 lore raw

PATCH 1/11 linux_compat: add kmemdup()

Quoted 18 lines · Heinrich Schuchardt
Quoted 17 lines · AKASHI Takahiro

Adding kmemdup() will help improve portability from linux kernel code (in my case, lib/crypto/x509_cert_parser.c and pkcs7_parser.c).

---
 include/linux/compat.h |  4 ++--
 lib/linux_compat.c     | 11 +++++++++++
 2 files changed, 13 insertions(+), 2 deletions(-)
include/linux/compat.h +0 -0
index d0f51baab407..d23ef50454ce 100644
@@ -117,6 +117,8 @@ static inline void kmem_cache_destroy(struct kmem_cache *cachep)
 	free(cachep);
 }
 
Quoted 4 lines

include/linux/compat.h static inline void kmem_cache_destroy(struct kmem_cache *cachep) · line 120

Quoted 1 line
include/linux/compat.h +1 -0
+void *kmemdup(const void *src, size_t size, int flags);

kememdup() is already implemented in fs/ubifs/ubifs.c and used both in drivers/mtd/ and in fs/ubifs/.

Good catch, I haven't noticed that.
ubifs.c is a very bad place for that function.
# drivers/mtd/mtd*.c doesn't use it as relevant code is guarded by
"#ifndef __UBOOT__" though.

Quoted 15 lines

Why would you want to use a different signature than both the Linux kernel and current U-Boot?

Quoted 12 lines
include/linux/compat.h +1 -2
+
 #define DECLARE_WAITQUEUE(...)	do { } while (0)
 #define add_wait_queue(...)	do { } while (0)
 #define remove_wait_queue(...)	do { } while (0)
@@ -346,8 +348,6 @@ struct writeback_control {
 	unsigned for_sync:1;		/* sync(2) WB_SYNC_ALL writeback */
 };
 
-void *kmemdup(const void *src, size_t len, gfp_t gfp);
-
 typedef int irqreturn_t;
 
Quoted 16 lines

lib/linux_compat.c void *kmem_cache_alloc(struct kmem_cache *obj, int flag) · line 45

Quoted 12 lines
include/linux/compat.h +0 -0
 struct timer_list {};
lib/linux_compat.c +3 -0
index 6373b4451eb3..dd1e5b3c2087 100644
@@ -40,3 +40,14 @@ void *kmem_cache_alloc(struct kmem_cache *obj, int flag)
 {
 	return malloc_cache_aligned(obj->sz);
 }
+
+void *kmemdup(const void *src, size_t size, int flags)
+{

You should not duplicate the implementation in the UBI filesystem. Instead move the existing implementation to a place that can be linked in your future work.

Okay, I'll remove it from ubifs.c.

Thanks,
-Takahiro Akashi

Quoted 14 lines

Best regards

Heinrich

Quoted 9 lines
lib/linux_compat.c +8 -0
+	void *p;
+
+	p = kmalloc(size, flags);
+	if (p)
+		memcpy(p, src, size);
+
+	return p;
+}
#4 lore raw

PATCH 2/11 include: time.h: define time64_t

In reply to AKASHI Takahiro

Adding time64_t definition will help improve portability of linux kernel code (in my case, lib/crypto/x509_cert_parser.c).

---
 include/linux/time.h | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)
include/linux/time.h +24 -0
index b8d298eb4d68..6186985856d7 100644
@@ -150,4 +150,28 @@ _DEFUN (ctime_r, (tim_p, result),
     return asctime_r (localtime_r (tim_p, &tm), result);
 }
 
+/* from <linux>/kernel/time/time.c */
+typedef __s64 time64_t;
+
+inline time64_t mktime64(const unsigned int year0, const unsigned int mon0,
+			 const unsigned int day, const unsigned int hour,
+			 const unsigned int min, const unsigned int sec)
+{
+	unsigned int mon = mon0, year = year0;
+
+	/* 1..12 -> 11,12,1..10 */
+	mon -= 2;
+	if (0 >= (int)mon) {
+		mon += 12;	/* Puts Feb last since it has leap day */
+		year -= 1;
+	}
+
+	return ((((time64_t)
+		  (year / 4 - year / 100 + year / 400 + 367 * mon / 12 + day) +
+		  year * 365 - 719499
+	  ) * 24 + hour /* now have hours - midnight tomorrow handled here */
+	  ) * 60 + min /* now have minutes */
+	) * 60 + sec; /* finally seconds */
+}
+
 #endif
Signature · 1 line
2.21.0
#5 lore raw

PATCH 2/11 include: time.h: define time64_t

Quoted 13 lines · AKASHI Takahiro

Adding time64_t definition will help improve portability of linux kernel code (in my case, lib/crypto/x509_cert_parser.c).

---
  include/linux/time.h | 24 ++++++++++++++++++++++++
  1 file changed, 24 insertions(+)
include/linux/time.h +0 -0
index b8d298eb4d68..6186985856d7 100644

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 154

Quoted 5 lines
include/linux/time.h +2 -0
@@ -150,4 +150,28 @@ _DEFUN (ctime_r, (tim_p, result),
      return asctime_r (localtime_r (tim_p, &tm), result);
  }
 
+/* from <linux>/kernel/time/time.c */
+typedef __s64 time64_t;

Wouldn't we want to put these definitions into a file include/linux/time64.h?

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 158

Quoted 4 lines
include/linux/time.h +4 -0
+
+inline time64_t mktime64(const unsigned int year0, const unsigned int mon0,
+			 const unsigned int day, const unsigned int hour,
+			 const unsigned int min, const unsigned int sec)

Where is the function description?

The Linux kernel does not make this function an inline function. Why should we inline it in U-Boot?

Quoted 6 lines
include/linux/time.h +6 -0
+{
+	unsigned int mon = mon0, year = year0;
+
+	/* 1..12 -> 11,12,1..10 */
+	mon -= 2;
+	if (0 >= (int)mon) {

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 174

Quoted 10 lines
include/linux/time.h +10 -0
+		mon += 12;	/* Puts Feb last since it has leap day */
+		year -= 1;
+	}
+
+	return ((((time64_t)
+		  (year / 4 - year / 100 + year / 400 + 367 * mon / 12 + day) +
+		  year * 365 - 719499
+	  ) * 24 + hour /* now have hours - midnight tomorrow handled here */
+	  ) * 60 + min /* now have minutes */
+	) * 60 + sec; /* finally seconds */

This code is duplicate to rtc_mktime().

Duplication could be avoided by letting rtc_mktime() call mktime64().

Best regards

Heinrich

Quoted 4 lines
include/linux/time.h +2 -0
+}
+
  #endif
#6 lore raw

PATCH 2/11 include: time.h: define time64_t

Quoted 17 lines · Heinrich Schuchardt
Quoted 16 lines · AKASHI Takahiro

Adding time64_t definition will help improve portability of linux kernel code (in my case, lib/crypto/x509_cert_parser.c).

---
 include/linux/time.h | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)
include/linux/time.h +0 -0
index b8d298eb4d68..6186985856d7 100644
@@ -150,4 +150,28 @@ _DEFUN (ctime_r, (tim_p, result),
     return asctime_r (localtime_r (tim_p, &tm), result);
 }
 
Quoted 5 lines

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 154

Quoted 2 lines
include/linux/time.h +2 -0
+/* from <linux>/kernel/time/time.c */
+typedef __s64 time64_t;

Wouldn't we want to put these definitions into a file include/linux/time64.h?

Obviously, there is no problem at following your suggestion, but I hesitate to do so as it adds only one line header file. Moreover, mktime64(), which is the only reason for this patch, is also declared in "linux/time.h" even in linux code.
# Please note that "linux/time64.h" is mainly for timespec64 helper functions, which are never used in U-Boot.

So I'd like to leave as it is. I think that we can re-visit this issue in the future when other definitions in time64.h are required.

Quoted 30 lines

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 158

Quoted 4 lines
include/linux/time.h +4 -0
+
+inline time64_t mktime64(const unsigned int year0, const unsigned int mon0,
+			 const unsigned int day, const unsigned int hour,
+			 const unsigned int min, const unsigned int sec)

Where is the function description?

The Linux kernel does not make this function an inline function. Why should we inline it in U-Boot?

Quoted 6 lines
include/linux/time.h +6 -0
+{
+	unsigned int mon = mon0, year = year0;
+
+	/* 1..12 -> 11,12,1..10 */
+	mon -= 2;
+	if (0 >= (int)mon) {

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 174

Quoted 10 lines
include/linux/time.h +10 -0
+		mon += 12;	/* Puts Feb last since it has leap day */
+		year -= 1;
+	}
+
+	return ((((time64_t)
+		  (year / 4 - year / 100 + year / 400 + 367 * mon / 12 + day) +
+		  year * 365 - 719499
+	  ) * 24 + hour /* now have hours - midnight tomorrow handled here */
+	  ) * 60 + min /* now have minutes */
+	) * 60 + sec; /* finally seconds */

This code is duplicate to rtc_mktime().

Quoted 1 line

Duplication could be avoided by letting rtc_mktime() call mktime64().

Okay, but as an inline function in this case.
In addition, drivers/rtc/date.c will be moved to lib/ for general use with CONFIG_LIB_DATE.

Thanks,
-Takahiro Akashi

Quoted 9 lines

Best regards

Heinrich

Quoted 4 lines
include/linux/time.h +2 -0
+}
+
 #endif
#7 lore raw

PATCH 2/11 include: time.h: define time64_t

Quoted 64 lines · AKASHI Takahiro
Quoted 17 lines · Heinrich Schuchardt
Quoted 16 lines · AKASHI Takahiro

Adding time64_t definition will help improve portability of linux kernel code (in my case, lib/crypto/x509_cert_parser.c).

---
  include/linux/time.h | 24 ++++++++++++++++++++++++
  1 file changed, 24 insertions(+)
include/linux/time.h +0 -0
index b8d298eb4d68..6186985856d7 100644
@@ -150,4 +150,28 @@ _DEFUN (ctime_r, (tim_p, result),
      return asctime_r (localtime_r (tim_p, &tm), result);
  }
 
Quoted 5 lines

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 154

Quoted 2 lines
include/linux/time.h +2 -0
+/* from <linux>/kernel/time/time.c */
+typedef __s64 time64_t;

Wouldn't we want to put these definitions into a file include/linux/time64.h?

Obviously, there is no problem at following your suggestion, but I hesitate to do so as it adds only one line header file. Moreover, mktime64(), which is the only reason for this patch, is also declared in "linux/time.h" even in linux code.
# Please note that "linux/time64.h" is mainly for timespec64 helper functions, which are never used in U-Boot.

So I'd like to leave as it is. I think that we can re-visit this issue in the future when other definitions in time64.h are required.

Quoted 30 lines

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 158

Quoted 4 lines
include/linux/time.h +4 -0
+
+inline time64_t mktime64(const unsigned int year0, const unsigned int mon0,
+			 const unsigned int day, const unsigned int hour,
+			 const unsigned int min, const unsigned int sec)

Where is the function description?

The Linux kernel does not make this function an inline function. Why should we inline it in U-Boot?

Quoted 6 lines
include/linux/time.h +6 -0
+{
+	unsigned int mon = mon0, year = year0;
+
+	/* 1..12 -> 11,12,1..10 */
+	mon -= 2;
+	if (0 >= (int)mon) {

include/linux/time.h _DEFUN (ctime_r, (tim_p, result), · line 174

Quoted 10 lines
include/linux/time.h +10 -0
+		mon += 12;	/* Puts Feb last since it has leap day */
+		year -= 1;
+	}
+
+	return ((((time64_t)
+		  (year / 4 - year / 100 + year / 400 + 367 * mon / 12 + day) +
+		  year * 365 - 719499
+	  ) * 24 + hour /* now have hours - midnight tomorrow handled here */
+	  ) * 60 + min /* now have minutes */
+	) * 60 + sec; /* finally seconds */

This code is duplicate to rtc_mktime().

Quoted 3 lines
Quoted 1 line

Duplication could be avoided by letting rtc_mktime() call mktime64().

Okay, but as an inline function in this case.

Inline use in two places adds more bytes to the binary. U-Boot should stay as small as possible.

Best regards

Heinrich

Quoted 17 lines

In addition, drivers/rtc/date.c will be moved to lib/ for general use with CONFIG_LIB_DATE.

Thanks,
-Takahiro Akashi

Quoted 9 lines

Best regards

Heinrich

Quoted 4 lines
include/linux/time.h +2 -0
+}
+
  #endif
#8 lore raw

PATCH 3/11 include: kernel.h: include printk.h

In reply to AKASHI Takahiro

Adding "printk.h" will help improve portability from linux kernel code (in my case, lib/asn1_decoder.c).

---
 include/linux/kernel.h | 2 ++
 1 file changed, 2 insertions(+)
include/linux/kernel.h +2 -0
index a85c15d8dc28..919d12bdf89c 100644
@@ -2,7 +2,9 @@
 #define _LINUX_KERNEL_H
 
 
+#include <vsprintf.h> /* for printf utilities */
 #include <linux/types.h>
+#include <linux/printk.h>
 
 #define USHRT_MAX	((u16)(~0U))
 #define SHRT_MAX	((s16)(USHRT_MAX>>1))
Signature · 1 line
2.21.0
#9 lore raw

PATCH 3/11 include: kernel.h: include printk.h

Quoted 13 lines · AKASHI Takahiro

Adding "printk.h" will help improve portability from linux kernel code (in my case, lib/asn1_decoder.c).

---
  include/linux/kernel.h | 2 ++
  1 file changed, 2 insertions(+)
include/linux/kernel.h +0 -0
index a85c15d8dc28..919d12bdf89c 100644

include/linux/kernel.h line 5

Quoted 4 lines
include/linux/kernel.h +1 -0
@@ -2,7 +2,9 @@
  #define _LINUX_KERNEL_H
 
 
+#include <vsprintf.h> /* for printf utilities */

Kernel code has no vsprintf.h. Linux's lib/asn1_decoder.c has not a single print statement. So why are you inserting vsprintf.h here?

include/linux/kernel.h line 7

Quoted 2 lines
include/linux/kernel.h +1 -0
  #include <linux/types.h>
+#include <linux/printk.h>

Linux include/linux/kernel.h also includes printk.h. OK.

Best regards

Heinrich

Quoted 4 lines
include/linux/kernel.h +0 -0
  #define USHRT_MAX	((u16)(~0U))
  #define SHRT_MAX	((s16)(USHRT_MAX>>1))
 
#10 lore raw

PATCH 3/11 include: kernel.h: include printk.h

Quoted 17 lines · Heinrich Schuchardt
Quoted 16 lines · AKASHI Takahiro

Adding "printk.h" will help improve portability from linux kernel code (in my case, lib/asn1_decoder.c).

---
 include/linux/kernel.h | 2 ++
 1 file changed, 2 insertions(+)
include/linux/kernel.h +0 -0
index a85c15d8dc28..919d12bdf89c 100644
@@ -2,7 +2,9 @@
 #define _LINUX_KERNEL_H
 
 
Quoted 4 lines

include/linux/kernel.h line 5

Quoted 1 line
include/linux/kernel.h +1 -0
+#include <vsprintf.h> /* for printf utilities */

Kernel code has no vsprintf.h. Linux's lib/asn1_decoder.c has not a single print statement. So why are you inserting vsprintf.h here?

I don't remember why I mentioned to asn1_decoder.c here, but When this statement is removed from kernel.h, some files cannot be compiled.
The fact is that sprintf() is used in time.h without including any related headers.
(There are still bunch of *missing headers* issues in U-Boot.)

I will add one more patch against time.h here.

Thanks,
-Takahiro Akashi

Quoted 13 lines

include/linux/kernel.h line 7

Quoted 2 lines
include/linux/kernel.h +1 -0
 #include <linux/types.h>
+#include <linux/printk.h>

Linux include/linux/kernel.h also includes printk.h. OK.

Best regards

Heinrich

Quoted 4 lines
include/linux/kernel.h +0 -0
 #define USHRT_MAX	((u16)(~0U))
 #define SHRT_MAX	((s16)(USHRT_MAX>>1))
 
#11 lore raw

PATCH 3/11 include: kernel.h: include printk.h

Quoted 18 lines · AKASHI Takahiro
Quoted 17 lines · Heinrich Schuchardt
Quoted 16 lines · AKASHI Takahiro

Adding "printk.h" will help improve portability from linux kernel code (in my case, lib/asn1_decoder.c).

---
 include/linux/kernel.h | 2 ++
 1 file changed, 2 insertions(+)
include/linux/kernel.h +0 -0
index a85c15d8dc28..919d12bdf89c 100644
@@ -2,7 +2,9 @@
 #define _LINUX_KERNEL_H
 
 
Quoted 6 lines
Quoted 4 lines

include/linux/kernel.h line 5

Quoted 1 line
include/linux/kernel.h +1 -0
+#include <vsprintf.h> /* for printf utilities */

Kernel code has no vsprintf.h. Linux's lib/asn1_decoder.c has not a single print statement. So why are you inserting vsprintf.h here?

I don't remember why I mentioned to asn1_decoder.c here, but

For the record, asn1_parser.c uses pr_debug/pr_devel and others in many places.

-Takahiro Akashi

Quoted 25 lines

When this statement is removed from kernel.h, some files cannot be compiled.
The fact is that sprintf() is used in time.h without including any related headers.
(There are still bunch of *missing headers* issues in U-Boot.)

I will add one more patch against time.h here.

Thanks,
-Takahiro Akashi

Quoted 13 lines

include/linux/kernel.h line 7

Quoted 2 lines
include/linux/kernel.h +1 -0
 #include <linux/types.h>
+#include <linux/printk.h>

Linux include/linux/kernel.h also includes printk.h. OK.

Best regards

Heinrich

Quoted 4 lines
include/linux/kernel.h +0 -0
 #define USHRT_MAX	((u16)(~0U))
 #define SHRT_MAX	((s16)(USHRT_MAX>>1))
 
#12 lore raw

PATCH 4/11 cmd: add asn1_compiler

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

This host command will be used to create a ASN1 parser, for example, for pkcs7 or x509.

---
 cmd/Kconfig                       |    3 +
 include/linux/asn1.h              |   65 ++
 include/linux/asn1_ber_bytecode.h |   89 ++
 include/linux/asn1_decoder.h      |   20 +
 scripts/Makefile                  |    3 +
 scripts/asn1_compiler.c           | 1611 +++++++++++++++++++++++++++++
 6 files changed, 1791 insertions(+)
 create mode 100644 include/linux/asn1.h
 create mode 100644 include/linux/asn1_ber_bytecode.h
 create mode 100644 include/linux/asn1_decoder.h
 create mode 100644 scripts/asn1_compiler.c
cmd/Kconfig +3 -0
index 4e61565aab16..605dcfdf4282 100644
@@ -184,6 +184,9 @@ endmenu
 config BUILD_BIN2C
 	bool
 
+config BUILD_ASN1
+	bool
+
 comment "Commands"
 
 menu "Info commands"
include/linux/asn1.h new file +65 -0

65 lines not on this page · raw message

include/linux/asn1_ber_bytecode.h new file +89 -0

89 lines not on this page · raw message

include/linux/asn1_decoder.h new file +20 -0
new file mode 100644
index 000000000000..83f9c6e1e5e9
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 decoder
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells ([email protected])
+ */
+
+#ifndef _LINUX_ASN1_DECODER_H
+#define _LINUX_ASN1_DECODER_H
+
+#include <linux/asn1.h>
+
+struct asn1_decoder;
+
+extern int asn1_ber_decoder(const struct asn1_decoder *decoder,
+			    void *context,
+			    const unsigned char *data,
+			    size_t datalen);
+
+#endif /* _LINUX_ASN1_DECODER_H */
scripts/Makefile +3 -0
index e7b353f77f43..c4e2ae5db2ea 100644
@@ -5,6 +5,9 @@
 # ---------------------------------------------------------------------------
 
 hostprogs-$(CONFIG_BUILD_BIN2C)		+= bin2c
+hostprogs-$(CONFIG_BUILD_ASN1)		+= asn1_compiler
+
+HOSTCFLAGS_asn1_compiler.o = -idirafter $(srctree)/include
 
 always		:= $(hostprogs-y)
 
scripts/asn1_compiler.c new file +1611 -0

1,611 lines not on this page · raw message

Signature · 1 line
2.21.0
#13 lore raw

PATCH 4/11 cmd: add asn1_compiler

Quoted 5 lines · AKASHI Takahiro

Imported from linux kernel v5.3.

This host command will be used to create a ASN1 parser, for example, for pkcs7 or x509.

Quoted 8 lines
---
  cmd/Kconfig                       |    3 +
  include/linux/asn1.h              |   65 ++
  include/linux/asn1_ber_bytecode.h |   89 ++
  include/linux/asn1_decoder.h      |   20 +
  scripts/Makefile                  |    3 +
  scripts/asn1_compiler.c           | 1611 +++++++++++++++++++++++++++++

Except for Makefile and Kconfig these are verbatime copies of the Linux files. OK.

Quoted 217 lines
  6 files changed, 1791 insertions(+)

create mode 100644 include/linux/asn1.h create mode 100644 include/linux/asn1_ber_bytecode.h create mode 100644 include/linux/asn1_decoder.h create mode 100644 scripts/asn1_compiler.c

cmd/Kconfig +3 -0
index 4e61565aab16..605dcfdf4282 100644
@@ -184,6 +184,9 @@ endmenu
  config BUILD_BIN2C
  	bool
 
+config BUILD_ASN1
+	bool
+
  comment "Commands"
 
  menu "Info commands"
include/linux/asn1.h new file +65 -0
new file mode 100644
index 000000000000..a4d0bdd10711
@@ -0,0 +1,65 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 BER/DER/CER encoding definitions
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_ASN1_H
+#define _LINUX_ASN1_H
+
+/* Class */
+enum asn1_class {
+	ASN1_UNIV	= 0,	/* Universal */
+	ASN1_APPL	= 1,	/* Application */
+	ASN1_CONT	= 2,	/* Context */
+	ASN1_PRIV	= 3	/* Private */
+};
+#define ASN1_CLASS_BITS	0xc0
+
+
+enum asn1_method {
+	ASN1_PRIM	= 0,	/* Primitive */
+	ASN1_CONS	= 1	/* Constructed */
+};
+#define ASN1_CONS_BIT	0x20
+
+/* Tag */
+enum asn1_tag {
+	ASN1_EOC	= 0,	/* End Of Contents or N/A */
+	ASN1_BOOL	= 1,	/* Boolean */
+	ASN1_INT	= 2,	/* Integer */
+	ASN1_BTS	= 3,	/* Bit String */
+	ASN1_OTS	= 4,	/* Octet String */
+	ASN1_NULL	= 5,	/* Null */
+	ASN1_OID	= 6,	/* Object Identifier  */
+	ASN1_ODE	= 7,	/* Object Description */
+	ASN1_EXT	= 8,	/* External */
+	ASN1_REAL	= 9,	/* Real float */
+	ASN1_ENUM	= 10,	/* Enumerated */
+	ASN1_EPDV	= 11,	/* Embedded PDV */
+	ASN1_UTF8STR	= 12,	/* UTF8 String */
+	ASN1_RELOID	= 13,	/* Relative OID */
+	/* 14 - Reserved */
+	/* 15 - Reserved */
+	ASN1_SEQ	= 16,	/* Sequence and Sequence of */
+	ASN1_SET	= 17,	/* Set and Set of */
+	ASN1_NUMSTR	= 18,	/* Numerical String */
+	ASN1_PRNSTR	= 19,	/* Printable String */
+	ASN1_TEXSTR	= 20,	/* T61 String / Teletext String */
+	ASN1_VIDSTR	= 21,	/* Videotex String */
+	ASN1_IA5STR	= 22,	/* IA5 String */
+	ASN1_UNITIM	= 23,	/* Universal Time */
+	ASN1_GENTIM	= 24,	/* General Time */
+	ASN1_GRASTR	= 25,	/* Graphic String */
+	ASN1_VISSTR	= 26,	/* Visible String */
+	ASN1_GENSTR	= 27,	/* General String */
+	ASN1_UNISTR	= 28,	/* Universal String */
+	ASN1_CHRSTR	= 29,	/* Character String */
+	ASN1_BMPSTR	= 30,	/* BMP String */
+	ASN1_LONG_TAG	= 31	/* Long form tag */
+};
+
+#define ASN1_INDEFINITE_LENGTH 0x80
+
+#endif /* _LINUX_ASN1_H */
include/linux/asn1_ber_bytecode.h new file +89 -0
new file mode 100644
index 000000000000..b38361953a48
@@ -0,0 +1,89 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 BER/DER/CER parsing state machine internal definitions
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_ASN1_BER_BYTECODE_H
+#define _LINUX_ASN1_BER_BYTECODE_H
+
+#ifdef __KERNEL__
+#include <linux/types.h>
+#endif
+#include <linux/asn1.h>
+
+typedef int (*asn1_action_t)(void *context,
+			     size_t hdrlen, /* In case of ANY type */
+			     unsigned char tag, /* In case of ANY type */
+			     const void *value, size_t vlen);
+
+struct asn1_decoder {
+	const unsigned char *machine;
+	size_t machlen;
+	const asn1_action_t *actions;
+};
+
+enum asn1_opcode {
+	/* The tag-matching ops come first and the odd-numbered slots
+	 * are for OR_SKIP ops.
+	 */
+#define ASN1_OP_MATCH__SKIP		  0x01
+#define ASN1_OP_MATCH__ACT		  0x02
+#define ASN1_OP_MATCH__JUMP		  0x04
+#define ASN1_OP_MATCH__ANY		  0x08
+#define ASN1_OP_MATCH__COND		  0x10
+
+	ASN1_OP_MATCH			= 0x00,
+	ASN1_OP_MATCH_OR_SKIP		= 0x01,
+	ASN1_OP_MATCH_ACT		= 0x02,
+	ASN1_OP_MATCH_ACT_OR_SKIP	= 0x03,
+	ASN1_OP_MATCH_JUMP		= 0x04,
+	ASN1_OP_MATCH_JUMP_OR_SKIP	= 0x05,
+	ASN1_OP_MATCH_ANY		= 0x08,
+	ASN1_OP_MATCH_ANY_OR_SKIP	= 0x09,
+	ASN1_OP_MATCH_ANY_ACT		= 0x0a,
+	ASN1_OP_MATCH_ANY_ACT_OR_SKIP	= 0x0b,
+	/* Everything before here matches unconditionally */
+
+	ASN1_OP_COND_MATCH_OR_SKIP	= 0x11,
+	ASN1_OP_COND_MATCH_ACT_OR_SKIP	= 0x13,
+	ASN1_OP_COND_MATCH_JUMP_OR_SKIP	= 0x15,
+	ASN1_OP_COND_MATCH_ANY		= 0x18,
+	ASN1_OP_COND_MATCH_ANY_OR_SKIP	= 0x19,
+	ASN1_OP_COND_MATCH_ANY_ACT	= 0x1a,
+	ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP = 0x1b,
+
+	/* Everything before here will want a tag from the data */
+#define ASN1_OP__MATCHES_TAG ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP
+
+	/* These are here to help fill up space */
+	ASN1_OP_COND_FAIL		= 0x1c,
+	ASN1_OP_COMPLETE		= 0x1d,
+	ASN1_OP_ACT			= 0x1e,
+	ASN1_OP_MAYBE_ACT		= 0x1f,
+
+	/* The following eight have bit 0 -> SET, 1 -> OF, 2 -> ACT */
+	ASN1_OP_END_SEQ			= 0x20,
+	ASN1_OP_END_SET			= 0x21,
+	ASN1_OP_END_SEQ_OF		= 0x22,
+	ASN1_OP_END_SET_OF		= 0x23,
+	ASN1_OP_END_SEQ_ACT		= 0x24,
+	ASN1_OP_END_SET_ACT		= 0x25,
+	ASN1_OP_END_SEQ_OF_ACT		= 0x26,
+	ASN1_OP_END_SET_OF_ACT		= 0x27,
+#define ASN1_OP_END__SET		  0x01
+#define ASN1_OP_END__OF			  0x02
+#define ASN1_OP_END__ACT		  0x04
+
+	ASN1_OP_RETURN			= 0x28,
+
+	ASN1_OP__NR
+};
+
+#define _tag(CLASS, CP, TAG) ((ASN1_##CLASS << 6) | (ASN1_##CP << 5) | ASN1_##TAG)
+#define _tagn(CLASS, CP, TAG) ((ASN1_##CLASS << 6) | (ASN1_##CP << 5) | TAG)
+#define _jump_target(N) (N)
+#define _action(N) (N)
+
+#endif /* _LINUX_ASN1_BER_BYTECODE_H */
include/linux/asn1_decoder.h new file +20 -0
new file mode 100644
index 000000000000..83f9c6e1e5e9
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 decoder
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_ASN1_DECODER_H
+#define _LINUX_ASN1_DECODER_H
+
+#include <linux/asn1.h>
+
+struct asn1_decoder;
+
+extern int asn1_ber_decoder(const struct asn1_decoder *decoder,
+			    void *context,
+			    const unsigned char *data,
+			    size_t datalen);
+
+#endif /* _LINUX_ASN1_DECODER_H */
scripts/Makefile +0 -0
index e7b353f77f43..c4e2ae5db2ea 100644

scripts/Makefile line 10

Quoted 6 lines
scripts/Makefile +3 -0
@@ -5,6 +5,9 @@
  # ---------------------------------------------------------------------------
 
  hostprogs-$(CONFIG_BUILD_BIN2C)		+= bin2c
+hostprogs-$(CONFIG_BUILD_ASN1)		+= asn1_compiler
+
+HOSTCFLAGS_asn1_compiler.o = -idirafter $(srctree)/include

Which problems did you run into that make you use -idirafter here?

Otherwise

Quoted 1,621 lines
scripts/Makefile +0 -0
  always		:= $(hostprogs-y)
 
scripts/asn1_compiler.c new file +1611 -0
new file mode 100644
index 000000000000..adabd4145264
@@ -0,0 +1,1611 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* Simplified ASN.1 notation parser
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <string.h>
+#include <ctype.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <sys/stat.h>
+#include <linux/asn1_ber_bytecode.h>
+
+enum token_type {
+	DIRECTIVE_ABSENT,
+	DIRECTIVE_ALL,
+	DIRECTIVE_ANY,
+	DIRECTIVE_APPLICATION,
+	DIRECTIVE_AUTOMATIC,
+	DIRECTIVE_BEGIN,
+	DIRECTIVE_BIT,
+	DIRECTIVE_BMPString,
+	DIRECTIVE_BOOLEAN,
+	DIRECTIVE_BY,
+	DIRECTIVE_CHARACTER,
+	DIRECTIVE_CHOICE,
+	DIRECTIVE_CLASS,
+	DIRECTIVE_COMPONENT,
+	DIRECTIVE_COMPONENTS,
+	DIRECTIVE_CONSTRAINED,
+	DIRECTIVE_CONTAINING,
+	DIRECTIVE_DEFAULT,
+	DIRECTIVE_DEFINED,
+	DIRECTIVE_DEFINITIONS,
+	DIRECTIVE_EMBEDDED,
+	DIRECTIVE_ENCODED,
+	DIRECTIVE_ENCODING_CONTROL,
+	DIRECTIVE_END,
+	DIRECTIVE_ENUMERATED,
+	DIRECTIVE_EXCEPT,
+	DIRECTIVE_EXPLICIT,
+	DIRECTIVE_EXPORTS,
+	DIRECTIVE_EXTENSIBILITY,
+	DIRECTIVE_EXTERNAL,
+	DIRECTIVE_FALSE,
+	DIRECTIVE_FROM,
+	DIRECTIVE_GeneralString,
+	DIRECTIVE_GeneralizedTime,
+	DIRECTIVE_GraphicString,
+	DIRECTIVE_IA5String,
+	DIRECTIVE_IDENTIFIER,
+	DIRECTIVE_IMPLICIT,
+	DIRECTIVE_IMPLIED,
+	DIRECTIVE_IMPORTS,
+	DIRECTIVE_INCLUDES,
+	DIRECTIVE_INSTANCE,
+	DIRECTIVE_INSTRUCTIONS,
+	DIRECTIVE_INTEGER,
+	DIRECTIVE_INTERSECTION,
+	DIRECTIVE_ISO646String,
+	DIRECTIVE_MAX,
+	DIRECTIVE_MIN,
+	DIRECTIVE_MINUS_INFINITY,
+	DIRECTIVE_NULL,
+	DIRECTIVE_NumericString,
+	DIRECTIVE_OBJECT,
+	DIRECTIVE_OCTET,
+	DIRECTIVE_OF,
+	DIRECTIVE_OPTIONAL,
+	DIRECTIVE_ObjectDescriptor,
+	DIRECTIVE_PATTERN,
+	DIRECTIVE_PDV,
+	DIRECTIVE_PLUS_INFINITY,
+	DIRECTIVE_PRESENT,
+	DIRECTIVE_PRIVATE,
+	DIRECTIVE_PrintableString,
+	DIRECTIVE_REAL,
+	DIRECTIVE_RELATIVE_OID,
+	DIRECTIVE_SEQUENCE,
+	DIRECTIVE_SET,
+	DIRECTIVE_SIZE,
+	DIRECTIVE_STRING,
+	DIRECTIVE_SYNTAX,
+	DIRECTIVE_T61String,
+	DIRECTIVE_TAGS,
+	DIRECTIVE_TRUE,
+	DIRECTIVE_TeletexString,
+	DIRECTIVE_UNION,
+	DIRECTIVE_UNIQUE,
+	DIRECTIVE_UNIVERSAL,
+	DIRECTIVE_UTCTime,
+	DIRECTIVE_UTF8String,
+	DIRECTIVE_UniversalString,
+	DIRECTIVE_VideotexString,
+	DIRECTIVE_VisibleString,
+	DIRECTIVE_WITH,
+	NR__DIRECTIVES,
+	TOKEN_ASSIGNMENT = NR__DIRECTIVES,
+	TOKEN_OPEN_CURLY,
+	TOKEN_CLOSE_CURLY,
+	TOKEN_OPEN_SQUARE,
+	TOKEN_CLOSE_SQUARE,
+	TOKEN_OPEN_ACTION,
+	TOKEN_CLOSE_ACTION,
+	TOKEN_COMMA,
+	TOKEN_NUMBER,
+	TOKEN_TYPE_NAME,
+	TOKEN_ELEMENT_NAME,
+	NR__TOKENS
+};
+
+static const unsigned char token_to_tag[NR__TOKENS] = {
+	/* EOC goes first */
+	[DIRECTIVE_BOOLEAN]		= ASN1_BOOL,
+	[DIRECTIVE_INTEGER]		= ASN1_INT,
+	[DIRECTIVE_BIT]			= ASN1_BTS,
+	[DIRECTIVE_OCTET]		= ASN1_OTS,
+	[DIRECTIVE_NULL]		= ASN1_NULL,
+	[DIRECTIVE_OBJECT]		= ASN1_OID,
+	[DIRECTIVE_ObjectDescriptor]	= ASN1_ODE,
+	[DIRECTIVE_EXTERNAL]		= ASN1_EXT,
+	[DIRECTIVE_REAL]		= ASN1_REAL,
+	[DIRECTIVE_ENUMERATED]		= ASN1_ENUM,
+	[DIRECTIVE_EMBEDDED]		= 0,
+	[DIRECTIVE_UTF8String]		= ASN1_UTF8STR,
+	[DIRECTIVE_RELATIVE_OID]	= ASN1_RELOID,
+	/* 14 */
+	/* 15 */
+	[DIRECTIVE_SEQUENCE]		= ASN1_SEQ,
+	[DIRECTIVE_SET]			= ASN1_SET,
+	[DIRECTIVE_NumericString]	= ASN1_NUMSTR,
+	[DIRECTIVE_PrintableString]	= ASN1_PRNSTR,
+	[DIRECTIVE_T61String]		= ASN1_TEXSTR,
+	[DIRECTIVE_TeletexString]	= ASN1_TEXSTR,
+	[DIRECTIVE_VideotexString]	= ASN1_VIDSTR,
+	[DIRECTIVE_IA5String]		= ASN1_IA5STR,
+	[DIRECTIVE_UTCTime]		= ASN1_UNITIM,
+	[DIRECTIVE_GeneralizedTime]	= ASN1_GENTIM,
+	[DIRECTIVE_GraphicString]	= ASN1_GRASTR,
+	[DIRECTIVE_VisibleString]	= ASN1_VISSTR,
+	[DIRECTIVE_GeneralString]	= ASN1_GENSTR,
+	[DIRECTIVE_UniversalString]	= ASN1_UNITIM,
+	[DIRECTIVE_CHARACTER]		= ASN1_CHRSTR,
+	[DIRECTIVE_BMPString]		= ASN1_BMPSTR,
+};
+
+static const char asn1_classes[4][5] = {
+	[ASN1_UNIV]	= "UNIV",
+	[ASN1_APPL]	= "APPL",
+	[ASN1_CONT]	= "CONT",
+	[ASN1_PRIV]	= "PRIV"
+};
+
+static const char asn1_methods[2][5] = {
+	[ASN1_UNIV]	= "PRIM",
+	[ASN1_APPL]	= "CONS"
+};
+
+static const char *const asn1_universal_tags[32] = {
+	"EOC",
+	"BOOL",
+	"INT",
+	"BTS",
+	"OTS",
+	"NULL",
+	"OID",
+	"ODE",
+	"EXT",
+	"REAL",
+	"ENUM",
+	"EPDV",
+	"UTF8STR",
+	"RELOID",
+	NULL,		/* 14 */
+	NULL,		/* 15 */
+	"SEQ",
+	"SET",
+	"NUMSTR",
+	"PRNSTR",
+	"TEXSTR",
+	"VIDSTR",
+	"IA5STR",
+	"UNITIM",
+	"GENTIM",
+	"GRASTR",
+	"VISSTR",
+	"GENSTR",
+	"UNISTR",
+	"CHRSTR",
+	"BMPSTR",
+	NULL		/* 31 */
+};
+
+static const char *filename;
+static const char *grammar_name;
+static const char *outputname;
+static const char *headername;
+
+static const char *const directives[NR__DIRECTIVES] = {
+#define _(X) [DIRECTIVE_##X] = #X
+	_(ABSENT),
+	_(ALL),
+	_(ANY),
+	_(APPLICATION),
+	_(AUTOMATIC),
+	_(BEGIN),
+	_(BIT),
+	_(BMPString),
+	_(BOOLEAN),
+	_(BY),
+	_(CHARACTER),
+	_(CHOICE),
+	_(CLASS),
+	_(COMPONENT),
+	_(COMPONENTS),
+	_(CONSTRAINED),
+	_(CONTAINING),
+	_(DEFAULT),
+	_(DEFINED),
+	_(DEFINITIONS),
+	_(EMBEDDED),
+	_(ENCODED),
+	[DIRECTIVE_ENCODING_CONTROL] = "ENCODING-CONTROL",
+	_(END),
+	_(ENUMERATED),
+	_(EXCEPT),
+	_(EXPLICIT),
+	_(EXPORTS),
+	_(EXTENSIBILITY),
+	_(EXTERNAL),
+	_(FALSE),
+	_(FROM),
+	_(GeneralString),
+	_(GeneralizedTime),
+	_(GraphicString),
+	_(IA5String),
+	_(IDENTIFIER),
+	_(IMPLICIT),
+	_(IMPLIED),
+	_(IMPORTS),
+	_(INCLUDES),
+	_(INSTANCE),
+	_(INSTRUCTIONS),
+	_(INTEGER),
+	_(INTERSECTION),
+	_(ISO646String),
+	_(MAX),
+	_(MIN),
+	[DIRECTIVE_MINUS_INFINITY] = "MINUS-INFINITY",
+	[DIRECTIVE_NULL] = "NULL",
+	_(NumericString),
+	_(OBJECT),
+	_(OCTET),
+	_(OF),
+	_(OPTIONAL),
+	_(ObjectDescriptor),
+	_(PATTERN),
+	_(PDV),
+	[DIRECTIVE_PLUS_INFINITY] = "PLUS-INFINITY",
+	_(PRESENT),
+	_(PRIVATE),
+	_(PrintableString),
+	_(REAL),
+	[DIRECTIVE_RELATIVE_OID] = "RELATIVE-OID",
+	_(SEQUENCE),
+	_(SET),
+	_(SIZE),
+	_(STRING),
+	_(SYNTAX),
+	_(T61String),
+	_(TAGS),
+	_(TRUE),
+	_(TeletexString),
+	_(UNION),
+	_(UNIQUE),
+	_(UNIVERSAL),
+	_(UTCTime),
+	_(UTF8String),
+	_(UniversalString),
+	_(VideotexString),
+	_(VisibleString),
+	_(WITH)
+};
+
+struct action {
+	struct action	*next;
+	char		*name;
+	unsigned char	index;
+};
+
+static struct action *action_list;
+static unsigned nr_actions;
+
+struct token {
+	unsigned short	line;
+	enum token_type	token_type : 8;
+	unsigned char	size;
+	struct action	*action;
+	char		*content;
+	struct type	*type;
+};
+
+static struct token *token_list;
+static unsigned nr_tokens;
+static bool verbose_opt;
+static bool debug_opt;
+
+#define verbose(fmt, ...) do { if (verbose_opt) printf(fmt, ## __VA_ARGS__); } while (0)
+#define debug(fmt, ...) do { if (debug_opt) printf(fmt, ## __VA_ARGS__); } while (0)
+
+static int directive_compare(const void *_key, const void *_pdir)
+{
+	const struct token *token = _key;
+	const char *const *pdir = _pdir, *dir = *pdir;
+	size_t dlen, clen;
+	int val;
+
+	dlen = strlen(dir);
+	clen = (dlen < token->size) ? dlen : token->size;
+
+	//debug("cmp(%s,%s) = ", token->content, dir);
+
+	val = memcmp(token->content, dir, clen);
+	if (val != 0) {
+		//debug("%d [cmp]\n", val);
+		return val;
+	}
+
+	if (dlen == token->size) {
+		//debug("0\n");
+		return 0;
+	}
+	//debug("%d\n", (int)dlen - (int)token->size);
+	return dlen - token->size; /* shorter -> negative */
+}
+
+/*
+ * Tokenise an ASN.1 grammar
+ */
+static void tokenise(char *buffer, char *end)
+{
+	struct token *tokens;
+	char *line, *nl, *start, *p, *q;
+	unsigned tix, lineno;
+
+	/* Assume we're going to have half as many tokens as we have
+	 * characters
+	 */
+	token_list = tokens = calloc((end - buffer) / 2, sizeof(struct token));
+	if (!tokens) {
+		perror(NULL);
+		exit(1);
+	}
+	tix = 0;
+
+	lineno = 0;
+	while (buffer < end) {
+		/* First of all, break out a line */
+		lineno++;
+		line = buffer;
+		nl = memchr(line, '\n', end - buffer);
+		if (!nl) {
+			buffer = nl = end;
+		} else {
+			buffer = nl + 1;
+			*nl = '\0';
+		}
+
+		/* Remove "--" comments */
+		p = line;
+	next_comment:
+		while ((p = memchr(p, '-', nl - p))) {
+			if (p[1] == '-') {
+				/* Found a comment; see if there's a terminator */
+				q = p + 2;
+				while ((q = memchr(q, '-', nl - q))) {
+					if (q[1] == '-') {
+						/* There is - excise the comment */
+						q += 2;
+						memmove(p, q, nl - q);
+						goto next_comment;
+					}
+					q++;
+				}
+				*p = '\0';
+				nl = p;
+				break;
+			} else {
+				p++;
+			}
+		}
+
+		p = line;
+		while (p < nl) {
+			/* Skip white space */
+			while (p < nl && isspace(*p))
+				*(p++) = 0;
+			if (p >= nl)
+				break;
+
+			tokens[tix].line = lineno;
+			start = p;
+
+			/* Handle string tokens */
+			if (isalpha(*p)) {
+				const char **dir;
+
+				/* Can be a directive, type name or element
+				 * name.  Find the end of the name.
+				 */
+				q = p + 1;
+				while (q < nl && (isalnum(*q) || *q == '-' || *q == '_'))
+					q++;
+				tokens[tix].size = q - p;
+				p = q;
+
+				tokens[tix].content = malloc(tokens[tix].size + 1);
+				if (!tokens[tix].content) {
+					perror(NULL);
+					exit(1);
+				}
+				memcpy(tokens[tix].content, start, tokens[tix].size);
+				tokens[tix].content[tokens[tix].size] = 0;
+
+				/* If it begins with a lowercase letter then
+				 * it's an element name
+				 */
+				if (islower(tokens[tix].content[0])) {
+					tokens[tix++].token_type = TOKEN_ELEMENT_NAME;
+					continue;
+				}
+
+				/* Otherwise we need to search the directive
+				 * table
+				 */
+				dir = bsearch(&tokens[tix], directives,
+					      sizeof(directives) / sizeof(directives[1]),
+					      sizeof(directives[1]),
+					      directive_compare);
+				if (dir) {
+					tokens[tix++].token_type = dir - directives;
+					continue;
+				}
+
+				tokens[tix++].token_type = TOKEN_TYPE_NAME;
+				continue;
+			}
+
+			/* Handle numbers */
+			if (isdigit(*p)) {
+				/* Find the end of the number */
+				q = p + 1;
+				while (q < nl && (isdigit(*q)))
+					q++;
+				tokens[tix].size = q - p;
+				p = q;
+				tokens[tix].content = malloc(tokens[tix].size + 1);
+				if (!tokens[tix].content) {
+					perror(NULL);
+					exit(1);
+				}
+				memcpy(tokens[tix].content, start, tokens[tix].size);
+				tokens[tix].content[tokens[tix].size] = 0;
+				tokens[tix++].token_type = TOKEN_NUMBER;
+				continue;
+			}
+
+			if (nl - p >= 3) {
+				if (memcmp(p, "::=", 3) == 0) {
+					p += 3;
+					tokens[tix].size = 3;
+					tokens[tix].content = "::=";
+					tokens[tix++].token_type = TOKEN_ASSIGNMENT;
+					continue;
+				}
+			}
+
+			if (nl - p >= 2) {
+				if (memcmp(p, "({", 2) == 0) {
+					p += 2;
+					tokens[tix].size = 2;
+					tokens[tix].content = "({";
+					tokens[tix++].token_type = TOKEN_OPEN_ACTION;
+					continue;
+				}
+				if (memcmp(p, "})", 2) == 0) {
+					p += 2;
+					tokens[tix].size = 2;
+					tokens[tix].content = "})";
+					tokens[tix++].token_type = TOKEN_CLOSE_ACTION;
+					continue;
+				}
+			}
+
+			if (nl - p >= 1) {
+				tokens[tix].size = 1;
+				switch (*p) {
+				case '{':
+					p += 1;
+					tokens[tix].content = "{";
+					tokens[tix++].token_type = TOKEN_OPEN_CURLY;
+					continue;
+				case '}':
+					p += 1;
+					tokens[tix].content = "}";
+					tokens[tix++].token_type = TOKEN_CLOSE_CURLY;
+					continue;
+				case '[':
+					p += 1;
+					tokens[tix].content = "[";
+					tokens[tix++].token_type = TOKEN_OPEN_SQUARE;
+					continue;
+				case ']':
+					p += 1;
+					tokens[tix].content = "]";
+					tokens[tix++].token_type = TOKEN_CLOSE_SQUARE;
+					continue;
+				case ',':
+					p += 1;
+					tokens[tix].content = ",";
+					tokens[tix++].token_type = TOKEN_COMMA;
+					continue;
+				default:
+					break;
+				}
+			}
+
+			fprintf(stderr, "%s:%u: Unknown character in grammar: '%c'\n",
+				filename, lineno, *p);
+			exit(1);
+		}
+	}
+
+	nr_tokens = tix;
+	verbose("Extracted %u tokens\n", nr_tokens);
+
+#if 0
+	{
+		int n;
+		for (n = 0; n < nr_tokens; n++)
+			debug("Token %3u: '%s'\n", n, token_list[n].content);
+	}
+#endif
+}
+
+static void build_type_list(void);
+static void parse(void);
+static void dump_elements(void);
+static void render(FILE *out, FILE *hdr);
+
+/*
+ *
+ */
+int main(int argc, char **argv)
+{
+	struct stat st;
+	ssize_t readlen;
+	FILE *out, *hdr;
+	char *buffer, *p;
+	char *kbuild_verbose;
+	int fd;
+
+	kbuild_verbose = getenv("KBUILD_VERBOSE");
+	if (kbuild_verbose)
+		verbose_opt = atoi(kbuild_verbose);
+
+	while (argc > 4) {
+		if (strcmp(argv[1], "-v") == 0)
+			verbose_opt = true;
+		else if (strcmp(argv[1], "-d") == 0)
+			debug_opt = true;
+		else
+			break;
+		memmove(&argv[1], &argv[2], (argc - 2) * sizeof(char *));
+		argc--;
+	}
+
+	if (argc != 4) {
+		fprintf(stderr, "Format: %s [-v] [-d] <grammar-file> <c-file> <hdr-file>\n",
+			argv[0]);
+		exit(2);
+	}
+
+	filename = argv[1];
+	outputname = argv[2];
+	headername = argv[3];
+
+	fd = open(filename, O_RDONLY);
+	if (fd < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (fstat(fd, &st) < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (!(buffer = malloc(st.st_size + 1))) {
+		perror(NULL);
+		exit(1);
+	}
+
+	if ((readlen = read(fd, buffer, st.st_size)) < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (close(fd) < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (readlen != st.st_size) {
+		fprintf(stderr, "%s: Short read\n", filename);
+		exit(1);
+	}
+
+	p = strrchr(argv[1], '/');
+	p = p ? p + 1 : argv[1];
+	grammar_name = strdup(p);
+	if (!p) {
+		perror(NULL);
+		exit(1);
+	}
+	p = strchr(grammar_name, '.');
+	if (p)
+		*p = '\0';
+
+	buffer[readlen] = 0;
+	tokenise(buffer, buffer + readlen);
+	build_type_list();
+	parse();
+	dump_elements();
+
+	out = fopen(outputname, "w");
+	if (!out) {
+		perror(outputname);
+		exit(1);
+	}
+
+	hdr = fopen(headername, "w");
+	if (!hdr) {
+		perror(headername);
+		exit(1);
+	}
+
+	render(out, hdr);
+
+	if (fclose(out) < 0) {
+		perror(outputname);
+		exit(1);
+	}
+
+	if (fclose(hdr) < 0) {
+		perror(headername);
+		exit(1);
+	}
+
+	return 0;
+}
+
+enum compound {
+	NOT_COMPOUND,
+	SET,
+	SET_OF,
+	SEQUENCE,
+	SEQUENCE_OF,
+	CHOICE,
+	ANY,
+	TYPE_REF,
+	TAG_OVERRIDE
+};
+
+struct element {
+	struct type	*type_def;
+	struct token	*name;
+	struct token	*type;
+	struct action	*action;
+	struct element	*children;
+	struct element	*next;
+	struct element	*render_next;
+	struct element	*list_next;
+	uint8_t		n_elements;
+	enum compound	compound : 8;
+	enum asn1_class	class : 8;
+	enum asn1_method method : 8;
+	uint8_t		tag;
+	unsigned	entry_index;
+	unsigned	flags;
+#define ELEMENT_IMPLICIT	0x0001
+#define ELEMENT_EXPLICIT	0x0002
+#define ELEMENT_TAG_SPECIFIED	0x0004
+#define ELEMENT_RENDERED	0x0008
+#define ELEMENT_SKIPPABLE	0x0010
+#define ELEMENT_CONDITIONAL	0x0020
+};
+
+struct type {
+	struct token	*name;
+	struct token	*def;
+	struct element	*element;
+	unsigned	ref_count;
+	unsigned	flags;
+#define TYPE_STOP_MARKER	0x0001
+#define TYPE_BEGIN		0x0002
+};
+
+static struct type *type_list;
+static struct type **type_index;
+static unsigned nr_types;
+
+static int type_index_compare(const void *_a, const void *_b)
+{
+	const struct type *const *a = _a, *const *b = _b;
+
+	if ((*a)->name->size != (*b)->name->size)
+		return (*a)->name->size - (*b)->name->size;
+	else
+		return memcmp((*a)->name->content, (*b)->name->content,
+			      (*a)->name->size);
+}
+
+static int type_finder(const void *_key, const void *_ti)
+{
+	const struct token *token = _key;
+	const struct type *const *ti = _ti;
+	const struct type *type = *ti;
+
+	if (token->size != type->name->size)
+		return token->size - type->name->size;
+	else
+		return memcmp(token->content, type->name->content,
+			      token->size);
+}
+
+/*
+ * Build up a list of types and a sorted index to that list.
+ */
+static void build_type_list(void)
+{
+	struct type *types;
+	unsigned nr, t, n;
+
+	nr = 0;
+	for (n = 0; n < nr_tokens - 1; n++)
+		if (token_list[n + 0].token_type == TOKEN_TYPE_NAME &&
+		    token_list[n + 1].token_type == TOKEN_ASSIGNMENT)
+			nr++;
+
+	if (nr == 0) {
+		fprintf(stderr, "%s: No defined types\n", filename);
+		exit(1);
+	}
+
+	nr_types = nr;
+	types = type_list = calloc(nr + 1, sizeof(type_list[0]));
+	if (!type_list) {
+		perror(NULL);
+		exit(1);
+	}
+	type_index = calloc(nr, sizeof(type_index[0]));
+	if (!type_index) {
+		perror(NULL);
+		exit(1);
+	}
+
+	t = 0;
+	types[t].flags |= TYPE_BEGIN;
+	for (n = 0; n < nr_tokens - 1; n++) {
+		if (token_list[n + 0].token_type == TOKEN_TYPE_NAME &&
+		    token_list[n + 1].token_type == TOKEN_ASSIGNMENT) {
+			types[t].name = &token_list[n];
+			type_index[t] = &types[t];
+			t++;
+		}
+	}
+	types[t].name = &token_list[n + 1];
+	types[t].flags |= TYPE_STOP_MARKER;
+
+	qsort(type_index, nr, sizeof(type_index[0]), type_index_compare);
+
+	verbose("Extracted %u types\n", nr_types);
+#if 0
+	for (n = 0; n < nr_types; n++) {
+		struct type *type = type_index[n];
+		debug("- %*.*s\n", type->name->content);
+	}
+#endif
+}
+
+static struct element *parse_type(struct token **_cursor, struct token *stop,
+				  struct token *name);
+
+/*
+ * Parse the token stream
+ */
+static void parse(void)
+{
+	struct token *cursor;
+	struct type *type;
+
+	/* Parse one type definition statement at a time */
+	type = type_list;
+	do {
+		cursor = type->name;
+
+		if (cursor[0].token_type != TOKEN_TYPE_NAME ||
+		    cursor[1].token_type != TOKEN_ASSIGNMENT)
+			abort();
+		cursor += 2;
+
+		type->element = parse_type(&cursor, type[1].name, NULL);
+		type->element->type_def = type;
+
+		if (cursor != type[1].name) {
+			fprintf(stderr, "%s:%d: Parse error at token '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+	} while (type++, !(type->flags & TYPE_STOP_MARKER));
+
+	verbose("Extracted %u actions\n", nr_actions);
+}
+
+static struct element *element_list;
+
+static struct element *alloc_elem(struct token *type)
+{
+	struct element *e = calloc(1, sizeof(*e));
+	if (!e) {
+		perror(NULL);
+		exit(1);
+	}
+	e->list_next = element_list;
+	element_list = e;
+	return e;
+}
+
+static struct element *parse_compound(struct token **_cursor, struct token *end,
+				      int alternates);
+
+/*
+ * Parse one type definition statement
+ */
+static struct element *parse_type(struct token **_cursor, struct token *end,
+				  struct token *name)
+{
+	struct element *top, *element;
+	struct action *action, **ppaction;
+	struct token *cursor = *_cursor;
+	struct type **ref;
+	char *p;
+	int labelled = 0, implicit = 0;
+
+	top = element = alloc_elem(cursor);
+	element->class = ASN1_UNIV;
+	element->method = ASN1_PRIM;
+	element->tag = token_to_tag[cursor->token_type];
+	element->name = name;
+
+	/* Extract the tag value if one given */
+	if (cursor->token_type == TOKEN_OPEN_SQUARE) {
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		switch (cursor->token_type) {
+		case DIRECTIVE_UNIVERSAL:
+			element->class = ASN1_UNIV;
+			cursor++;
+			break;
+		case DIRECTIVE_APPLICATION:
+			element->class = ASN1_APPL;
+			cursor++;
+			break;
+		case TOKEN_NUMBER:
+			element->class = ASN1_CONT;
+			break;
+		case DIRECTIVE_PRIVATE:
+			element->class = ASN1_PRIV;
+			cursor++;
+			break;
+		default:
+			fprintf(stderr, "%s:%d: Unrecognised tag class token '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_NUMBER) {
+			fprintf(stderr, "%s:%d: Missing tag number '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+		element->tag &= ~0x1f;
+		element->tag |= strtoul(cursor->content, &p, 10);
+		element->flags |= ELEMENT_TAG_SPECIFIED;
+		if (p - cursor->content != cursor->size)
+			abort();
+		cursor++;
+
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_CLOSE_SQUARE) {
+			fprintf(stderr, "%s:%d: Missing closing square bracket '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		labelled = 1;
+	}
+
+	/* Handle implicit and explicit markers */
+	if (cursor->token_type == DIRECTIVE_IMPLICIT) {
+		element->flags |= ELEMENT_IMPLICIT;
+		implicit = 1;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+	} else if (cursor->token_type == DIRECTIVE_EXPLICIT) {
+		element->flags |= ELEMENT_EXPLICIT;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+	}
+
+	if (labelled) {
+		if (!implicit)
+			element->method |= ASN1_CONS;
+		element->compound = implicit ? TAG_OVERRIDE : SEQUENCE;
+		element->children = alloc_elem(cursor);
+		element = element->children;
+		element->class = ASN1_UNIV;
+		element->method = ASN1_PRIM;
+		element->tag = token_to_tag[cursor->token_type];
+		element->name = name;
+	}
+
+	/* Extract the type we're expecting here */
+	element->type = cursor;
+	switch (cursor->token_type) {
+	case DIRECTIVE_ANY:
+		element->compound = ANY;
+		cursor++;
+		break;
+
+	case DIRECTIVE_NULL:
+	case DIRECTIVE_BOOLEAN:
+	case DIRECTIVE_ENUMERATED:
+	case DIRECTIVE_INTEGER:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		break;
+
+	case DIRECTIVE_EXTERNAL:
+		element->method = ASN1_CONS;
+
+	case DIRECTIVE_BMPString:
+	case DIRECTIVE_GeneralString:
+	case DIRECTIVE_GraphicString:
+	case DIRECTIVE_IA5String:
+	case DIRECTIVE_ISO646String:
+	case DIRECTIVE_NumericString:
+	case DIRECTIVE_PrintableString:
+	case DIRECTIVE_T61String:
+	case DIRECTIVE_TeletexString:
+	case DIRECTIVE_UniversalString:
+	case DIRECTIVE_UTF8String:
+	case DIRECTIVE_VideotexString:
+	case DIRECTIVE_VisibleString:
+	case DIRECTIVE_ObjectDescriptor:
+	case DIRECTIVE_GeneralizedTime:
+	case DIRECTIVE_UTCTime:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		break;
+
+	case DIRECTIVE_BIT:
+	case DIRECTIVE_OCTET:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != DIRECTIVE_STRING)
+			goto parse_error;
+		cursor++;
+		break;
+
+	case DIRECTIVE_OBJECT:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != DIRECTIVE_IDENTIFIER)
+			goto parse_error;
+		cursor++;
+		break;
+
+	case TOKEN_TYPE_NAME:
+		element->compound = TYPE_REF;
+		ref = bsearch(cursor, type_index, nr_types, sizeof(type_index[0]),
+			      type_finder);
+		if (!ref) {
+			fprintf(stderr, "%s:%d: Type '%s' undefined\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+		cursor->type = *ref;
+		(*ref)->ref_count++;
+		cursor++;
+		break;
+
+	case DIRECTIVE_CHOICE:
+		element->compound = CHOICE;
+		cursor++;
+		element->children = parse_compound(&cursor, end, 1);
+		break;
+
+	case DIRECTIVE_SEQUENCE:
+		element->compound = SEQUENCE;
+		element->method = ASN1_CONS;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type == DIRECTIVE_OF) {
+			element->compound = SEQUENCE_OF;
+			cursor++;
+			if (cursor >= end)
+				goto overrun_error;
+			element->children = parse_type(&cursor, end, NULL);
+		} else {
+			element->children = parse_compound(&cursor, end, 0);
+		}
+		break;
+
+	case DIRECTIVE_SET:
+		element->compound = SET;
+		element->method = ASN1_CONS;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type == DIRECTIVE_OF) {
+			element->compound = SET_OF;
+			cursor++;
+			if (cursor >= end)
+				goto parse_error;
+			element->children = parse_type(&cursor, end, NULL);
+		} else {
+			element->children = parse_compound(&cursor, end, 1);
+		}
+		break;
+
+	default:
+		fprintf(stderr, "%s:%d: Token '%s' does not introduce a type\n",
+			filename, cursor->line, cursor->content);
+		exit(1);
+	}
+
+	/* Handle elements that are optional */
+	if (cursor < end && (cursor->token_type == DIRECTIVE_OPTIONAL ||
+			     cursor->token_type == DIRECTIVE_DEFAULT)
+	    ) {
+		cursor++;
+		top->flags |= ELEMENT_SKIPPABLE;
+	}
+
+	if (cursor < end && cursor->token_type == TOKEN_OPEN_ACTION) {
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_ELEMENT_NAME) {
+			fprintf(stderr, "%s:%d: Token '%s' is not an action function name\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+		action = malloc(sizeof(struct action));
+		if (!action) {
+			perror(NULL);
+			exit(1);
+		}
+		action->index = 0;
+		action->name = cursor->content;
+
+		for (ppaction = &action_list;
+		     *ppaction;
+		     ppaction = &(*ppaction)->next
+		     ) {
+			int cmp = strcmp(action->name, (*ppaction)->name);
+			if (cmp == 0) {
+				free(action);
+				action = *ppaction;
+				goto found;
+			}
+			if (cmp < 0) {
+				action->next = *ppaction;
+				*ppaction = action;
+				nr_actions++;
+				goto found;
+			}
+		}
+		action->next = NULL;
+		*ppaction = action;
+		nr_actions++;
+	found:
+
+		element->action = action;
+		cursor->action = action;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_CLOSE_ACTION) {
+			fprintf(stderr, "%s:%d: Missing close action, got '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+		cursor++;
+	}
+
+	*_cursor = cursor;
+	return top;
+
+parse_error:
+	fprintf(stderr, "%s:%d: Unexpected token '%s'\n",
+		filename, cursor->line, cursor->content);
+	exit(1);
+
+overrun_error:
+	fprintf(stderr, "%s: Unexpectedly hit EOF\n", filename);
+	exit(1);
+}
+
+/*
+ * Parse a compound type list
+ */
+static struct element *parse_compound(struct token **_cursor, struct token *end,
+				      int alternates)
+{
+	struct element *children, **child_p = &children, *element;
+	struct token *cursor = *_cursor, *name;
+
+	if (cursor->token_type != TOKEN_OPEN_CURLY) {
+		fprintf(stderr, "%s:%d: Expected compound to start with brace not '%s'\n",
+			filename, cursor->line, cursor->content);
+		exit(1);
+	}
+	cursor++;
+	if (cursor >= end)
+		goto overrun_error;
+
+	if (cursor->token_type == TOKEN_OPEN_CURLY) {
+		fprintf(stderr, "%s:%d: Empty compound\n",
+			filename, cursor->line);
+		exit(1);
+	}
+
+	for (;;) {
+		name = NULL;
+		if (cursor->token_type == TOKEN_ELEMENT_NAME) {
+			name = cursor;
+			cursor++;
+			if (cursor >= end)
+				goto overrun_error;
+		}
+
+		element = parse_type(&cursor, end, name);
+		if (alternates)
+			element->flags |= ELEMENT_SKIPPABLE | ELEMENT_CONDITIONAL;
+
+		*child_p = element;
+		child_p = &element->next;
+
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_COMMA)
+			break;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+	}
+
+	children->flags &= ~ELEMENT_CONDITIONAL;
+
+	if (cursor->token_type != TOKEN_CLOSE_CURLY) {
+		fprintf(stderr, "%s:%d: Expected compound closure, got '%s'\n",
+			filename, cursor->line, cursor->content);
+		exit(1);
+	}
+	cursor++;
+
+	*_cursor = cursor;
+	return children;
+
+overrun_error:
+	fprintf(stderr, "%s: Unexpectedly hit EOF\n", filename);
+	exit(1);
+}
+
+static void dump_element(const struct element *e, int level)
+{
+	const struct element *c;
+	const struct type *t = e->type_def;
+	const char *name = e->name ? e->name->content : ".";
+	const char *tname = t && t->name ? t->name->content : ".";
+	char tag[32];
+
+	if (e->class == 0 && e->method == 0 && e->tag == 0)
+		strcpy(tag, "<...>");
+	else if (e->class == ASN1_UNIV)
+		sprintf(tag, "%s %s %s",
+			asn1_classes[e->class],
+			asn1_methods[e->method],
+			asn1_universal_tags[e->tag]);
+	else
+		sprintf(tag, "%s %s %u",
+			asn1_classes[e->class],
+			asn1_methods[e->method],
+			e->tag);
+
+	printf("%c%c%c%c%c %c %*s[*] \e[33m%s\e[m %s %s \e[35m%s\e[m\n",
+	       e->flags & ELEMENT_IMPLICIT ? 'I' : '-',
+	       e->flags & ELEMENT_EXPLICIT ? 'E' : '-',
+	       e->flags & ELEMENT_TAG_SPECIFIED ? 'T' : '-',
+	       e->flags & ELEMENT_SKIPPABLE ? 'S' : '-',
+	       e->flags & ELEMENT_CONDITIONAL ? 'C' : '-',
+	       "-tTqQcaro"[e->compound],
+	       level, "",
+	       tag,
+	       tname,
+	       name,
+	       e->action ? e->action->name : "");
+	if (e->compound == TYPE_REF)
+		dump_element(e->type->type->element, level + 3);
+	else
+		for (c = e->children; c; c = c->next)
+			dump_element(c, level + 3);
+}
+
+static void dump_elements(void)
+{
+	if (debug_opt)
+		dump_element(type_list[0].element, 0);
+}
+
+static void render_element(FILE *out, struct element *e, struct element *tag);
+static void render_out_of_line_list(FILE *out);
+
+static int nr_entries;
+static int render_depth = 1;
+static struct element *render_list, **render_list_p = &render_list;
+
+__attribute__((format(printf, 2, 3)))
+static void render_opcode(FILE *out, const char *fmt, ...)
+{
+	va_list va;
+
+	if (out) {
+		fprintf(out, "\t[%4d] =%*s", nr_entries, render_depth, "");
+		va_start(va, fmt);
+		vfprintf(out, fmt, va);
+		va_end(va);
+	}
+	nr_entries++;
+}
+
+__attribute__((format(printf, 2, 3)))
+static void render_more(FILE *out, const char *fmt, ...)
+{
+	va_list va;
+
+	if (out) {
+		va_start(va, fmt);
+		vfprintf(out, fmt, va);
+		va_end(va);
+	}
+}
+
+/*
+ * Render the grammar into a state machine definition.
+ */
+static void render(FILE *out, FILE *hdr)
+{
+	struct element *e;
+	struct action *action;
+	struct type *root;
+	int index;
+
+	fprintf(hdr, "/*\n");
+	fprintf(hdr, " * Automatically generated by asn1_compiler.  Do not edit\n");
+	fprintf(hdr, " *\n");
+	fprintf(hdr, " * ASN.1 parser for %s\n", grammar_name);
+	fprintf(hdr, " */\n");
+	fprintf(hdr, "#include <linux/asn1_decoder.h>\n");
+	fprintf(hdr, "\n");
+	fprintf(hdr, "extern const struct asn1_decoder %s_decoder;\n", grammar_name);
+	if (ferror(hdr)) {
+		perror(headername);
+		exit(1);
+	}
+
+	fprintf(out, "/*\n");
+	fprintf(out, " * Automatically generated by asn1_compiler.  Do not edit\n");
+	fprintf(out, " *\n");
+	fprintf(out, " * ASN.1 parser for %s\n", grammar_name);
+	fprintf(out, " */\n");
+	fprintf(out, "#include <linux/asn1_ber_bytecode.h>\n");
+	fprintf(out, "#include \"%s.asn1.h\"\n", grammar_name);
+	fprintf(out, "\n");
+	if (ferror(out)) {
+		perror(outputname);
+		exit(1);
+	}
+
+	/* Tabulate the action functions we might have to call */
+	fprintf(hdr, "\n");
+	index = 0;
+	for (action = action_list; action; action = action->next) {
+		action->index = index++;
+		fprintf(hdr,
+			"extern int %s(void *, size_t, unsigned char,"
+			" const void *, size_t);\n",
+			action->name);
+	}
+	fprintf(hdr, "\n");
+
+	fprintf(out, "enum %s_actions {\n", grammar_name);
+	for (action = action_list; action; action = action->next)
+		fprintf(out, "\tACT_%s = %u,\n",
+			action->name, action->index);
+	fprintf(out, "\tNR__%s_actions = %u\n", grammar_name, nr_actions);
+	fprintf(out, "};\n");
+
+	fprintf(out, "\n");
+	fprintf(out, "static const asn1_action_t %s_action_table[NR__%s_actions] = {\n",
+		grammar_name, grammar_name);
+	for (action = action_list; action; action = action->next)
+		fprintf(out, "\t[%4u] = %s,\n", action->index, action->name);
+	fprintf(out, "};\n");
+
+	if (ferror(out)) {
+		perror(outputname);
+		exit(1);
+	}
+
+	/* We do two passes - the first one calculates all the offsets */
+	verbose("Pass 1\n");
+	nr_entries = 0;
+	root = &type_list[0];
+	render_element(NULL, root->element, NULL);
+	render_opcode(NULL, "ASN1_OP_COMPLETE,\n");
+	render_out_of_line_list(NULL);
+
+	for (e = element_list; e; e = e->list_next)
+		e->flags &= ~ELEMENT_RENDERED;
+
+	/* And then we actually render */
+	verbose("Pass 2\n");
+	fprintf(out, "\n");
+	fprintf(out, "static const unsigned char %s_machine[] = {\n",
+		grammar_name);
+
+	nr_entries = 0;
+	root = &type_list[0];
+	render_element(out, root->element, NULL);
+	render_opcode(out, "ASN1_OP_COMPLETE,\n");
+	render_out_of_line_list(out);
+
+	fprintf(out, "};\n");
+
+	fprintf(out, "\n");
+	fprintf(out, "const struct asn1_decoder %s_decoder = {\n", grammar_name);
+	fprintf(out, "\t.machine = %s_machine,\n", grammar_name);
+	fprintf(out, "\t.machlen = sizeof(%s_machine),\n", grammar_name);
+	fprintf(out, "\t.actions = %s_action_table,\n", grammar_name);
+	fprintf(out, "};\n");
+}
+
+/*
+ * Render the out-of-line elements
+ */
+static void render_out_of_line_list(FILE *out)
+{
+	struct element *e, *ce;
+	const char *act;
+	int entry;
+
+	while ((e = render_list)) {
+		render_list = e->render_next;
+		if (!render_list)
+			render_list_p = &render_list;
+
+		render_more(out, "\n");
+		e->entry_index = entry = nr_entries;
+		render_depth++;
+		for (ce = e->children; ce; ce = ce->next)
+			render_element(out, ce, NULL);
+		render_depth--;
+
+		act = e->action ? "_ACT" : "";
+		switch (e->compound) {
+		case SEQUENCE:
+			render_opcode(out, "ASN1_OP_END_SEQ%s,\n", act);
+			break;
+		case SEQUENCE_OF:
+			render_opcode(out, "ASN1_OP_END_SEQ_OF%s,\n", act);
+			render_opcode(out, "_jump_target(%u),\n", entry);
+			break;
+		case SET:
+			render_opcode(out, "ASN1_OP_END_SET%s,\n", act);
+			break;
+		case SET_OF:
+			render_opcode(out, "ASN1_OP_END_SET_OF%s,\n", act);
+			render_opcode(out, "_jump_target(%u),\n", entry);
+			break;
+		default:
+			break;
+		}
+		if (e->action)
+			render_opcode(out, "_action(ACT_%s),\n",
+				      e->action->name);
+		render_opcode(out, "ASN1_OP_RETURN,\n");
+	}
+}
+
+/*
+ * Render an element.
+ */
+static void render_element(FILE *out, struct element *e, struct element *tag)
+{
+	struct element *ec, *x;
+	const char *cond, *act;
+	int entry, skippable = 0, outofline = 0;
+
+	if (e->flags & ELEMENT_SKIPPABLE ||
+	    (tag && tag->flags & ELEMENT_SKIPPABLE))
+		skippable = 1;
+
+	if ((e->type_def && e->type_def->ref_count > 1) ||
+	    skippable)
+		outofline = 1;
+
+	if (e->type_def && out) {
+		render_more(out, "\t// %s\n", e->type_def->name->content);
+	}
+
+	/* Render the operation */
+	cond = (e->flags & ELEMENT_CONDITIONAL ||
+		(tag && tag->flags & ELEMENT_CONDITIONAL)) ? "COND_" : "";
+	act = e->action ? "_ACT" : "";
+	switch (e->compound) {
+	case ANY:
+		render_opcode(out, "ASN1_OP_%sMATCH_ANY%s%s,",
+			      cond, act, skippable ? "_OR_SKIP" : "");
+		if (e->name)
+			render_more(out, "\t\t// %s", e->name->content);
+		render_more(out, "\n");
+		goto dont_render_tag;
+
+	case TAG_OVERRIDE:
+		render_element(out, e->children, e);
+		return;
+
+	case SEQUENCE:
+	case SEQUENCE_OF:
+	case SET:
+	case SET_OF:
+		render_opcode(out, "ASN1_OP_%sMATCH%s%s,",
+			      cond,
+			      outofline ? "_JUMP" : "",
+			      skippable ? "_OR_SKIP" : "");
+		break;
+
+	case CHOICE:
+		goto dont_render_tag;
+
+	case TYPE_REF:
+		if (e->class == ASN1_UNIV && e->method == ASN1_PRIM && e->tag == 0)
+			goto dont_render_tag;
+	default:
+		render_opcode(out, "ASN1_OP_%sMATCH%s%s,",
+			      cond, act,
+			      skippable ? "_OR_SKIP" : "");
+		break;
+	}
+
+	x = tag ?: e;
+	if (x->name)
+		render_more(out, "\t\t// %s", x->name->content);
+	render_more(out, "\n");
+
+	/* Render the tag */
+	if (!tag || !(tag->flags & ELEMENT_TAG_SPECIFIED))
+		tag = e;
+
+	if (tag->class == ASN1_UNIV &&
+	    tag->tag != 14 &&
+	    tag->tag != 15 &&
+	    tag->tag != 31)
+		render_opcode(out, "_tag(%s, %s, %s),\n",
+			      asn1_classes[tag->class],
+			      asn1_methods[tag->method | e->method],
+			      asn1_universal_tags[tag->tag]);
+	else
+		render_opcode(out, "_tagn(%s, %s, %2u),\n",
+			      asn1_classes[tag->class],
+			      asn1_methods[tag->method | e->method],
+			      tag->tag);
+	tag = NULL;
+dont_render_tag:
+
+	/* Deal with compound types */
+	switch (e->compound) {
+	case TYPE_REF:
+		render_element(out, e->type->type->element, tag);
+		if (e->action)
+			render_opcode(out, "ASN1_OP_%sACT,\n",
+				      skippable ? "MAYBE_" : "");
+		break;
+
+	case SEQUENCE:
+		if (outofline) {
+			/* Render out-of-line for multiple use or
+			 * skipability */
+			render_opcode(out, "_jump_target(%u),", e->entry_index);
+			if (e->type_def && e->type_def->name)
+				render_more(out, "\t\t// --> %s",
+					    e->type_def->name->content);
+			render_more(out, "\n");
+			if (!(e->flags & ELEMENT_RENDERED)) {
+				e->flags |= ELEMENT_RENDERED;
+				*render_list_p = e;
+				render_list_p = &e->render_next;
+			}
+			return;
+		} else {
+			/* Render inline for single use */
+			render_depth++;
+			for (ec = e->children; ec; ec = ec->next)
+				render_element(out, ec, NULL);
+			render_depth--;
+			render_opcode(out, "ASN1_OP_END_SEQ%s,\n", act);
+		}
+		break;
+
+	case SEQUENCE_OF:
+	case SET_OF:
+		if (outofline) {
+			/* Render out-of-line for multiple use or
+			 * skipability */
+			render_opcode(out, "_jump_target(%u),", e->entry_index);
+			if (e->type_def && e->type_def->name)
+				render_more(out, "\t\t// --> %s",
+					    e->type_def->name->content);
+			render_more(out, "\n");
+			if (!(e->flags & ELEMENT_RENDERED)) {
+				e->flags |= ELEMENT_RENDERED;
+				*render_list_p = e;
+				render_list_p = &e->render_next;
+			}
+			return;
+		} else {
+			/* Render inline for single use */
+			entry = nr_entries;
+			render_depth++;
+			render_element(out, e->children, NULL);
+			render_depth--;
+			if (e->compound == SEQUENCE_OF)
+				render_opcode(out, "ASN1_OP_END_SEQ_OF%s,\n", act);
+			else
+				render_opcode(out, "ASN1_OP_END_SET_OF%s,\n", act);
+			render_opcode(out, "_jump_target(%u),\n", entry);
+		}
+		break;
+
+	case SET:
+		/* I can't think of a nice way to do SET support without having
+		 * a stack of bitmasks to make sure no element is repeated.
+		 * The bitmask has also to be checked that no non-optional
+		 * elements are left out whilst not preventing optional
+		 * elements from being left out.
+		 */
+		fprintf(stderr, "The ASN.1 SET type is not currently supported.\n");
+		exit(1);
+
+	case CHOICE:
+		for (ec = e->children; ec; ec = ec->next)
+			render_element(out, ec, ec);
+		if (!skippable)
+			render_opcode(out, "ASN1_OP_COND_FAIL,\n");
+		if (e->action)
+			render_opcode(out, "ASN1_OP_ACT,\n");
+		break;
+
+	default:
+		break;
+	}
+
+	if (e->action)
+		render_opcode(out, "_action(ACT_%s),\n", e->action->name);
+}
#14 lore raw

PATCH 4/11 cmd: add asn1_compiler

Quoted 237 lines · Heinrich Schuchardt
Quoted 5 lines · AKASHI Takahiro

Imported from linux kernel v5.3.

This host command will be used to create a ASN1 parser, for example, for pkcs7 or x509.

Quoted 8 lines
---
 cmd/Kconfig                       |    3 +
 include/linux/asn1.h              |   65 ++
 include/linux/asn1_ber_bytecode.h |   89 ++
 include/linux/asn1_decoder.h      |   20 +
 scripts/Makefile                  |    3 +
 scripts/asn1_compiler.c           | 1611 +++++++++++++++++++++++++++++

Except for Makefile and Kconfig these are verbatime copies of the Linux files. OK.

Quoted 219 lines
 6 files changed, 1791 insertions(+)
 create mode 100644 include/linux/asn1.h
 create mode 100644 include/linux/asn1_ber_bytecode.h
 create mode 100644 include/linux/asn1_decoder.h
 create mode 100644 scripts/asn1_compiler.c
cmd/Kconfig +3 -0
index 4e61565aab16..605dcfdf4282 100644
@@ -184,6 +184,9 @@ endmenu
 config BUILD_BIN2C
 	bool
 
+config BUILD_ASN1
+	bool
+
 comment "Commands"
 
 menu "Info commands"
include/linux/asn1.h new file +65 -0
new file mode 100644
index 000000000000..a4d0bdd10711
@@ -0,0 +1,65 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 BER/DER/CER encoding definitions
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_ASN1_H
+#define _LINUX_ASN1_H
+
+/* Class */
+enum asn1_class {
+	ASN1_UNIV	= 0,	/* Universal */
+	ASN1_APPL	= 1,	/* Application */
+	ASN1_CONT	= 2,	/* Context */
+	ASN1_PRIV	= 3	/* Private */
+};
+#define ASN1_CLASS_BITS	0xc0
+
+
+enum asn1_method {
+	ASN1_PRIM	= 0,	/* Primitive */
+	ASN1_CONS	= 1	/* Constructed */
+};
+#define ASN1_CONS_BIT	0x20
+
+/* Tag */
+enum asn1_tag {
+	ASN1_EOC	= 0,	/* End Of Contents or N/A */
+	ASN1_BOOL	= 1,	/* Boolean */
+	ASN1_INT	= 2,	/* Integer */
+	ASN1_BTS	= 3,	/* Bit String */
+	ASN1_OTS	= 4,	/* Octet String */
+	ASN1_NULL	= 5,	/* Null */
+	ASN1_OID	= 6,	/* Object Identifier  */
+	ASN1_ODE	= 7,	/* Object Description */
+	ASN1_EXT	= 8,	/* External */
+	ASN1_REAL	= 9,	/* Real float */
+	ASN1_ENUM	= 10,	/* Enumerated */
+	ASN1_EPDV	= 11,	/* Embedded PDV */
+	ASN1_UTF8STR	= 12,	/* UTF8 String */
+	ASN1_RELOID	= 13,	/* Relative OID */
+	/* 14 - Reserved */
+	/* 15 - Reserved */
+	ASN1_SEQ	= 16,	/* Sequence and Sequence of */
+	ASN1_SET	= 17,	/* Set and Set of */
+	ASN1_NUMSTR	= 18,	/* Numerical String */
+	ASN1_PRNSTR	= 19,	/* Printable String */
+	ASN1_TEXSTR	= 20,	/* T61 String / Teletext String */
+	ASN1_VIDSTR	= 21,	/* Videotex String */
+	ASN1_IA5STR	= 22,	/* IA5 String */
+	ASN1_UNITIM	= 23,	/* Universal Time */
+	ASN1_GENTIM	= 24,	/* General Time */
+	ASN1_GRASTR	= 25,	/* Graphic String */
+	ASN1_VISSTR	= 26,	/* Visible String */
+	ASN1_GENSTR	= 27,	/* General String */
+	ASN1_UNISTR	= 28,	/* Universal String */
+	ASN1_CHRSTR	= 29,	/* Character String */
+	ASN1_BMPSTR	= 30,	/* BMP String */
+	ASN1_LONG_TAG	= 31	/* Long form tag */
+};
+
+#define ASN1_INDEFINITE_LENGTH 0x80
+
+#endif /* _LINUX_ASN1_H */
include/linux/asn1_ber_bytecode.h new file +89 -0
new file mode 100644
index 000000000000..b38361953a48
@@ -0,0 +1,89 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 BER/DER/CER parsing state machine internal definitions
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_ASN1_BER_BYTECODE_H
+#define _LINUX_ASN1_BER_BYTECODE_H
+
+#ifdef __KERNEL__
+#include <linux/types.h>
+#endif
+#include <linux/asn1.h>
+
+typedef int (*asn1_action_t)(void *context,
+			     size_t hdrlen, /* In case of ANY type */
+			     unsigned char tag, /* In case of ANY type */
+			     const void *value, size_t vlen);
+
+struct asn1_decoder {
+	const unsigned char *machine;
+	size_t machlen;
+	const asn1_action_t *actions;
+};
+
+enum asn1_opcode {
+	/* The tag-matching ops come first and the odd-numbered slots
+	 * are for OR_SKIP ops.
+	 */
+#define ASN1_OP_MATCH__SKIP		  0x01
+#define ASN1_OP_MATCH__ACT		  0x02
+#define ASN1_OP_MATCH__JUMP		  0x04
+#define ASN1_OP_MATCH__ANY		  0x08
+#define ASN1_OP_MATCH__COND		  0x10
+
+	ASN1_OP_MATCH			= 0x00,
+	ASN1_OP_MATCH_OR_SKIP		= 0x01,
+	ASN1_OP_MATCH_ACT		= 0x02,
+	ASN1_OP_MATCH_ACT_OR_SKIP	= 0x03,
+	ASN1_OP_MATCH_JUMP		= 0x04,
+	ASN1_OP_MATCH_JUMP_OR_SKIP	= 0x05,
+	ASN1_OP_MATCH_ANY		= 0x08,
+	ASN1_OP_MATCH_ANY_OR_SKIP	= 0x09,
+	ASN1_OP_MATCH_ANY_ACT		= 0x0a,
+	ASN1_OP_MATCH_ANY_ACT_OR_SKIP	= 0x0b,
+	/* Everything before here matches unconditionally */
+
+	ASN1_OP_COND_MATCH_OR_SKIP	= 0x11,
+	ASN1_OP_COND_MATCH_ACT_OR_SKIP	= 0x13,
+	ASN1_OP_COND_MATCH_JUMP_OR_SKIP	= 0x15,
+	ASN1_OP_COND_MATCH_ANY		= 0x18,
+	ASN1_OP_COND_MATCH_ANY_OR_SKIP	= 0x19,
+	ASN1_OP_COND_MATCH_ANY_ACT	= 0x1a,
+	ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP = 0x1b,
+
+	/* Everything before here will want a tag from the data */
+#define ASN1_OP__MATCHES_TAG ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP
+
+	/* These are here to help fill up space */
+	ASN1_OP_COND_FAIL		= 0x1c,
+	ASN1_OP_COMPLETE		= 0x1d,
+	ASN1_OP_ACT			= 0x1e,
+	ASN1_OP_MAYBE_ACT		= 0x1f,
+
+	/* The following eight have bit 0 -> SET, 1 -> OF, 2 -> ACT */
+	ASN1_OP_END_SEQ			= 0x20,
+	ASN1_OP_END_SET			= 0x21,
+	ASN1_OP_END_SEQ_OF		= 0x22,
+	ASN1_OP_END_SET_OF		= 0x23,
+	ASN1_OP_END_SEQ_ACT		= 0x24,
+	ASN1_OP_END_SET_ACT		= 0x25,
+	ASN1_OP_END_SEQ_OF_ACT		= 0x26,
+	ASN1_OP_END_SET_OF_ACT		= 0x27,
+#define ASN1_OP_END__SET		  0x01
+#define ASN1_OP_END__OF			  0x02
+#define ASN1_OP_END__ACT		  0x04
+
+	ASN1_OP_RETURN			= 0x28,
+
+	ASN1_OP__NR
+};
+
+#define _tag(CLASS, CP, TAG) ((ASN1_##CLASS << 6) | (ASN1_##CP << 5) | ASN1_##TAG)
+#define _tagn(CLASS, CP, TAG) ((ASN1_##CLASS << 6) | (ASN1_##CP << 5) | TAG)
+#define _jump_target(N) (N)
+#define _action(N) (N)
+
+#endif /* _LINUX_ASN1_BER_BYTECODE_H */
include/linux/asn1_decoder.h new file +20 -0
new file mode 100644
index 000000000000..83f9c6e1e5e9
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 decoder
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_ASN1_DECODER_H
+#define _LINUX_ASN1_DECODER_H
+
+#include <linux/asn1.h>
+
+struct asn1_decoder;
+
+extern int asn1_ber_decoder(const struct asn1_decoder *decoder,
+			    void *context,
+			    const unsigned char *data,
+			    size_t datalen);
+
+#endif /* _LINUX_ASN1_DECODER_H */
scripts/Makefile +0 -0
index e7b353f77f43..c4e2ae5db2ea 100644
@@ -5,6 +5,9 @@
 # ---------------------------------------------------------------------------
 
Quoted 6 lines

scripts/Makefile line 10

Quoted 4 lines
scripts/Makefile +3 -0
 hostprogs-$(CONFIG_BUILD_BIN2C)		+= bin2c
+hostprogs-$(CONFIG_BUILD_ASN1)		+= asn1_compiler
+
+HOSTCFLAGS_asn1_compiler.o = -idirafter $(srctree)/include

Which problems did you run into that make you use -idirafter here?

Without this directive, a compiler supersedes *host's* include over U-Boot's include in matching "linux/", then it will miss out include/linux/asn1_ber_bytecode.h.

-Takahiro Akashi

Quoted 1,627 lines

Otherwise

Quoted 1,621 lines
scripts/Makefile +0 -0
 always		:= $(hostprogs-y)
 
scripts/asn1_compiler.c new file +1611 -0
new file mode 100644
index 000000000000..adabd4145264
@@ -0,0 +1,1611 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* Simplified ASN.1 notation parser
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <string.h>
+#include <ctype.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <sys/stat.h>
+#include <linux/asn1_ber_bytecode.h>
+
+enum token_type {
+	DIRECTIVE_ABSENT,
+	DIRECTIVE_ALL,
+	DIRECTIVE_ANY,
+	DIRECTIVE_APPLICATION,
+	DIRECTIVE_AUTOMATIC,
+	DIRECTIVE_BEGIN,
+	DIRECTIVE_BIT,
+	DIRECTIVE_BMPString,
+	DIRECTIVE_BOOLEAN,
+	DIRECTIVE_BY,
+	DIRECTIVE_CHARACTER,
+	DIRECTIVE_CHOICE,
+	DIRECTIVE_CLASS,
+	DIRECTIVE_COMPONENT,
+	DIRECTIVE_COMPONENTS,
+	DIRECTIVE_CONSTRAINED,
+	DIRECTIVE_CONTAINING,
+	DIRECTIVE_DEFAULT,
+	DIRECTIVE_DEFINED,
+	DIRECTIVE_DEFINITIONS,
+	DIRECTIVE_EMBEDDED,
+	DIRECTIVE_ENCODED,
+	DIRECTIVE_ENCODING_CONTROL,
+	DIRECTIVE_END,
+	DIRECTIVE_ENUMERATED,
+	DIRECTIVE_EXCEPT,
+	DIRECTIVE_EXPLICIT,
+	DIRECTIVE_EXPORTS,
+	DIRECTIVE_EXTENSIBILITY,
+	DIRECTIVE_EXTERNAL,
+	DIRECTIVE_FALSE,
+	DIRECTIVE_FROM,
+	DIRECTIVE_GeneralString,
+	DIRECTIVE_GeneralizedTime,
+	DIRECTIVE_GraphicString,
+	DIRECTIVE_IA5String,
+	DIRECTIVE_IDENTIFIER,
+	DIRECTIVE_IMPLICIT,
+	DIRECTIVE_IMPLIED,
+	DIRECTIVE_IMPORTS,
+	DIRECTIVE_INCLUDES,
+	DIRECTIVE_INSTANCE,
+	DIRECTIVE_INSTRUCTIONS,
+	DIRECTIVE_INTEGER,
+	DIRECTIVE_INTERSECTION,
+	DIRECTIVE_ISO646String,
+	DIRECTIVE_MAX,
+	DIRECTIVE_MIN,
+	DIRECTIVE_MINUS_INFINITY,
+	DIRECTIVE_NULL,
+	DIRECTIVE_NumericString,
+	DIRECTIVE_OBJECT,
+	DIRECTIVE_OCTET,
+	DIRECTIVE_OF,
+	DIRECTIVE_OPTIONAL,
+	DIRECTIVE_ObjectDescriptor,
+	DIRECTIVE_PATTERN,
+	DIRECTIVE_PDV,
+	DIRECTIVE_PLUS_INFINITY,
+	DIRECTIVE_PRESENT,
+	DIRECTIVE_PRIVATE,
+	DIRECTIVE_PrintableString,
+	DIRECTIVE_REAL,
+	DIRECTIVE_RELATIVE_OID,
+	DIRECTIVE_SEQUENCE,
+	DIRECTIVE_SET,
+	DIRECTIVE_SIZE,
+	DIRECTIVE_STRING,
+	DIRECTIVE_SYNTAX,
+	DIRECTIVE_T61String,
+	DIRECTIVE_TAGS,
+	DIRECTIVE_TRUE,
+	DIRECTIVE_TeletexString,
+	DIRECTIVE_UNION,
+	DIRECTIVE_UNIQUE,
+	DIRECTIVE_UNIVERSAL,
+	DIRECTIVE_UTCTime,
+	DIRECTIVE_UTF8String,
+	DIRECTIVE_UniversalString,
+	DIRECTIVE_VideotexString,
+	DIRECTIVE_VisibleString,
+	DIRECTIVE_WITH,
+	NR__DIRECTIVES,
+	TOKEN_ASSIGNMENT = NR__DIRECTIVES,
+	TOKEN_OPEN_CURLY,
+	TOKEN_CLOSE_CURLY,
+	TOKEN_OPEN_SQUARE,
+	TOKEN_CLOSE_SQUARE,
+	TOKEN_OPEN_ACTION,
+	TOKEN_CLOSE_ACTION,
+	TOKEN_COMMA,
+	TOKEN_NUMBER,
+	TOKEN_TYPE_NAME,
+	TOKEN_ELEMENT_NAME,
+	NR__TOKENS
+};
+
+static const unsigned char token_to_tag[NR__TOKENS] = {
+	/* EOC goes first */
+	[DIRECTIVE_BOOLEAN]		= ASN1_BOOL,
+	[DIRECTIVE_INTEGER]		= ASN1_INT,
+	[DIRECTIVE_BIT]			= ASN1_BTS,
+	[DIRECTIVE_OCTET]		= ASN1_OTS,
+	[DIRECTIVE_NULL]		= ASN1_NULL,
+	[DIRECTIVE_OBJECT]		= ASN1_OID,
+	[DIRECTIVE_ObjectDescriptor]	= ASN1_ODE,
+	[DIRECTIVE_EXTERNAL]		= ASN1_EXT,
+	[DIRECTIVE_REAL]		= ASN1_REAL,
+	[DIRECTIVE_ENUMERATED]		= ASN1_ENUM,
+	[DIRECTIVE_EMBEDDED]		= 0,
+	[DIRECTIVE_UTF8String]		= ASN1_UTF8STR,
+	[DIRECTIVE_RELATIVE_OID]	= ASN1_RELOID,
+	/* 14 */
+	/* 15 */
+	[DIRECTIVE_SEQUENCE]		= ASN1_SEQ,
+	[DIRECTIVE_SET]			= ASN1_SET,
+	[DIRECTIVE_NumericString]	= ASN1_NUMSTR,
+	[DIRECTIVE_PrintableString]	= ASN1_PRNSTR,
+	[DIRECTIVE_T61String]		= ASN1_TEXSTR,
+	[DIRECTIVE_TeletexString]	= ASN1_TEXSTR,
+	[DIRECTIVE_VideotexString]	= ASN1_VIDSTR,
+	[DIRECTIVE_IA5String]		= ASN1_IA5STR,
+	[DIRECTIVE_UTCTime]		= ASN1_UNITIM,
+	[DIRECTIVE_GeneralizedTime]	= ASN1_GENTIM,
+	[DIRECTIVE_GraphicString]	= ASN1_GRASTR,
+	[DIRECTIVE_VisibleString]	= ASN1_VISSTR,
+	[DIRECTIVE_GeneralString]	= ASN1_GENSTR,
+	[DIRECTIVE_UniversalString]	= ASN1_UNITIM,
+	[DIRECTIVE_CHARACTER]		= ASN1_CHRSTR,
+	[DIRECTIVE_BMPString]		= ASN1_BMPSTR,
+};
+
+static const char asn1_classes[4][5] = {
+	[ASN1_UNIV]	= "UNIV",
+	[ASN1_APPL]	= "APPL",
+	[ASN1_CONT]	= "CONT",
+	[ASN1_PRIV]	= "PRIV"
+};
+
+static const char asn1_methods[2][5] = {
+	[ASN1_UNIV]	= "PRIM",
+	[ASN1_APPL]	= "CONS"
+};
+
+static const char *const asn1_universal_tags[32] = {
+	"EOC",
+	"BOOL",
+	"INT",
+	"BTS",
+	"OTS",
+	"NULL",
+	"OID",
+	"ODE",
+	"EXT",
+	"REAL",
+	"ENUM",
+	"EPDV",
+	"UTF8STR",
+	"RELOID",
+	NULL,		/* 14 */
+	NULL,		/* 15 */
+	"SEQ",
+	"SET",
+	"NUMSTR",
+	"PRNSTR",
+	"TEXSTR",
+	"VIDSTR",
+	"IA5STR",
+	"UNITIM",
+	"GENTIM",
+	"GRASTR",
+	"VISSTR",
+	"GENSTR",
+	"UNISTR",
+	"CHRSTR",
+	"BMPSTR",
+	NULL		/* 31 */
+};
+
+static const char *filename;
+static const char *grammar_name;
+static const char *outputname;
+static const char *headername;
+
+static const char *const directives[NR__DIRECTIVES] = {
+#define _(X) [DIRECTIVE_##X] = #X
+	_(ABSENT),
+	_(ALL),
+	_(ANY),
+	_(APPLICATION),
+	_(AUTOMATIC),
+	_(BEGIN),
+	_(BIT),
+	_(BMPString),
+	_(BOOLEAN),
+	_(BY),
+	_(CHARACTER),
+	_(CHOICE),
+	_(CLASS),
+	_(COMPONENT),
+	_(COMPONENTS),
+	_(CONSTRAINED),
+	_(CONTAINING),
+	_(DEFAULT),
+	_(DEFINED),
+	_(DEFINITIONS),
+	_(EMBEDDED),
+	_(ENCODED),
+	[DIRECTIVE_ENCODING_CONTROL] = "ENCODING-CONTROL",
+	_(END),
+	_(ENUMERATED),
+	_(EXCEPT),
+	_(EXPLICIT),
+	_(EXPORTS),
+	_(EXTENSIBILITY),
+	_(EXTERNAL),
+	_(FALSE),
+	_(FROM),
+	_(GeneralString),
+	_(GeneralizedTime),
+	_(GraphicString),
+	_(IA5String),
+	_(IDENTIFIER),
+	_(IMPLICIT),
+	_(IMPLIED),
+	_(IMPORTS),
+	_(INCLUDES),
+	_(INSTANCE),
+	_(INSTRUCTIONS),
+	_(INTEGER),
+	_(INTERSECTION),
+	_(ISO646String),
+	_(MAX),
+	_(MIN),
+	[DIRECTIVE_MINUS_INFINITY] = "MINUS-INFINITY",
+	[DIRECTIVE_NULL] = "NULL",
+	_(NumericString),
+	_(OBJECT),
+	_(OCTET),
+	_(OF),
+	_(OPTIONAL),
+	_(ObjectDescriptor),
+	_(PATTERN),
+	_(PDV),
+	[DIRECTIVE_PLUS_INFINITY] = "PLUS-INFINITY",
+	_(PRESENT),
+	_(PRIVATE),
+	_(PrintableString),
+	_(REAL),
+	[DIRECTIVE_RELATIVE_OID] = "RELATIVE-OID",
+	_(SEQUENCE),
+	_(SET),
+	_(SIZE),
+	_(STRING),
+	_(SYNTAX),
+	_(T61String),
+	_(TAGS),
+	_(TRUE),
+	_(TeletexString),
+	_(UNION),
+	_(UNIQUE),
+	_(UNIVERSAL),
+	_(UTCTime),
+	_(UTF8String),
+	_(UniversalString),
+	_(VideotexString),
+	_(VisibleString),
+	_(WITH)
+};
+
+struct action {
+	struct action	*next;
+	char		*name;
+	unsigned char	index;
+};
+
+static struct action *action_list;
+static unsigned nr_actions;
+
+struct token {
+	unsigned short	line;
+	enum token_type	token_type : 8;
+	unsigned char	size;
+	struct action	*action;
+	char		*content;
+	struct type	*type;
+};
+
+static struct token *token_list;
+static unsigned nr_tokens;
+static bool verbose_opt;
+static bool debug_opt;
+
+#define verbose(fmt, ...) do { if (verbose_opt) printf(fmt, ## __VA_ARGS__); } while (0)
+#define debug(fmt, ...) do { if (debug_opt) printf(fmt, ## __VA_ARGS__); } while (0)
+
+static int directive_compare(const void *_key, const void *_pdir)
+{
+	const struct token *token = _key;
+	const char *const *pdir = _pdir, *dir = *pdir;
+	size_t dlen, clen;
+	int val;
+
+	dlen = strlen(dir);
+	clen = (dlen < token->size) ? dlen : token->size;
+
+	//debug("cmp(%s,%s) = ", token->content, dir);
+
+	val = memcmp(token->content, dir, clen);
+	if (val != 0) {
+		//debug("%d [cmp]\n", val);
+		return val;
+	}
+
+	if (dlen == token->size) {
+		//debug("0\n");
+		return 0;
+	}
+	//debug("%d\n", (int)dlen - (int)token->size);
+	return dlen - token->size; /* shorter -> negative */
+}
+
+/*
+ * Tokenise an ASN.1 grammar
+ */
+static void tokenise(char *buffer, char *end)
+{
+	struct token *tokens;
+	char *line, *nl, *start, *p, *q;
+	unsigned tix, lineno;
+
+	/* Assume we're going to have half as many tokens as we have
+	 * characters
+	 */
+	token_list = tokens = calloc((end - buffer) / 2, sizeof(struct token));
+	if (!tokens) {
+		perror(NULL);
+		exit(1);
+	}
+	tix = 0;
+
+	lineno = 0;
+	while (buffer < end) {
+		/* First of all, break out a line */
+		lineno++;
+		line = buffer;
+		nl = memchr(line, '\n', end - buffer);
+		if (!nl) {
+			buffer = nl = end;
+		} else {
+			buffer = nl + 1;
+			*nl = '\0';
+		}
+
+		/* Remove "--" comments */
+		p = line;
+	next_comment:
+		while ((p = memchr(p, '-', nl - p))) {
+			if (p[1] == '-') {
+				/* Found a comment; see if there's a terminator */
+				q = p + 2;
+				while ((q = memchr(q, '-', nl - q))) {
+					if (q[1] == '-') {
+						/* There is - excise the comment */
+						q += 2;
+						memmove(p, q, nl - q);
+						goto next_comment;
+					}
+					q++;
+				}
+				*p = '\0';
+				nl = p;
+				break;
+			} else {
+				p++;
+			}
+		}
+
+		p = line;
+		while (p < nl) {
+			/* Skip white space */
+			while (p < nl && isspace(*p))
+				*(p++) = 0;
+			if (p >= nl)
+				break;
+
+			tokens[tix].line = lineno;
+			start = p;
+
+			/* Handle string tokens */
+			if (isalpha(*p)) {
+				const char **dir;
+
+				/* Can be a directive, type name or element
+				 * name.  Find the end of the name.
+				 */
+				q = p + 1;
+				while (q < nl && (isalnum(*q) || *q == '-' || *q == '_'))
+					q++;
+				tokens[tix].size = q - p;
+				p = q;
+
+				tokens[tix].content = malloc(tokens[tix].size + 1);
+				if (!tokens[tix].content) {
+					perror(NULL);
+					exit(1);
+				}
+				memcpy(tokens[tix].content, start, tokens[tix].size);
+				tokens[tix].content[tokens[tix].size] = 0;
+
+				/* If it begins with a lowercase letter then
+				 * it's an element name
+				 */
+				if (islower(tokens[tix].content[0])) {
+					tokens[tix++].token_type = TOKEN_ELEMENT_NAME;
+					continue;
+				}
+
+				/* Otherwise we need to search the directive
+				 * table
+				 */
+				dir = bsearch(&tokens[tix], directives,
+					      sizeof(directives) / sizeof(directives[1]),
+					      sizeof(directives[1]),
+					      directive_compare);
+				if (dir) {
+					tokens[tix++].token_type = dir - directives;
+					continue;
+				}
+
+				tokens[tix++].token_type = TOKEN_TYPE_NAME;
+				continue;
+			}
+
+			/* Handle numbers */
+			if (isdigit(*p)) {
+				/* Find the end of the number */
+				q = p + 1;
+				while (q < nl && (isdigit(*q)))
+					q++;
+				tokens[tix].size = q - p;
+				p = q;
+				tokens[tix].content = malloc(tokens[tix].size + 1);
+				if (!tokens[tix].content) {
+					perror(NULL);
+					exit(1);
+				}
+				memcpy(tokens[tix].content, start, tokens[tix].size);
+				tokens[tix].content[tokens[tix].size] = 0;
+				tokens[tix++].token_type = TOKEN_NUMBER;
+				continue;
+			}
+
+			if (nl - p >= 3) {
+				if (memcmp(p, "::=", 3) == 0) {
+					p += 3;
+					tokens[tix].size = 3;
+					tokens[tix].content = "::=";
+					tokens[tix++].token_type = TOKEN_ASSIGNMENT;
+					continue;
+				}
+			}
+
+			if (nl - p >= 2) {
+				if (memcmp(p, "({", 2) == 0) {
+					p += 2;
+					tokens[tix].size = 2;
+					tokens[tix].content = "({";
+					tokens[tix++].token_type = TOKEN_OPEN_ACTION;
+					continue;
+				}
+				if (memcmp(p, "})", 2) == 0) {
+					p += 2;
+					tokens[tix].size = 2;
+					tokens[tix].content = "})";
+					tokens[tix++].token_type = TOKEN_CLOSE_ACTION;
+					continue;
+				}
+			}
+
+			if (nl - p >= 1) {
+				tokens[tix].size = 1;
+				switch (*p) {
+				case '{':
+					p += 1;
+					tokens[tix].content = "{";
+					tokens[tix++].token_type = TOKEN_OPEN_CURLY;
+					continue;
+				case '}':
+					p += 1;
+					tokens[tix].content = "}";
+					tokens[tix++].token_type = TOKEN_CLOSE_CURLY;
+					continue;
+				case '[':
+					p += 1;
+					tokens[tix].content = "[";
+					tokens[tix++].token_type = TOKEN_OPEN_SQUARE;
+					continue;
+				case ']':
+					p += 1;
+					tokens[tix].content = "]";
+					tokens[tix++].token_type = TOKEN_CLOSE_SQUARE;
+					continue;
+				case ',':
+					p += 1;
+					tokens[tix].content = ",";
+					tokens[tix++].token_type = TOKEN_COMMA;
+					continue;
+				default:
+					break;
+				}
+			}
+
+			fprintf(stderr, "%s:%u: Unknown character in grammar: '%c'\n",
+				filename, lineno, *p);
+			exit(1);
+		}
+	}
+
+	nr_tokens = tix;
+	verbose("Extracted %u tokens\n", nr_tokens);
+
+#if 0
+	{
+		int n;
+		for (n = 0; n < nr_tokens; n++)
+			debug("Token %3u: '%s'\n", n, token_list[n].content);
+	}
+#endif
+}
+
+static void build_type_list(void);
+static void parse(void);
+static void dump_elements(void);
+static void render(FILE *out, FILE *hdr);
+
+/*
+ *
+ */
+int main(int argc, char **argv)
+{
+	struct stat st;
+	ssize_t readlen;
+	FILE *out, *hdr;
+	char *buffer, *p;
+	char *kbuild_verbose;
+	int fd;
+
+	kbuild_verbose = getenv("KBUILD_VERBOSE");
+	if (kbuild_verbose)
+		verbose_opt = atoi(kbuild_verbose);
+
+	while (argc > 4) {
+		if (strcmp(argv[1], "-v") == 0)
+			verbose_opt = true;
+		else if (strcmp(argv[1], "-d") == 0)
+			debug_opt = true;
+		else
+			break;
+		memmove(&argv[1], &argv[2], (argc - 2) * sizeof(char *));
+		argc--;
+	}
+
+	if (argc != 4) {
+		fprintf(stderr, "Format: %s [-v] [-d] <grammar-file> <c-file> <hdr-file>\n",
+			argv[0]);
+		exit(2);
+	}
+
+	filename = argv[1];
+	outputname = argv[2];
+	headername = argv[3];
+
+	fd = open(filename, O_RDONLY);
+	if (fd < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (fstat(fd, &st) < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (!(buffer = malloc(st.st_size + 1))) {
+		perror(NULL);
+		exit(1);
+	}
+
+	if ((readlen = read(fd, buffer, st.st_size)) < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (close(fd) < 0) {
+		perror(filename);
+		exit(1);
+	}
+
+	if (readlen != st.st_size) {
+		fprintf(stderr, "%s: Short read\n", filename);
+		exit(1);
+	}
+
+	p = strrchr(argv[1], '/');
+	p = p ? p + 1 : argv[1];
+	grammar_name = strdup(p);
+	if (!p) {
+		perror(NULL);
+		exit(1);
+	}
+	p = strchr(grammar_name, '.');
+	if (p)
+		*p = '\0';
+
+	buffer[readlen] = 0;
+	tokenise(buffer, buffer + readlen);
+	build_type_list();
+	parse();
+	dump_elements();
+
+	out = fopen(outputname, "w");
+	if (!out) {
+		perror(outputname);
+		exit(1);
+	}
+
+	hdr = fopen(headername, "w");
+	if (!hdr) {
+		perror(headername);
+		exit(1);
+	}
+
+	render(out, hdr);
+
+	if (fclose(out) < 0) {
+		perror(outputname);
+		exit(1);
+	}
+
+	if (fclose(hdr) < 0) {
+		perror(headername);
+		exit(1);
+	}
+
+	return 0;
+}
+
+enum compound {
+	NOT_COMPOUND,
+	SET,
+	SET_OF,
+	SEQUENCE,
+	SEQUENCE_OF,
+	CHOICE,
+	ANY,
+	TYPE_REF,
+	TAG_OVERRIDE
+};
+
+struct element {
+	struct type	*type_def;
+	struct token	*name;
+	struct token	*type;
+	struct action	*action;
+	struct element	*children;
+	struct element	*next;
+	struct element	*render_next;
+	struct element	*list_next;
+	uint8_t		n_elements;
+	enum compound	compound : 8;
+	enum asn1_class	class : 8;
+	enum asn1_method method : 8;
+	uint8_t		tag;
+	unsigned	entry_index;
+	unsigned	flags;
+#define ELEMENT_IMPLICIT	0x0001
+#define ELEMENT_EXPLICIT	0x0002
+#define ELEMENT_TAG_SPECIFIED	0x0004
+#define ELEMENT_RENDERED	0x0008
+#define ELEMENT_SKIPPABLE	0x0010
+#define ELEMENT_CONDITIONAL	0x0020
+};
+
+struct type {
+	struct token	*name;
+	struct token	*def;
+	struct element	*element;
+	unsigned	ref_count;
+	unsigned	flags;
+#define TYPE_STOP_MARKER	0x0001
+#define TYPE_BEGIN		0x0002
+};
+
+static struct type *type_list;
+static struct type **type_index;
+static unsigned nr_types;
+
+static int type_index_compare(const void *_a, const void *_b)
+{
+	const struct type *const *a = _a, *const *b = _b;
+
+	if ((*a)->name->size != (*b)->name->size)
+		return (*a)->name->size - (*b)->name->size;
+	else
+		return memcmp((*a)->name->content, (*b)->name->content,
+			      (*a)->name->size);
+}
+
+static int type_finder(const void *_key, const void *_ti)
+{
+	const struct token *token = _key;
+	const struct type *const *ti = _ti;
+	const struct type *type = *ti;
+
+	if (token->size != type->name->size)
+		return token->size - type->name->size;
+	else
+		return memcmp(token->content, type->name->content,
+			      token->size);
+}
+
+/*
+ * Build up a list of types and a sorted index to that list.
+ */
+static void build_type_list(void)
+{
+	struct type *types;
+	unsigned nr, t, n;
+
+	nr = 0;
+	for (n = 0; n < nr_tokens - 1; n++)
+		if (token_list[n + 0].token_type == TOKEN_TYPE_NAME &&
+		    token_list[n + 1].token_type == TOKEN_ASSIGNMENT)
+			nr++;
+
+	if (nr == 0) {
+		fprintf(stderr, "%s: No defined types\n", filename);
+		exit(1);
+	}
+
+	nr_types = nr;
+	types = type_list = calloc(nr + 1, sizeof(type_list[0]));
+	if (!type_list) {
+		perror(NULL);
+		exit(1);
+	}
+	type_index = calloc(nr, sizeof(type_index[0]));
+	if (!type_index) {
+		perror(NULL);
+		exit(1);
+	}
+
+	t = 0;
+	types[t].flags |= TYPE_BEGIN;
+	for (n = 0; n < nr_tokens - 1; n++) {
+		if (token_list[n + 0].token_type == TOKEN_TYPE_NAME &&
+		    token_list[n + 1].token_type == TOKEN_ASSIGNMENT) {
+			types[t].name = &token_list[n];
+			type_index[t] = &types[t];
+			t++;
+		}
+	}
+	types[t].name = &token_list[n + 1];
+	types[t].flags |= TYPE_STOP_MARKER;
+
+	qsort(type_index, nr, sizeof(type_index[0]), type_index_compare);
+
+	verbose("Extracted %u types\n", nr_types);
+#if 0
+	for (n = 0; n < nr_types; n++) {
+		struct type *type = type_index[n];
+		debug("- %*.*s\n", type->name->content);
+	}
+#endif
+}
+
+static struct element *parse_type(struct token **_cursor, struct token *stop,
+				  struct token *name);
+
+/*
+ * Parse the token stream
+ */
+static void parse(void)
+{
+	struct token *cursor;
+	struct type *type;
+
+	/* Parse one type definition statement at a time */
+	type = type_list;
+	do {
+		cursor = type->name;
+
+		if (cursor[0].token_type != TOKEN_TYPE_NAME ||
+		    cursor[1].token_type != TOKEN_ASSIGNMENT)
+			abort();
+		cursor += 2;
+
+		type->element = parse_type(&cursor, type[1].name, NULL);
+		type->element->type_def = type;
+
+		if (cursor != type[1].name) {
+			fprintf(stderr, "%s:%d: Parse error at token '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+	} while (type++, !(type->flags & TYPE_STOP_MARKER));
+
+	verbose("Extracted %u actions\n", nr_actions);
+}
+
+static struct element *element_list;
+
+static struct element *alloc_elem(struct token *type)
+{
+	struct element *e = calloc(1, sizeof(*e));
+	if (!e) {
+		perror(NULL);
+		exit(1);
+	}
+	e->list_next = element_list;
+	element_list = e;
+	return e;
+}
+
+static struct element *parse_compound(struct token **_cursor, struct token *end,
+				      int alternates);
+
+/*
+ * Parse one type definition statement
+ */
+static struct element *parse_type(struct token **_cursor, struct token *end,
+				  struct token *name)
+{
+	struct element *top, *element;
+	struct action *action, **ppaction;
+	struct token *cursor = *_cursor;
+	struct type **ref;
+	char *p;
+	int labelled = 0, implicit = 0;
+
+	top = element = alloc_elem(cursor);
+	element->class = ASN1_UNIV;
+	element->method = ASN1_PRIM;
+	element->tag = token_to_tag[cursor->token_type];
+	element->name = name;
+
+	/* Extract the tag value if one given */
+	if (cursor->token_type == TOKEN_OPEN_SQUARE) {
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		switch (cursor->token_type) {
+		case DIRECTIVE_UNIVERSAL:
+			element->class = ASN1_UNIV;
+			cursor++;
+			break;
+		case DIRECTIVE_APPLICATION:
+			element->class = ASN1_APPL;
+			cursor++;
+			break;
+		case TOKEN_NUMBER:
+			element->class = ASN1_CONT;
+			break;
+		case DIRECTIVE_PRIVATE:
+			element->class = ASN1_PRIV;
+			cursor++;
+			break;
+		default:
+			fprintf(stderr, "%s:%d: Unrecognised tag class token '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_NUMBER) {
+			fprintf(stderr, "%s:%d: Missing tag number '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+		element->tag &= ~0x1f;
+		element->tag |= strtoul(cursor->content, &p, 10);
+		element->flags |= ELEMENT_TAG_SPECIFIED;
+		if (p - cursor->content != cursor->size)
+			abort();
+		cursor++;
+
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_CLOSE_SQUARE) {
+			fprintf(stderr, "%s:%d: Missing closing square bracket '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		labelled = 1;
+	}
+
+	/* Handle implicit and explicit markers */
+	if (cursor->token_type == DIRECTIVE_IMPLICIT) {
+		element->flags |= ELEMENT_IMPLICIT;
+		implicit = 1;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+	} else if (cursor->token_type == DIRECTIVE_EXPLICIT) {
+		element->flags |= ELEMENT_EXPLICIT;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+	}
+
+	if (labelled) {
+		if (!implicit)
+			element->method |= ASN1_CONS;
+		element->compound = implicit ? TAG_OVERRIDE : SEQUENCE;
+		element->children = alloc_elem(cursor);
+		element = element->children;
+		element->class = ASN1_UNIV;
+		element->method = ASN1_PRIM;
+		element->tag = token_to_tag[cursor->token_type];
+		element->name = name;
+	}
+
+	/* Extract the type we're expecting here */
+	element->type = cursor;
+	switch (cursor->token_type) {
+	case DIRECTIVE_ANY:
+		element->compound = ANY;
+		cursor++;
+		break;
+
+	case DIRECTIVE_NULL:
+	case DIRECTIVE_BOOLEAN:
+	case DIRECTIVE_ENUMERATED:
+	case DIRECTIVE_INTEGER:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		break;
+
+	case DIRECTIVE_EXTERNAL:
+		element->method = ASN1_CONS;
+
+	case DIRECTIVE_BMPString:
+	case DIRECTIVE_GeneralString:
+	case DIRECTIVE_GraphicString:
+	case DIRECTIVE_IA5String:
+	case DIRECTIVE_ISO646String:
+	case DIRECTIVE_NumericString:
+	case DIRECTIVE_PrintableString:
+	case DIRECTIVE_T61String:
+	case DIRECTIVE_TeletexString:
+	case DIRECTIVE_UniversalString:
+	case DIRECTIVE_UTF8String:
+	case DIRECTIVE_VideotexString:
+	case DIRECTIVE_VisibleString:
+	case DIRECTIVE_ObjectDescriptor:
+	case DIRECTIVE_GeneralizedTime:
+	case DIRECTIVE_UTCTime:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		break;
+
+	case DIRECTIVE_BIT:
+	case DIRECTIVE_OCTET:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != DIRECTIVE_STRING)
+			goto parse_error;
+		cursor++;
+		break;
+
+	case DIRECTIVE_OBJECT:
+		element->compound = NOT_COMPOUND;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != DIRECTIVE_IDENTIFIER)
+			goto parse_error;
+		cursor++;
+		break;
+
+	case TOKEN_TYPE_NAME:
+		element->compound = TYPE_REF;
+		ref = bsearch(cursor, type_index, nr_types, sizeof(type_index[0]),
+			      type_finder);
+		if (!ref) {
+			fprintf(stderr, "%s:%d: Type '%s' undefined\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+		cursor->type = *ref;
+		(*ref)->ref_count++;
+		cursor++;
+		break;
+
+	case DIRECTIVE_CHOICE:
+		element->compound = CHOICE;
+		cursor++;
+		element->children = parse_compound(&cursor, end, 1);
+		break;
+
+	case DIRECTIVE_SEQUENCE:
+		element->compound = SEQUENCE;
+		element->method = ASN1_CONS;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type == DIRECTIVE_OF) {
+			element->compound = SEQUENCE_OF;
+			cursor++;
+			if (cursor >= end)
+				goto overrun_error;
+			element->children = parse_type(&cursor, end, NULL);
+		} else {
+			element->children = parse_compound(&cursor, end, 0);
+		}
+		break;
+
+	case DIRECTIVE_SET:
+		element->compound = SET;
+		element->method = ASN1_CONS;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type == DIRECTIVE_OF) {
+			element->compound = SET_OF;
+			cursor++;
+			if (cursor >= end)
+				goto parse_error;
+			element->children = parse_type(&cursor, end, NULL);
+		} else {
+			element->children = parse_compound(&cursor, end, 1);
+		}
+		break;
+
+	default:
+		fprintf(stderr, "%s:%d: Token '%s' does not introduce a type\n",
+			filename, cursor->line, cursor->content);
+		exit(1);
+	}
+
+	/* Handle elements that are optional */
+	if (cursor < end && (cursor->token_type == DIRECTIVE_OPTIONAL ||
+			     cursor->token_type == DIRECTIVE_DEFAULT)
+	    ) {
+		cursor++;
+		top->flags |= ELEMENT_SKIPPABLE;
+	}
+
+	if (cursor < end && cursor->token_type == TOKEN_OPEN_ACTION) {
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_ELEMENT_NAME) {
+			fprintf(stderr, "%s:%d: Token '%s' is not an action function name\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+
+		action = malloc(sizeof(struct action));
+		if (!action) {
+			perror(NULL);
+			exit(1);
+		}
+		action->index = 0;
+		action->name = cursor->content;
+
+		for (ppaction = &action_list;
+		     *ppaction;
+		     ppaction = &(*ppaction)->next
+		     ) {
+			int cmp = strcmp(action->name, (*ppaction)->name);
+			if (cmp == 0) {
+				free(action);
+				action = *ppaction;
+				goto found;
+			}
+			if (cmp < 0) {
+				action->next = *ppaction;
+				*ppaction = action;
+				nr_actions++;
+				goto found;
+			}
+		}
+		action->next = NULL;
+		*ppaction = action;
+		nr_actions++;
+	found:
+
+		element->action = action;
+		cursor->action = action;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_CLOSE_ACTION) {
+			fprintf(stderr, "%s:%d: Missing close action, got '%s'\n",
+				filename, cursor->line, cursor->content);
+			exit(1);
+		}
+		cursor++;
+	}
+
+	*_cursor = cursor;
+	return top;
+
+parse_error:
+	fprintf(stderr, "%s:%d: Unexpected token '%s'\n",
+		filename, cursor->line, cursor->content);
+	exit(1);
+
+overrun_error:
+	fprintf(stderr, "%s: Unexpectedly hit EOF\n", filename);
+	exit(1);
+}
+
+/*
+ * Parse a compound type list
+ */
+static struct element *parse_compound(struct token **_cursor, struct token *end,
+				      int alternates)
+{
+	struct element *children, **child_p = &children, *element;
+	struct token *cursor = *_cursor, *name;
+
+	if (cursor->token_type != TOKEN_OPEN_CURLY) {
+		fprintf(stderr, "%s:%d: Expected compound to start with brace not '%s'\n",
+			filename, cursor->line, cursor->content);
+		exit(1);
+	}
+	cursor++;
+	if (cursor >= end)
+		goto overrun_error;
+
+	if (cursor->token_type == TOKEN_OPEN_CURLY) {
+		fprintf(stderr, "%s:%d: Empty compound\n",
+			filename, cursor->line);
+		exit(1);
+	}
+
+	for (;;) {
+		name = NULL;
+		if (cursor->token_type == TOKEN_ELEMENT_NAME) {
+			name = cursor;
+			cursor++;
+			if (cursor >= end)
+				goto overrun_error;
+		}
+
+		element = parse_type(&cursor, end, name);
+		if (alternates)
+			element->flags |= ELEMENT_SKIPPABLE | ELEMENT_CONDITIONAL;
+
+		*child_p = element;
+		child_p = &element->next;
+
+		if (cursor >= end)
+			goto overrun_error;
+		if (cursor->token_type != TOKEN_COMMA)
+			break;
+		cursor++;
+		if (cursor >= end)
+			goto overrun_error;
+	}
+
+	children->flags &= ~ELEMENT_CONDITIONAL;
+
+	if (cursor->token_type != TOKEN_CLOSE_CURLY) {
+		fprintf(stderr, "%s:%d: Expected compound closure, got '%s'\n",
+			filename, cursor->line, cursor->content);
+		exit(1);
+	}
+	cursor++;
+
+	*_cursor = cursor;
+	return children;
+
+overrun_error:
+	fprintf(stderr, "%s: Unexpectedly hit EOF\n", filename);
+	exit(1);
+}
+
+static void dump_element(const struct element *e, int level)
+{
+	const struct element *c;
+	const struct type *t = e->type_def;
+	const char *name = e->name ? e->name->content : ".";
+	const char *tname = t && t->name ? t->name->content : ".";
+	char tag[32];
+
+	if (e->class == 0 && e->method == 0 && e->tag == 0)
+		strcpy(tag, "<...>");
+	else if (e->class == ASN1_UNIV)
+		sprintf(tag, "%s %s %s",
+			asn1_classes[e->class],
+			asn1_methods[e->method],
+			asn1_universal_tags[e->tag]);
+	else
+		sprintf(tag, "%s %s %u",
+			asn1_classes[e->class],
+			asn1_methods[e->method],
+			e->tag);
+
+	printf("%c%c%c%c%c %c %*s[*] \e[33m%s\e[m %s %s \e[35m%s\e[m\n",
+	       e->flags & ELEMENT_IMPLICIT ? 'I' : '-',
+	       e->flags & ELEMENT_EXPLICIT ? 'E' : '-',
+	       e->flags & ELEMENT_TAG_SPECIFIED ? 'T' : '-',
+	       e->flags & ELEMENT_SKIPPABLE ? 'S' : '-',
+	       e->flags & ELEMENT_CONDITIONAL ? 'C' : '-',
+	       "-tTqQcaro"[e->compound],
+	       level, "",
+	       tag,
+	       tname,
+	       name,
+	       e->action ? e->action->name : "");
+	if (e->compound == TYPE_REF)
+		dump_element(e->type->type->element, level + 3);
+	else
+		for (c = e->children; c; c = c->next)
+			dump_element(c, level + 3);
+}
+
+static void dump_elements(void)
+{
+	if (debug_opt)
+		dump_element(type_list[0].element, 0);
+}
+
+static void render_element(FILE *out, struct element *e, struct element *tag);
+static void render_out_of_line_list(FILE *out);
+
+static int nr_entries;
+static int render_depth = 1;
+static struct element *render_list, **render_list_p = &render_list;
+
+__attribute__((format(printf, 2, 3)))
+static void render_opcode(FILE *out, const char *fmt, ...)
+{
+	va_list va;
+
+	if (out) {
+		fprintf(out, "\t[%4d] =%*s", nr_entries, render_depth, "");
+		va_start(va, fmt);
+		vfprintf(out, fmt, va);
+		va_end(va);
+	}
+	nr_entries++;
+}
+
+__attribute__((format(printf, 2, 3)))
+static void render_more(FILE *out, const char *fmt, ...)
+{
+	va_list va;
+
+	if (out) {
+		va_start(va, fmt);
+		vfprintf(out, fmt, va);
+		va_end(va);
+	}
+}
+
+/*
+ * Render the grammar into a state machine definition.
+ */
+static void render(FILE *out, FILE *hdr)
+{
+	struct element *e;
+	struct action *action;
+	struct type *root;
+	int index;
+
+	fprintf(hdr, "/*\n");
+	fprintf(hdr, " * Automatically generated by asn1_compiler.  Do not edit\n");
+	fprintf(hdr, " *\n");
+	fprintf(hdr, " * ASN.1 parser for %s\n", grammar_name);
+	fprintf(hdr, " */\n");
+	fprintf(hdr, "#include <linux/asn1_decoder.h>\n");
+	fprintf(hdr, "\n");
+	fprintf(hdr, "extern const struct asn1_decoder %s_decoder;\n", grammar_name);
+	if (ferror(hdr)) {
+		perror(headername);
+		exit(1);
+	}
+
+	fprintf(out, "/*\n");
+	fprintf(out, " * Automatically generated by asn1_compiler.  Do not edit\n");
+	fprintf(out, " *\n");
+	fprintf(out, " * ASN.1 parser for %s\n", grammar_name);
+	fprintf(out, " */\n");
+	fprintf(out, "#include <linux/asn1_ber_bytecode.h>\n");
+	fprintf(out, "#include \"%s.asn1.h\"\n", grammar_name);
+	fprintf(out, "\n");
+	if (ferror(out)) {
+		perror(outputname);
+		exit(1);
+	}
+
+	/* Tabulate the action functions we might have to call */
+	fprintf(hdr, "\n");
+	index = 0;
+	for (action = action_list; action; action = action->next) {
+		action->index = index++;
+		fprintf(hdr,
+			"extern int %s(void *, size_t, unsigned char,"
+			" const void *, size_t);\n",
+			action->name);
+	}
+	fprintf(hdr, "\n");
+
+	fprintf(out, "enum %s_actions {\n", grammar_name);
+	for (action = action_list; action; action = action->next)
+		fprintf(out, "\tACT_%s = %u,\n",
+			action->name, action->index);
+	fprintf(out, "\tNR__%s_actions = %u\n", grammar_name, nr_actions);
+	fprintf(out, "};\n");
+
+	fprintf(out, "\n");
+	fprintf(out, "static const asn1_action_t %s_action_table[NR__%s_actions] = {\n",
+		grammar_name, grammar_name);
+	for (action = action_list; action; action = action->next)
+		fprintf(out, "\t[%4u] = %s,\n", action->index, action->name);
+	fprintf(out, "};\n");
+
+	if (ferror(out)) {
+		perror(outputname);
+		exit(1);
+	}
+
+	/* We do two passes - the first one calculates all the offsets */
+	verbose("Pass 1\n");
+	nr_entries = 0;
+	root = &type_list[0];
+	render_element(NULL, root->element, NULL);
+	render_opcode(NULL, "ASN1_OP_COMPLETE,\n");
+	render_out_of_line_list(NULL);
+
+	for (e = element_list; e; e = e->list_next)
+		e->flags &= ~ELEMENT_RENDERED;
+
+	/* And then we actually render */
+	verbose("Pass 2\n");
+	fprintf(out, "\n");
+	fprintf(out, "static const unsigned char %s_machine[] = {\n",
+		grammar_name);
+
+	nr_entries = 0;
+	root = &type_list[0];
+	render_element(out, root->element, NULL);
+	render_opcode(out, "ASN1_OP_COMPLETE,\n");
+	render_out_of_line_list(out);
+
+	fprintf(out, "};\n");
+
+	fprintf(out, "\n");
+	fprintf(out, "const struct asn1_decoder %s_decoder = {\n", grammar_name);
+	fprintf(out, "\t.machine = %s_machine,\n", grammar_name);
+	fprintf(out, "\t.machlen = sizeof(%s_machine),\n", grammar_name);
+	fprintf(out, "\t.actions = %s_action_table,\n", grammar_name);
+	fprintf(out, "};\n");
+}
+
+/*
+ * Render the out-of-line elements
+ */
+static void render_out_of_line_list(FILE *out)
+{
+	struct element *e, *ce;
+	const char *act;
+	int entry;
+
+	while ((e = render_list)) {
+		render_list = e->render_next;
+		if (!render_list)
+			render_list_p = &render_list;
+
+		render_more(out, "\n");
+		e->entry_index = entry = nr_entries;
+		render_depth++;
+		for (ce = e->children; ce; ce = ce->next)
+			render_element(out, ce, NULL);
+		render_depth--;
+
+		act = e->action ? "_ACT" : "";
+		switch (e->compound) {
+		case SEQUENCE:
+			render_opcode(out, "ASN1_OP_END_SEQ%s,\n", act);
+			break;
+		case SEQUENCE_OF:
+			render_opcode(out, "ASN1_OP_END_SEQ_OF%s,\n", act);
+			render_opcode(out, "_jump_target(%u),\n", entry);
+			break;
+		case SET:
+			render_opcode(out, "ASN1_OP_END_SET%s,\n", act);
+			break;
+		case SET_OF:
+			render_opcode(out, "ASN1_OP_END_SET_OF%s,\n", act);
+			render_opcode(out, "_jump_target(%u),\n", entry);
+			break;
+		default:
+			break;
+		}
+		if (e->action)
+			render_opcode(out, "_action(ACT_%s),\n",
+				      e->action->name);
+		render_opcode(out, "ASN1_OP_RETURN,\n");
+	}
+}
+
+/*
+ * Render an element.
+ */
+static void render_element(FILE *out, struct element *e, struct element *tag)
+{
+	struct element *ec, *x;
+	const char *cond, *act;
+	int entry, skippable = 0, outofline = 0;
+
+	if (e->flags & ELEMENT_SKIPPABLE ||
+	    (tag && tag->flags & ELEMENT_SKIPPABLE))
+		skippable = 1;
+
+	if ((e->type_def && e->type_def->ref_count > 1) ||
+	    skippable)
+		outofline = 1;
+
+	if (e->type_def && out) {
+		render_more(out, "\t// %s\n", e->type_def->name->content);
+	}
+
+	/* Render the operation */
+	cond = (e->flags & ELEMENT_CONDITIONAL ||
+		(tag && tag->flags & ELEMENT_CONDITIONAL)) ? "COND_" : "";
+	act = e->action ? "_ACT" : "";
+	switch (e->compound) {
+	case ANY:
+		render_opcode(out, "ASN1_OP_%sMATCH_ANY%s%s,",
+			      cond, act, skippable ? "_OR_SKIP" : "");
+		if (e->name)
+			render_more(out, "\t\t// %s", e->name->content);
+		render_more(out, "\n");
+		goto dont_render_tag;
+
+	case TAG_OVERRIDE:
+		render_element(out, e->children, e);
+		return;
+
+	case SEQUENCE:
+	case SEQUENCE_OF:
+	case SET:
+	case SET_OF:
+		render_opcode(out, "ASN1_OP_%sMATCH%s%s,",
+			      cond,
+			      outofline ? "_JUMP" : "",
+			      skippable ? "_OR_SKIP" : "");
+		break;
+
+	case CHOICE:
+		goto dont_render_tag;
+
+	case TYPE_REF:
+		if (e->class == ASN1_UNIV && e->method == ASN1_PRIM && e->tag == 0)
+			goto dont_render_tag;
+	default:
+		render_opcode(out, "ASN1_OP_%sMATCH%s%s,",
+			      cond, act,
+			      skippable ? "_OR_SKIP" : "");
+		break;
+	}
+
+	x = tag ?: e;
+	if (x->name)
+		render_more(out, "\t\t// %s", x->name->content);
+	render_more(out, "\n");
+
+	/* Render the tag */
+	if (!tag || !(tag->flags & ELEMENT_TAG_SPECIFIED))
+		tag = e;
+
+	if (tag->class == ASN1_UNIV &&
+	    tag->tag != 14 &&
+	    tag->tag != 15 &&
+	    tag->tag != 31)
+		render_opcode(out, "_tag(%s, %s, %s),\n",
+			      asn1_classes[tag->class],
+			      asn1_methods[tag->method | e->method],
+			      asn1_universal_tags[tag->tag]);
+	else
+		render_opcode(out, "_tagn(%s, %s, %2u),\n",
+			      asn1_classes[tag->class],
+			      asn1_methods[tag->method | e->method],
+			      tag->tag);
+	tag = NULL;
+dont_render_tag:
+
+	/* Deal with compound types */
+	switch (e->compound) {
+	case TYPE_REF:
+		render_element(out, e->type->type->element, tag);
+		if (e->action)
+			render_opcode(out, "ASN1_OP_%sACT,\n",
+				      skippable ? "MAYBE_" : "");
+		break;
+
+	case SEQUENCE:
+		if (outofline) {
+			/* Render out-of-line for multiple use or
+			 * skipability */
+			render_opcode(out, "_jump_target(%u),", e->entry_index);
+			if (e->type_def && e->type_def->name)
+				render_more(out, "\t\t// --> %s",
+					    e->type_def->name->content);
+			render_more(out, "\n");
+			if (!(e->flags & ELEMENT_RENDERED)) {
+				e->flags |= ELEMENT_RENDERED;
+				*render_list_p = e;
+				render_list_p = &e->render_next;
+			}
+			return;
+		} else {
+			/* Render inline for single use */
+			render_depth++;
+			for (ec = e->children; ec; ec = ec->next)
+				render_element(out, ec, NULL);
+			render_depth--;
+			render_opcode(out, "ASN1_OP_END_SEQ%s,\n", act);
+		}
+		break;
+
+	case SEQUENCE_OF:
+	case SET_OF:
+		if (outofline) {
+			/* Render out-of-line for multiple use or
+			 * skipability */
+			render_opcode(out, "_jump_target(%u),", e->entry_index);
+			if (e->type_def && e->type_def->name)
+				render_more(out, "\t\t// --> %s",
+					    e->type_def->name->content);
+			render_more(out, "\n");
+			if (!(e->flags & ELEMENT_RENDERED)) {
+				e->flags |= ELEMENT_RENDERED;
+				*render_list_p = e;
+				render_list_p = &e->render_next;
+			}
+			return;
+		} else {
+			/* Render inline for single use */
+			entry = nr_entries;
+			render_depth++;
+			render_element(out, e->children, NULL);
+			render_depth--;
+			if (e->compound == SEQUENCE_OF)
+				render_opcode(out, "ASN1_OP_END_SEQ_OF%s,\n", act);
+			else
+				render_opcode(out, "ASN1_OP_END_SET_OF%s,\n", act);
+			render_opcode(out, "_jump_target(%u),\n", entry);
+		}
+		break;
+
+	case SET:
+		/* I can't think of a nice way to do SET support without having
+		 * a stack of bitmasks to make sure no element is repeated.
+		 * The bitmask has also to be checked that no non-optional
+		 * elements are left out whilst not preventing optional
+		 * elements from being left out.
+		 */
+		fprintf(stderr, "The ASN.1 SET type is not currently supported.\n");
+		exit(1);
+
+	case CHOICE:
+		for (ec = e->children; ec; ec = ec->next)
+			render_element(out, ec, ec);
+		if (!skippable)
+			render_opcode(out, "ASN1_OP_COND_FAIL,\n");
+		if (e->action)
+			render_opcode(out, "ASN1_OP_ACT,\n");
+		break;
+
+	default:
+		break;
+	}
+
+	if (e->action)
+		render_opcode(out, "_action(ACT_%s),\n", e->action->name);
+}
#15 lore raw

PATCH 5/11 Makefile: add build script for asn1 parsers

In reply to AKASHI Takahiro

This rule will be used to build x509 and pkcs7 parsers.

---
 scripts/Makefile.build | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
scripts/Makefile.build +4 -2
index f7a041296d3d..9357d310e50b 100644
@@ -331,7 +331,7 @@ quiet_cmd_asn1_compiler = ASN.1   $@
       cmd_asn1_compiler = $(objtree)/scripts/asn1_compiler $< \
 				$(subst .h,.c,$@) $(subst .c,.h,$@)
 
-$(obj)/%-asn1.c $(obj)/%-asn1.h: $(src)/%.asn1 $(objtree)/scripts/asn1_compiler
+$(obj)/%.asn1.c $(obj)/%.asn1.h: $(src)/%.asn1 $(objtree)/scripts/asn1_compiler
 	$(call cmd,asn1_compiler)
 
 # Build the compiled-in targets
@@ -419,9 +419,11 @@ targets += $(multi-used-y) $(multi-used-m)
 intermediate_targets = $(foreach sfx, $(2), \
 				$(patsubst %$(strip $(1)),%$(sfx), \
 					$(filter %$(strip $(1)), $(targets))))
+# %.asn1.o <- %.asn1.[ch] <- %.asn1
 # %.lex.o <- %.lex.c <- %.l
 # %.tab.o <- %.tab.[ch] <- %.y
-targets += $(call intermediate_targets, .lex.o, .lex.c) \
+targets += $(call intermediate_targets, .asn1.o, .asn1.c .asn1.h) \
+	   $(call intermediate_targets, .lex.o, .lex.c) \
 	   $(call intermediate_targets, .tab.o, .tab.c .tab.h)
 
 # Descending
Signature · 1 line
2.21.0
#16 lore raw

PATCH 5/11 Makefile: add build script for asn1 parsers

Quoted 3 lines · AKASHI Takahiro

This rule will be used to build x509 and pkcs7 parsers.

The ASN1 compiler is generating both *.asn.h and *.asn.c files.

Then I would expect that these files are deleted by `make mrproper'.

Please, adjust the clean: target in /Makefile accordingly.

Best regards

Heinrich

Quoted 31 lines
---
  scripts/Makefile.build | 6 ++++--
  1 file changed, 4 insertions(+), 2 deletions(-)
scripts/Makefile.build +4 -2
index f7a041296d3d..9357d310e50b 100644
@@ -331,7 +331,7 @@ quiet_cmd_asn1_compiler = ASN.1   $@
        cmd_asn1_compiler = $(objtree)/scripts/asn1_compiler $< \
  				$(subst .h,.c,$@) $(subst .c,.h,$@)
 
-$(obj)/%-asn1.c $(obj)/%-asn1.h: $(src)/%.asn1 $(objtree)/scripts/asn1_compiler
+$(obj)/%.asn1.c $(obj)/%.asn1.h: $(src)/%.asn1 $(objtree)/scripts/asn1_compiler
  	$(call cmd,asn1_compiler)
 
  # Build the compiled-in targets
@@ -419,9 +419,11 @@ targets += $(multi-used-y) $(multi-used-m)
  intermediate_targets = $(foreach sfx, $(2), \
  				$(patsubst %$(strip $(1)),%$(sfx), \
  					$(filter %$(strip $(1)), $(targets))))
+# %.asn1.o <- %.asn1.[ch] <- %.asn1
  # %.lex.o <- %.lex.c <- %.l
  # %.tab.o <- %.tab.[ch] <- %.y
-targets += $(call intermediate_targets, .lex.o, .lex.c) \
+targets += $(call intermediate_targets, .asn1.o, .asn1.c .asn1.h) \
+	   $(call intermediate_targets, .lex.o, .lex.c) \
  	   $(call intermediate_targets, .tab.o, .tab.c .tab.h)
 
  # Descending
#17 lore raw

PATCH 5/11 Makefile: add build script for asn1 parsers

Quoted 10 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

This rule will be used to build x509 and pkcs7 parsers.

The ASN1 compiler is generating both *.asn.h and *.asn.c files.

Then I would expect that these files are deleted by `make mrproper'.

Please, adjust the clean: target in /Makefile accordingly.

Good catch, thank you.

-Takahiro Akashi

Quoted 36 lines

Best regards

Heinrich

Quoted 31 lines
---
 scripts/Makefile.build | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
scripts/Makefile.build +4 -2
index f7a041296d3d..9357d310e50b 100644
@@ -331,7 +331,7 @@ quiet_cmd_asn1_compiler = ASN.1   $@
       cmd_asn1_compiler = $(objtree)/scripts/asn1_compiler $< \
 				$(subst .h,.c,$@) $(subst .c,.h,$@)
 
-$(obj)/%-asn1.c $(obj)/%-asn1.h: $(src)/%.asn1 $(objtree)/scripts/asn1_compiler
+$(obj)/%.asn1.c $(obj)/%.asn1.h: $(src)/%.asn1 $(objtree)/scripts/asn1_compiler
 	$(call cmd,asn1_compiler)
 
 # Build the compiled-in targets
@@ -419,9 +419,11 @@ targets += $(multi-used-y) $(multi-used-m)
 intermediate_targets = $(foreach sfx, $(2), \
 				$(patsubst %$(strip $(1)),%$(sfx), \
 					$(filter %$(strip $(1)), $(targets))))
+# %.asn1.o <- %.asn1.[ch] <- %.asn1
 # %.lex.o <- %.lex.c <- %.l
 # %.tab.o <- %.tab.[ch] <- %.y
-targets += $(call intermediate_targets, .lex.o, .lex.c) \
+targets += $(call intermediate_targets, .asn1.o, .asn1.c .asn1.h) \
+	   $(call intermediate_targets, .lex.o, .lex.c) \
 	   $(call intermediate_targets, .tab.o, .tab.c .tab.h)
 
 # Descending
#18 lore raw

PATCH 6/11 lib: add asn1 decoder

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

---
 lib/Kconfig        |   6 +
 lib/Makefile       |   1 +
 lib/asn1_decoder.c | 527 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 534 insertions(+)
 create mode 100644 lib/asn1_decoder.c
lib/Kconfig +6 -0
index 3da45a5ec322..26c94f49ecd2 100644
@@ -521,6 +521,12 @@ config SMBIOS_PRODUCT_NAME
 
 endmenu
 
+config ASN1
+	bool
+	select BUILD_ASN1
+	help
+	  Enable asn1 decoder library.
+
 source lib/efi/Kconfig
 source lib/efi_loader/Kconfig
 source lib/optee/Kconfig
lib/Makefile +1 -0
index 2fffd68f943c..eb3a675fb8c2 100644
@@ -17,6 +17,7 @@ obj-$(CONFIG_OF_LIVE) += of_live.o
 obj-$(CONFIG_CMD_DHRYSTONE) += dhry/
 obj-$(CONFIG_ARCH_AT91) += at91/
 obj-$(CONFIG_OPTEE) += optee/
+obj-$(CONFIG_BUILD_ASN1) += asn1_decoder.o
 
 obj-$(CONFIG_AES) += aes.o
 
lib/asn1_decoder.c new file +527 -0

527 lines not on this page · raw message

Signature · 1 line
2.21.0
#19 lore raw

PATCH 6/11 lib: add asn1 decoder

Quoted 7 lines · AKASHI Takahiro

Imported from linux kernel v5.3.

---
  lib/Kconfig        |   6 +
  lib/Makefile       |   1 +
  lib/asn1_decoder.c | 527 +++++++++++++++++++++++++++++++++++++++++++++

Except for adjustments of include files lib/asn1_decoder.c is a verbatime copy of the Linux kernel file. OK.

Quoted 8 lines
  3 files changed, 534 insertions(+)

create mode 100644 lib/asn1_decoder.c

lib/Kconfig +0 -0
index 3da45a5ec322..26c94f49ecd2 100644

lib/Kconfig config SMBIOS_PRODUCT_NAME · line 526

Quoted 6 lines
lib/Kconfig +3 -0
@@ -521,6 +521,12 @@ config SMBIOS_PRODUCT_NAME
 
  endmenu
 
+config ASN1
+	bool
+	select BUILD_ASN1

BUILD_ASN1 does not really fit into cmd/Kconfig.

Should we move it here?

Why do we need two separate flags ASN1 and BUILD_ASN1?

Best regards

Heinrich

Quoted 552 lines
lib/Kconfig +3 -0
+	help
+	  Enable asn1 decoder library.
+
  source lib/efi/Kconfig
  source lib/efi_loader/Kconfig
  source lib/optee/Kconfig
lib/Makefile +1 -0
index 2fffd68f943c..eb3a675fb8c2 100644
@@ -17,6 +17,7 @@ obj-$(CONFIG_OF_LIVE) += of_live.o
  obj-$(CONFIG_CMD_DHRYSTONE) += dhry/
  obj-$(CONFIG_ARCH_AT91) += at91/
  obj-$(CONFIG_OPTEE) += optee/
+obj-$(CONFIG_BUILD_ASN1) += asn1_decoder.o
 
  obj-$(CONFIG_AES) += aes.o
 
lib/asn1_decoder.c new file +527 -0
new file mode 100644
index 000000000000..db222625dd0f
@@ -0,0 +1,527 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* Decoder for ASN.1 BER/DER/CER encoded bytestream
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifdef __UBOOT__
+#include <linux/compat.h>
+#else
+#include <linux/export.h>
+#endif
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#ifndef __UBOOT__
+#include <linux/module.h>
+#endif
+#include <linux/asn1_decoder.h>
+#include <linux/asn1_ber_bytecode.h>
+
+static const unsigned char asn1_op_lengths[ASN1_OP__NR] = {
+	/*					OPC TAG JMP ACT */
+	[ASN1_OP_MATCH]				= 1 + 1,
+	[ASN1_OP_MATCH_OR_SKIP]			= 1 + 1,
+	[ASN1_OP_MATCH_ACT]			= 1 + 1     + 1,
+	[ASN1_OP_MATCH_ACT_OR_SKIP]		= 1 + 1     + 1,
+	[ASN1_OP_MATCH_JUMP]			= 1 + 1 + 1,
+	[ASN1_OP_MATCH_JUMP_OR_SKIP]		= 1 + 1 + 1,
+	[ASN1_OP_MATCH_ANY]			= 1,
+	[ASN1_OP_MATCH_ANY_OR_SKIP]		= 1,
+	[ASN1_OP_MATCH_ANY_ACT]			= 1         + 1,
+	[ASN1_OP_MATCH_ANY_ACT_OR_SKIP]		= 1         + 1,
+	[ASN1_OP_COND_MATCH_OR_SKIP]		= 1 + 1,
+	[ASN1_OP_COND_MATCH_ACT_OR_SKIP]	= 1 + 1     + 1,
+	[ASN1_OP_COND_MATCH_JUMP_OR_SKIP]	= 1 + 1 + 1,
+	[ASN1_OP_COND_MATCH_ANY]		= 1,
+	[ASN1_OP_COND_MATCH_ANY_OR_SKIP]	= 1,
+	[ASN1_OP_COND_MATCH_ANY_ACT]		= 1         + 1,
+	[ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP]	= 1         + 1,
+	[ASN1_OP_COND_FAIL]			= 1,
+	[ASN1_OP_COMPLETE]			= 1,
+	[ASN1_OP_ACT]				= 1         + 1,
+	[ASN1_OP_MAYBE_ACT]			= 1         + 1,
+	[ASN1_OP_RETURN]			= 1,
+	[ASN1_OP_END_SEQ]			= 1,
+	[ASN1_OP_END_SEQ_OF]			= 1     + 1,
+	[ASN1_OP_END_SET]			= 1,
+	[ASN1_OP_END_SET_OF]			= 1     + 1,
+	[ASN1_OP_END_SEQ_ACT]			= 1         + 1,
+	[ASN1_OP_END_SEQ_OF_ACT]		= 1     + 1 + 1,
+	[ASN1_OP_END_SET_ACT]			= 1         + 1,
+	[ASN1_OP_END_SET_OF_ACT]		= 1     + 1 + 1,
+};
+
+/*
+ * Find the length of an indefinite length object
+ * @data: The data buffer
+ * @datalen: The end of the innermost containing element in the buffer
+ * @_dp: The data parse cursor (updated before returning)
+ * @_len: Where to return the size of the element.
+ * @_errmsg: Where to return a pointer to an error message on error
+ */
+static int asn1_find_indefinite_length(const unsigned char *data, size_t datalen,
+				       size_t *_dp, size_t *_len,
+				       const char **_errmsg)
+{
+	unsigned char tag, tmp;
+	size_t dp = *_dp, len, n;
+	int indef_level = 1;
+
+next_tag:
+	if (unlikely(datalen - dp < 2)) {
+		if (datalen == dp)
+			goto missing_eoc;
+		goto data_overrun_error;
+	}
+
+	/* Extract a tag from the data */
+	tag = data[dp++];
+	if (tag == ASN1_EOC) {
+		/* It appears to be an EOC. */
+		if (data[dp++] != 0)
+			goto invalid_eoc;
+		if (--indef_level <= 0) {
+			*_len = dp - *_dp;
+			*_dp = dp;
+			return 0;
+		}
+		goto next_tag;
+	}
+
+	if (unlikely((tag & 0x1f) == ASN1_LONG_TAG)) {
+		do {
+			if (unlikely(datalen - dp < 2))
+				goto data_overrun_error;
+			tmp = data[dp++];
+		} while (tmp & 0x80);
+	}
+
+	/* Extract the length */
+	len = data[dp++];
+	if (len <= 0x7f)
+		goto check_length;
+
+	if (unlikely(len == ASN1_INDEFINITE_LENGTH)) {
+		/* Indefinite length */
+		if (unlikely((tag & ASN1_CONS_BIT) == ASN1_PRIM << 5))
+			goto indefinite_len_primitive;
+		indef_level++;
+		goto next_tag;
+	}
+
+	n = len - 0x80;
+	if (unlikely(n > sizeof(len) - 1))
+		goto length_too_long;
+	if (unlikely(n > datalen - dp))
+		goto data_overrun_error;
+	len = 0;
+	for (; n > 0; n--) {
+		len <<= 8;
+		len |= data[dp++];
+	}
+check_length:
+	if (len > datalen - dp)
+		goto data_overrun_error;
+	dp += len;
+	goto next_tag;
+
+length_too_long:
+	*_errmsg = "Unsupported length";
+	goto error;
+indefinite_len_primitive:
+	*_errmsg = "Indefinite len primitive not permitted";
+	goto error;
+invalid_eoc:
+	*_errmsg = "Invalid length EOC";
+	goto error;
+data_overrun_error:
+	*_errmsg = "Data overrun error";
+	goto error;
+missing_eoc:
+	*_errmsg = "Missing EOC in indefinite len cons";
+error:
+	*_dp = dp;
+	return -1;
+}
+
+/**
+ * asn1_ber_decoder - Decoder BER/DER/CER ASN.1 according to pattern
+ * @decoder: The decoder definition (produced by asn1_compiler)
+ * @context: The caller's context (to be passed to the action functions)
+ * @data: The encoded data
+ * @datalen: The size of the encoded data
+ *
+ * Decode BER/DER/CER encoded ASN.1 data according to a bytecode pattern
+ * produced by asn1_compiler.  Action functions are called on marked tags to
+ * allow the caller to retrieve significant data.
+ *
+ * LIMITATIONS:
+ *
+ * To keep down the amount of stack used by this function, the following limits
+ * have been imposed:
+ *
+ *  (1) This won't handle datalen > 65535 without increasing the size of the
+ *	cons stack elements and length_too_long checking.
+ *
+ *  (2) The stack of constructed types is 10 deep.  If the depth of non-leaf
+ *	constructed types exceeds this, the decode will fail.
+ *
+ *  (3) The SET type (not the SET OF type) isn't really supported as tracking
+ *	what members of the set have been seen is a pain.
+ */
+int asn1_ber_decoder(const struct asn1_decoder *decoder,
+		     void *context,
+		     const unsigned char *data,
+		     size_t datalen)
+{
+	const unsigned char *machine = decoder->machine;
+	const asn1_action_t *actions = decoder->actions;
+	size_t machlen = decoder->machlen;
+	enum asn1_opcode op;
+	unsigned char tag = 0, csp = 0, jsp = 0, optag = 0, hdr = 0;
+	const char *errmsg;
+	size_t pc = 0, dp = 0, tdp = 0, len = 0;
+	int ret;
+
+	unsigned char flags = 0;
+#define FLAG_INDEFINITE_LENGTH	0x01
+#define FLAG_MATCHED		0x02
+#define FLAG_LAST_MATCHED	0x04 /* Last tag matched */
+#define FLAG_CONS		0x20 /* Corresponds to CONS bit in the opcode tag
+				      * - ie. whether or not we are going to parse
+				      *   a compound type.
+				      */
+
+#define NR_CONS_STACK 10
+	unsigned short cons_dp_stack[NR_CONS_STACK];
+	unsigned short cons_datalen_stack[NR_CONS_STACK];
+	unsigned char cons_hdrlen_stack[NR_CONS_STACK];
+#define NR_JUMP_STACK 10
+	unsigned char jump_stack[NR_JUMP_STACK];
+
+	if (datalen > 65535)
+		return -EMSGSIZE;
+
+next_op:
+	pr_debug("next_op: pc=\e[32m%zu\e[m/%zu dp=\e[33m%zu\e[m/%zu C=%d J=%d\n",
+		 pc, machlen, dp, datalen, csp, jsp);
+	if (unlikely(pc >= machlen))
+		goto machine_overrun_error;
+	op = machine[pc];
+	if (unlikely(pc + asn1_op_lengths[op] > machlen))
+		goto machine_overrun_error;
+
+	/* If this command is meant to match a tag, then do that before
+	 * evaluating the command.
+	 */
+	if (op <= ASN1_OP__MATCHES_TAG) {
+		unsigned char tmp;
+
+		/* Skip conditional matches if possible */
+		if ((op & ASN1_OP_MATCH__COND && flags & FLAG_MATCHED) ||
+		    (op & ASN1_OP_MATCH__SKIP && dp == datalen)) {
+			flags &= ~FLAG_LAST_MATCHED;
+			pc += asn1_op_lengths[op];
+			goto next_op;
+		}
+
+		flags = 0;
+		hdr = 2;
+
+		/* Extract a tag from the data */
+		if (unlikely(datalen - dp < 2))
+			goto data_overrun_error;
+		tag = data[dp++];
+		if (unlikely((tag & 0x1f) == ASN1_LONG_TAG))
+			goto long_tag_not_supported;
+
+		if (op & ASN1_OP_MATCH__ANY) {
+			pr_debug("- any %02x\n", tag);
+		} else {
+			/* Extract the tag from the machine
+			 * - Either CONS or PRIM are permitted in the data if
+			 *   CONS is not set in the op stream, otherwise CONS
+			 *   is mandatory.
+			 */
+			optag = machine[pc + 1];
+			flags |= optag & FLAG_CONS;
+
+			/* Determine whether the tag matched */
+			tmp = optag ^ tag;
+			tmp &= ~(optag & ASN1_CONS_BIT);
+			pr_debug("- match? %02x %02x %02x\n", tag, optag, tmp);
+			if (tmp != 0) {
+				/* All odd-numbered tags are MATCH_OR_SKIP. */
+				if (op & ASN1_OP_MATCH__SKIP) {
+					pc += asn1_op_lengths[op];
+					dp--;
+					goto next_op;
+				}
+				goto tag_mismatch;
+			}
+		}
+		flags |= FLAG_MATCHED;
+
+		len = data[dp++];
+		if (len > 0x7f) {
+			if (unlikely(len == ASN1_INDEFINITE_LENGTH)) {
+				/* Indefinite length */
+				if (unlikely(!(tag & ASN1_CONS_BIT)))
+					goto indefinite_len_primitive;
+				flags |= FLAG_INDEFINITE_LENGTH;
+				if (unlikely(2 > datalen - dp))
+					goto data_overrun_error;
+			} else {
+				int n = len - 0x80;
+				if (unlikely(n > 2))
+					goto length_too_long;
+				if (unlikely(n > datalen - dp))
+					goto data_overrun_error;
+				hdr += n;
+				for (len = 0; n > 0; n--) {
+					len <<= 8;
+					len |= data[dp++];
+				}
+				if (unlikely(len > datalen - dp))
+					goto data_overrun_error;
+			}
+		} else {
+			if (unlikely(len > datalen - dp))
+				goto data_overrun_error;
+		}
+
+		if (flags & FLAG_CONS) {
+			/* For expected compound forms, we stack the positions
+			 * of the start and end of the data.
+			 */
+			if (unlikely(csp >= NR_CONS_STACK))
+				goto cons_stack_overflow;
+			cons_dp_stack[csp] = dp;
+			cons_hdrlen_stack[csp] = hdr;
+			if (!(flags & FLAG_INDEFINITE_LENGTH)) {
+				cons_datalen_stack[csp] = datalen;
+				datalen = dp + len;
+			} else {
+				cons_datalen_stack[csp] = 0;
+			}
+			csp++;
+		}
+
+		pr_debug("- TAG: %02x %zu%s\n",
+			 tag, len, flags & FLAG_CONS ? " CONS" : "");
+		tdp = dp;
+	}
+
+	/* Decide how to handle the operation */
+	switch (op) {
+	case ASN1_OP_MATCH:
+	case ASN1_OP_MATCH_OR_SKIP:
+	case ASN1_OP_MATCH_ACT:
+	case ASN1_OP_MATCH_ACT_OR_SKIP:
+	case ASN1_OP_MATCH_ANY:
+	case ASN1_OP_MATCH_ANY_OR_SKIP:
+	case ASN1_OP_MATCH_ANY_ACT:
+	case ASN1_OP_MATCH_ANY_ACT_OR_SKIP:
+	case ASN1_OP_COND_MATCH_OR_SKIP:
+	case ASN1_OP_COND_MATCH_ACT_OR_SKIP:
+	case ASN1_OP_COND_MATCH_ANY:
+	case ASN1_OP_COND_MATCH_ANY_OR_SKIP:
+	case ASN1_OP_COND_MATCH_ANY_ACT:
+	case ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP:
+
+		if (!(flags & FLAG_CONS)) {
+			if (flags & FLAG_INDEFINITE_LENGTH) {
+				size_t tmp = dp;
+
+				ret = asn1_find_indefinite_length(
+					data, datalen, &tmp, &len, &errmsg);
+				if (ret < 0)
+					goto error;
+			}
+			pr_debug("- LEAF: %zu\n", len);
+		}
+
+		if (op & ASN1_OP_MATCH__ACT) {
+			unsigned char act;
+
+			if (op & ASN1_OP_MATCH__ANY)
+				act = machine[pc + 1];
+			else
+				act = machine[pc + 2];
+			ret = actions[act](context, hdr, tag, data + dp, len);
+			if (ret < 0)
+				return ret;
+		}
+
+		if (!(flags & FLAG_CONS))
+			dp += len;
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_MATCH_JUMP:
+	case ASN1_OP_MATCH_JUMP_OR_SKIP:
+	case ASN1_OP_COND_MATCH_JUMP_OR_SKIP:
+		pr_debug("- MATCH_JUMP\n");
+		if (unlikely(jsp == NR_JUMP_STACK))
+			goto jump_stack_overflow;
+		jump_stack[jsp++] = pc + asn1_op_lengths[op];
+		pc = machine[pc + 2];
+		goto next_op;
+
+	case ASN1_OP_COND_FAIL:
+		if (unlikely(!(flags & FLAG_MATCHED)))
+			goto tag_mismatch;
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_COMPLETE:
+		if (unlikely(jsp != 0 || csp != 0)) {
+			pr_err("ASN.1 decoder error: Stacks not empty at completion (%u, %u)\n",
+			       jsp, csp);
+			return -EBADMSG;
+		}
+		return 0;
+
+	case ASN1_OP_END_SET:
+	case ASN1_OP_END_SET_ACT:
+		if (unlikely(!(flags & FLAG_MATCHED)))
+			goto tag_mismatch;
+		/* fall through */
+
+	case ASN1_OP_END_SEQ:
+	case ASN1_OP_END_SET_OF:
+	case ASN1_OP_END_SEQ_OF:
+	case ASN1_OP_END_SEQ_ACT:
+	case ASN1_OP_END_SET_OF_ACT:
+	case ASN1_OP_END_SEQ_OF_ACT:
+		if (unlikely(csp <= 0))
+			goto cons_stack_underflow;
+		csp--;
+		tdp = cons_dp_stack[csp];
+		hdr = cons_hdrlen_stack[csp];
+		len = datalen;
+		datalen = cons_datalen_stack[csp];
+		pr_debug("- end cons t=%zu dp=%zu l=%zu/%zu\n",
+			 tdp, dp, len, datalen);
+		if (datalen == 0) {
+			/* Indefinite length - check for the EOC. */
+			datalen = len;
+			if (unlikely(datalen - dp < 2))
+				goto data_overrun_error;
+			if (data[dp++] != 0) {
+				if (op & ASN1_OP_END__OF) {
+					dp--;
+					csp++;
+					pc = machine[pc + 1];
+					pr_debug("- continue\n");
+					goto next_op;
+				}
+				goto missing_eoc;
+			}
+			if (data[dp++] != 0)
+				goto invalid_eoc;
+			len = dp - tdp - 2;
+		} else {
+			if (dp < len && (op & ASN1_OP_END__OF)) {
+				datalen = len;
+				csp++;
+				pc = machine[pc + 1];
+				pr_debug("- continue\n");
+				goto next_op;
+			}
+			if (dp != len)
+				goto cons_length_error;
+			len -= tdp;
+			pr_debug("- cons len l=%zu d=%zu\n", len, dp - tdp);
+		}
+
+		if (op & ASN1_OP_END__ACT) {
+			unsigned char act;
+			if (op & ASN1_OP_END__OF)
+				act = machine[pc + 2];
+			else
+				act = machine[pc + 1];
+			ret = actions[act](context, hdr, 0, data + tdp, len);
+			if (ret < 0)
+				return ret;
+		}
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_MAYBE_ACT:
+		if (!(flags & FLAG_LAST_MATCHED)) {
+			pc += asn1_op_lengths[op];
+			goto next_op;
+		}
+		/* fall through */
+
+	case ASN1_OP_ACT:
+		ret = actions[machine[pc + 1]](context, hdr, tag, data + tdp, len);
+		if (ret < 0)
+			return ret;
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_RETURN:
+		if (unlikely(jsp <= 0))
+			goto jump_stack_underflow;
+		pc = jump_stack[--jsp];
+		flags |= FLAG_MATCHED | FLAG_LAST_MATCHED;
+		goto next_op;
+
+	default:
+		break;
+	}
+
+	/* Shouldn't reach here */
+	pr_err("ASN.1 decoder error: Found reserved opcode (%u) pc=%zu\n",
+	       op, pc);
+	return -EBADMSG;
+
+data_overrun_error:
+	errmsg = "Data overrun error";
+	goto error;
+machine_overrun_error:
+	errmsg = "Machine overrun error";
+	goto error;
+jump_stack_underflow:
+	errmsg = "Jump stack underflow";
+	goto error;
+jump_stack_overflow:
+	errmsg = "Jump stack overflow";
+	goto error;
+cons_stack_underflow:
+	errmsg = "Cons stack underflow";
+	goto error;
+cons_stack_overflow:
+	errmsg = "Cons stack overflow";
+	goto error;
+cons_length_error:
+	errmsg = "Cons length error";
+	goto error;
+missing_eoc:
+	errmsg = "Missing EOC in indefinite len cons";
+	goto error;
+invalid_eoc:
+	errmsg = "Invalid length EOC";
+	goto error;
+length_too_long:
+	errmsg = "Unsupported length";
+	goto error;
+indefinite_len_primitive:
+	errmsg = "Indefinite len primitive not permitted";
+	goto error;
+tag_mismatch:
+	errmsg = "Unexpected tag";
+	goto error;
+long_tag_not_supported:
+	errmsg = "Long tag not supported";
+error:
+	pr_debug("\nASN1: %s [m=%zu d=%zu ot=%02x t=%02x l=%zu]\n",
+		 errmsg, pc, dp, optag, tag, len);
+	return -EBADMSG;
+}
+EXPORT_SYMBOL_GPL(asn1_ber_decoder);
+
+MODULE_LICENSE("GPL");
#20 lore raw

PATCH 6/11 lib: add asn1 decoder

Quoted 21 lines · Heinrich Schuchardt
Quoted 7 lines · AKASHI Takahiro

Imported from linux kernel v5.3.

---
 lib/Kconfig        |   6 +
 lib/Makefile       |   1 +
 lib/asn1_decoder.c | 527 +++++++++++++++++++++++++++++++++++++++++++++

Except for adjustments of include files lib/asn1_decoder.c is a verbatime copy of the Linux kernel file. OK.

Quoted 9 lines
 3 files changed, 534 insertions(+)
 create mode 100644 lib/asn1_decoder.c
lib/Kconfig +0 -0
index 3da45a5ec322..26c94f49ecd2 100644
@@ -521,6 +521,12 @@ config SMBIOS_PRODUCT_NAME
 
Quoted 7 lines

lib/Kconfig config SMBIOS_PRODUCT_NAME · line 526

Quoted 5 lines
lib/Kconfig +3 -0
 endmenu
 
+config ASN1
+	bool
+	select BUILD_ASN1

BUILD_ASN1 does not really fit into cmd/Kconfig.

I just follows the way of BUILD_BIN2C, but

Quoted 1 line

Should we move it here?

yeah, bin2c is only used under cmd and so moving BUILD_ASN1 to lib/ is fine to me.

Quoted 1 line

Why do we need two separate flags ASN1 and BUILD_ASN1?

My intension here is that CONFIG_ASN1 is for asn1_decoder.o and CONFIG_BUILD_ASN1 is for asn1 compiler.
(lib/Makefile is wrong in this sense.)
I believe that having different config's here makes sense as those two targets are closely related but different objects.

So I will change their names to clarify this.

-Takahiro Akashi

Quoted 557 lines

Best regards

Heinrich

Quoted 552 lines
lib/Kconfig +3 -0
+	help
+	  Enable asn1 decoder library.
+
 source lib/efi/Kconfig
 source lib/efi_loader/Kconfig
 source lib/optee/Kconfig
lib/Makefile +1 -0
index 2fffd68f943c..eb3a675fb8c2 100644
@@ -17,6 +17,7 @@ obj-$(CONFIG_OF_LIVE) += of_live.o
 obj-$(CONFIG_CMD_DHRYSTONE) += dhry/
 obj-$(CONFIG_ARCH_AT91) += at91/
 obj-$(CONFIG_OPTEE) += optee/
+obj-$(CONFIG_BUILD_ASN1) += asn1_decoder.o
 
 obj-$(CONFIG_AES) += aes.o
 
lib/asn1_decoder.c new file +527 -0
new file mode 100644
index 000000000000..db222625dd0f
@@ -0,0 +1,527 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* Decoder for ASN.1 BER/DER/CER encoded bytestream
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifdef __UBOOT__
+#include <linux/compat.h>
+#else
+#include <linux/export.h>
+#endif
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#ifndef __UBOOT__
+#include <linux/module.h>
+#endif
+#include <linux/asn1_decoder.h>
+#include <linux/asn1_ber_bytecode.h>
+
+static const unsigned char asn1_op_lengths[ASN1_OP__NR] = {
+	/*					OPC TAG JMP ACT */
+	[ASN1_OP_MATCH]				= 1 + 1,
+	[ASN1_OP_MATCH_OR_SKIP]			= 1 + 1,
+	[ASN1_OP_MATCH_ACT]			= 1 + 1     + 1,
+	[ASN1_OP_MATCH_ACT_OR_SKIP]		= 1 + 1     + 1,
+	[ASN1_OP_MATCH_JUMP]			= 1 + 1 + 1,
+	[ASN1_OP_MATCH_JUMP_OR_SKIP]		= 1 + 1 + 1,
+	[ASN1_OP_MATCH_ANY]			= 1,
+	[ASN1_OP_MATCH_ANY_OR_SKIP]		= 1,
+	[ASN1_OP_MATCH_ANY_ACT]			= 1         + 1,
+	[ASN1_OP_MATCH_ANY_ACT_OR_SKIP]		= 1         + 1,
+	[ASN1_OP_COND_MATCH_OR_SKIP]		= 1 + 1,
+	[ASN1_OP_COND_MATCH_ACT_OR_SKIP]	= 1 + 1     + 1,
+	[ASN1_OP_COND_MATCH_JUMP_OR_SKIP]	= 1 + 1 + 1,
+	[ASN1_OP_COND_MATCH_ANY]		= 1,
+	[ASN1_OP_COND_MATCH_ANY_OR_SKIP]	= 1,
+	[ASN1_OP_COND_MATCH_ANY_ACT]		= 1         + 1,
+	[ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP]	= 1         + 1,
+	[ASN1_OP_COND_FAIL]			= 1,
+	[ASN1_OP_COMPLETE]			= 1,
+	[ASN1_OP_ACT]				= 1         + 1,
+	[ASN1_OP_MAYBE_ACT]			= 1         + 1,
+	[ASN1_OP_RETURN]			= 1,
+	[ASN1_OP_END_SEQ]			= 1,
+	[ASN1_OP_END_SEQ_OF]			= 1     + 1,
+	[ASN1_OP_END_SET]			= 1,
+	[ASN1_OP_END_SET_OF]			= 1     + 1,
+	[ASN1_OP_END_SEQ_ACT]			= 1         + 1,
+	[ASN1_OP_END_SEQ_OF_ACT]		= 1     + 1 + 1,
+	[ASN1_OP_END_SET_ACT]			= 1         + 1,
+	[ASN1_OP_END_SET_OF_ACT]		= 1     + 1 + 1,
+};
+
+/*
+ * Find the length of an indefinite length object
+ * @data: The data buffer
+ * @datalen: The end of the innermost containing element in the buffer
+ * @_dp: The data parse cursor (updated before returning)
+ * @_len: Where to return the size of the element.
+ * @_errmsg: Where to return a pointer to an error message on error
+ */
+static int asn1_find_indefinite_length(const unsigned char *data, size_t datalen,
+				       size_t *_dp, size_t *_len,
+				       const char **_errmsg)
+{
+	unsigned char tag, tmp;
+	size_t dp = *_dp, len, n;
+	int indef_level = 1;
+
+next_tag:
+	if (unlikely(datalen - dp < 2)) {
+		if (datalen == dp)
+			goto missing_eoc;
+		goto data_overrun_error;
+	}
+
+	/* Extract a tag from the data */
+	tag = data[dp++];
+	if (tag == ASN1_EOC) {
+		/* It appears to be an EOC. */
+		if (data[dp++] != 0)
+			goto invalid_eoc;
+		if (--indef_level <= 0) {
+			*_len = dp - *_dp;
+			*_dp = dp;
+			return 0;
+		}
+		goto next_tag;
+	}
+
+	if (unlikely((tag & 0x1f) == ASN1_LONG_TAG)) {
+		do {
+			if (unlikely(datalen - dp < 2))
+				goto data_overrun_error;
+			tmp = data[dp++];
+		} while (tmp & 0x80);
+	}
+
+	/* Extract the length */
+	len = data[dp++];
+	if (len <= 0x7f)
+		goto check_length;
+
+	if (unlikely(len == ASN1_INDEFINITE_LENGTH)) {
+		/* Indefinite length */
+		if (unlikely((tag & ASN1_CONS_BIT) == ASN1_PRIM << 5))
+			goto indefinite_len_primitive;
+		indef_level++;
+		goto next_tag;
+	}
+
+	n = len - 0x80;
+	if (unlikely(n > sizeof(len) - 1))
+		goto length_too_long;
+	if (unlikely(n > datalen - dp))
+		goto data_overrun_error;
+	len = 0;
+	for (; n > 0; n--) {
+		len <<= 8;
+		len |= data[dp++];
+	}
+check_length:
+	if (len > datalen - dp)
+		goto data_overrun_error;
+	dp += len;
+	goto next_tag;
+
+length_too_long:
+	*_errmsg = "Unsupported length";
+	goto error;
+indefinite_len_primitive:
+	*_errmsg = "Indefinite len primitive not permitted";
+	goto error;
+invalid_eoc:
+	*_errmsg = "Invalid length EOC";
+	goto error;
+data_overrun_error:
+	*_errmsg = "Data overrun error";
+	goto error;
+missing_eoc:
+	*_errmsg = "Missing EOC in indefinite len cons";
+error:
+	*_dp = dp;
+	return -1;
+}
+
+/**
+ * asn1_ber_decoder - Decoder BER/DER/CER ASN.1 according to pattern
+ * @decoder: The decoder definition (produced by asn1_compiler)
+ * @context: The caller's context (to be passed to the action functions)
+ * @data: The encoded data
+ * @datalen: The size of the encoded data
+ *
+ * Decode BER/DER/CER encoded ASN.1 data according to a bytecode pattern
+ * produced by asn1_compiler.  Action functions are called on marked tags to
+ * allow the caller to retrieve significant data.
+ *
+ * LIMITATIONS:
+ *
+ * To keep down the amount of stack used by this function, the following limits
+ * have been imposed:
+ *
+ *  (1) This won't handle datalen > 65535 without increasing the size of the
+ *	cons stack elements and length_too_long checking.
+ *
+ *  (2) The stack of constructed types is 10 deep.  If the depth of non-leaf
+ *	constructed types exceeds this, the decode will fail.
+ *
+ *  (3) The SET type (not the SET OF type) isn't really supported as tracking
+ *	what members of the set have been seen is a pain.
+ */
+int asn1_ber_decoder(const struct asn1_decoder *decoder,
+		     void *context,
+		     const unsigned char *data,
+		     size_t datalen)
+{
+	const unsigned char *machine = decoder->machine;
+	const asn1_action_t *actions = decoder->actions;
+	size_t machlen = decoder->machlen;
+	enum asn1_opcode op;
+	unsigned char tag = 0, csp = 0, jsp = 0, optag = 0, hdr = 0;
+	const char *errmsg;
+	size_t pc = 0, dp = 0, tdp = 0, len = 0;
+	int ret;
+
+	unsigned char flags = 0;
+#define FLAG_INDEFINITE_LENGTH	0x01
+#define FLAG_MATCHED		0x02
+#define FLAG_LAST_MATCHED	0x04 /* Last tag matched */
+#define FLAG_CONS		0x20 /* Corresponds to CONS bit in the opcode tag
+				      * - ie. whether or not we are going to parse
+				      *   a compound type.
+				      */
+
+#define NR_CONS_STACK 10
+	unsigned short cons_dp_stack[NR_CONS_STACK];
+	unsigned short cons_datalen_stack[NR_CONS_STACK];
+	unsigned char cons_hdrlen_stack[NR_CONS_STACK];
+#define NR_JUMP_STACK 10
+	unsigned char jump_stack[NR_JUMP_STACK];
+
+	if (datalen > 65535)
+		return -EMSGSIZE;
+
+next_op:
+	pr_debug("next_op: pc=\e[32m%zu\e[m/%zu dp=\e[33m%zu\e[m/%zu C=%d J=%d\n",
+		 pc, machlen, dp, datalen, csp, jsp);
+	if (unlikely(pc >= machlen))
+		goto machine_overrun_error;
+	op = machine[pc];
+	if (unlikely(pc + asn1_op_lengths[op] > machlen))
+		goto machine_overrun_error;
+
+	/* If this command is meant to match a tag, then do that before
+	 * evaluating the command.
+	 */
+	if (op <= ASN1_OP__MATCHES_TAG) {
+		unsigned char tmp;
+
+		/* Skip conditional matches if possible */
+		if ((op & ASN1_OP_MATCH__COND && flags & FLAG_MATCHED) ||
+		    (op & ASN1_OP_MATCH__SKIP && dp == datalen)) {
+			flags &= ~FLAG_LAST_MATCHED;
+			pc += asn1_op_lengths[op];
+			goto next_op;
+		}
+
+		flags = 0;
+		hdr = 2;
+
+		/* Extract a tag from the data */
+		if (unlikely(datalen - dp < 2))
+			goto data_overrun_error;
+		tag = data[dp++];
+		if (unlikely((tag & 0x1f) == ASN1_LONG_TAG))
+			goto long_tag_not_supported;
+
+		if (op & ASN1_OP_MATCH__ANY) {
+			pr_debug("- any %02x\n", tag);
+		} else {
+			/* Extract the tag from the machine
+			 * - Either CONS or PRIM are permitted in the data if
+			 *   CONS is not set in the op stream, otherwise CONS
+			 *   is mandatory.
+			 */
+			optag = machine[pc + 1];
+			flags |= optag & FLAG_CONS;
+
+			/* Determine whether the tag matched */
+			tmp = optag ^ tag;
+			tmp &= ~(optag & ASN1_CONS_BIT);
+			pr_debug("- match? %02x %02x %02x\n", tag, optag, tmp);
+			if (tmp != 0) {
+				/* All odd-numbered tags are MATCH_OR_SKIP. */
+				if (op & ASN1_OP_MATCH__SKIP) {
+					pc += asn1_op_lengths[op];
+					dp--;
+					goto next_op;
+				}
+				goto tag_mismatch;
+			}
+		}
+		flags |= FLAG_MATCHED;
+
+		len = data[dp++];
+		if (len > 0x7f) {
+			if (unlikely(len == ASN1_INDEFINITE_LENGTH)) {
+				/* Indefinite length */
+				if (unlikely(!(tag & ASN1_CONS_BIT)))
+					goto indefinite_len_primitive;
+				flags |= FLAG_INDEFINITE_LENGTH;
+				if (unlikely(2 > datalen - dp))
+					goto data_overrun_error;
+			} else {
+				int n = len - 0x80;
+				if (unlikely(n > 2))
+					goto length_too_long;
+				if (unlikely(n > datalen - dp))
+					goto data_overrun_error;
+				hdr += n;
+				for (len = 0; n > 0; n--) {
+					len <<= 8;
+					len |= data[dp++];
+				}
+				if (unlikely(len > datalen - dp))
+					goto data_overrun_error;
+			}
+		} else {
+			if (unlikely(len > datalen - dp))
+				goto data_overrun_error;
+		}
+
+		if (flags & FLAG_CONS) {
+			/* For expected compound forms, we stack the positions
+			 * of the start and end of the data.
+			 */
+			if (unlikely(csp >= NR_CONS_STACK))
+				goto cons_stack_overflow;
+			cons_dp_stack[csp] = dp;
+			cons_hdrlen_stack[csp] = hdr;
+			if (!(flags & FLAG_INDEFINITE_LENGTH)) {
+				cons_datalen_stack[csp] = datalen;
+				datalen = dp + len;
+			} else {
+				cons_datalen_stack[csp] = 0;
+			}
+			csp++;
+		}
+
+		pr_debug("- TAG: %02x %zu%s\n",
+			 tag, len, flags & FLAG_CONS ? " CONS" : "");
+		tdp = dp;
+	}
+
+	/* Decide how to handle the operation */
+	switch (op) {
+	case ASN1_OP_MATCH:
+	case ASN1_OP_MATCH_OR_SKIP:
+	case ASN1_OP_MATCH_ACT:
+	case ASN1_OP_MATCH_ACT_OR_SKIP:
+	case ASN1_OP_MATCH_ANY:
+	case ASN1_OP_MATCH_ANY_OR_SKIP:
+	case ASN1_OP_MATCH_ANY_ACT:
+	case ASN1_OP_MATCH_ANY_ACT_OR_SKIP:
+	case ASN1_OP_COND_MATCH_OR_SKIP:
+	case ASN1_OP_COND_MATCH_ACT_OR_SKIP:
+	case ASN1_OP_COND_MATCH_ANY:
+	case ASN1_OP_COND_MATCH_ANY_OR_SKIP:
+	case ASN1_OP_COND_MATCH_ANY_ACT:
+	case ASN1_OP_COND_MATCH_ANY_ACT_OR_SKIP:
+
+		if (!(flags & FLAG_CONS)) {
+			if (flags & FLAG_INDEFINITE_LENGTH) {
+				size_t tmp = dp;
+
+				ret = asn1_find_indefinite_length(
+					data, datalen, &tmp, &len, &errmsg);
+				if (ret < 0)
+					goto error;
+			}
+			pr_debug("- LEAF: %zu\n", len);
+		}
+
+		if (op & ASN1_OP_MATCH__ACT) {
+			unsigned char act;
+
+			if (op & ASN1_OP_MATCH__ANY)
+				act = machine[pc + 1];
+			else
+				act = machine[pc + 2];
+			ret = actions[act](context, hdr, tag, data + dp, len);
+			if (ret < 0)
+				return ret;
+		}
+
+		if (!(flags & FLAG_CONS))
+			dp += len;
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_MATCH_JUMP:
+	case ASN1_OP_MATCH_JUMP_OR_SKIP:
+	case ASN1_OP_COND_MATCH_JUMP_OR_SKIP:
+		pr_debug("- MATCH_JUMP\n");
+		if (unlikely(jsp == NR_JUMP_STACK))
+			goto jump_stack_overflow;
+		jump_stack[jsp++] = pc + asn1_op_lengths[op];
+		pc = machine[pc + 2];
+		goto next_op;
+
+	case ASN1_OP_COND_FAIL:
+		if (unlikely(!(flags & FLAG_MATCHED)))
+			goto tag_mismatch;
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_COMPLETE:
+		if (unlikely(jsp != 0 || csp != 0)) {
+			pr_err("ASN.1 decoder error: Stacks not empty at completion (%u, %u)\n",
+			       jsp, csp);
+			return -EBADMSG;
+		}
+		return 0;
+
+	case ASN1_OP_END_SET:
+	case ASN1_OP_END_SET_ACT:
+		if (unlikely(!(flags & FLAG_MATCHED)))
+			goto tag_mismatch;
+		/* fall through */
+
+	case ASN1_OP_END_SEQ:
+	case ASN1_OP_END_SET_OF:
+	case ASN1_OP_END_SEQ_OF:
+	case ASN1_OP_END_SEQ_ACT:
+	case ASN1_OP_END_SET_OF_ACT:
+	case ASN1_OP_END_SEQ_OF_ACT:
+		if (unlikely(csp <= 0))
+			goto cons_stack_underflow;
+		csp--;
+		tdp = cons_dp_stack[csp];
+		hdr = cons_hdrlen_stack[csp];
+		len = datalen;
+		datalen = cons_datalen_stack[csp];
+		pr_debug("- end cons t=%zu dp=%zu l=%zu/%zu\n",
+			 tdp, dp, len, datalen);
+		if (datalen == 0) {
+			/* Indefinite length - check for the EOC. */
+			datalen = len;
+			if (unlikely(datalen - dp < 2))
+				goto data_overrun_error;
+			if (data[dp++] != 0) {
+				if (op & ASN1_OP_END__OF) {
+					dp--;
+					csp++;
+					pc = machine[pc + 1];
+					pr_debug("- continue\n");
+					goto next_op;
+				}
+				goto missing_eoc;
+			}
+			if (data[dp++] != 0)
+				goto invalid_eoc;
+			len = dp - tdp - 2;
+		} else {
+			if (dp < len && (op & ASN1_OP_END__OF)) {
+				datalen = len;
+				csp++;
+				pc = machine[pc + 1];
+				pr_debug("- continue\n");
+				goto next_op;
+			}
+			if (dp != len)
+				goto cons_length_error;
+			len -= tdp;
+			pr_debug("- cons len l=%zu d=%zu\n", len, dp - tdp);
+		}
+
+		if (op & ASN1_OP_END__ACT) {
+			unsigned char act;
+			if (op & ASN1_OP_END__OF)
+				act = machine[pc + 2];
+			else
+				act = machine[pc + 1];
+			ret = actions[act](context, hdr, 0, data + tdp, len);
+			if (ret < 0)
+				return ret;
+		}
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_MAYBE_ACT:
+		if (!(flags & FLAG_LAST_MATCHED)) {
+			pc += asn1_op_lengths[op];
+			goto next_op;
+		}
+		/* fall through */
+
+	case ASN1_OP_ACT:
+		ret = actions[machine[pc + 1]](context, hdr, tag, data + tdp, len);
+		if (ret < 0)
+			return ret;
+		pc += asn1_op_lengths[op];
+		goto next_op;
+
+	case ASN1_OP_RETURN:
+		if (unlikely(jsp <= 0))
+			goto jump_stack_underflow;
+		pc = jump_stack[--jsp];
+		flags |= FLAG_MATCHED | FLAG_LAST_MATCHED;
+		goto next_op;
+
+	default:
+		break;
+	}
+
+	/* Shouldn't reach here */
+	pr_err("ASN.1 decoder error: Found reserved opcode (%u) pc=%zu\n",
+	       op, pc);
+	return -EBADMSG;
+
+data_overrun_error:
+	errmsg = "Data overrun error";
+	goto error;
+machine_overrun_error:
+	errmsg = "Machine overrun error";
+	goto error;
+jump_stack_underflow:
+	errmsg = "Jump stack underflow";
+	goto error;
+jump_stack_overflow:
+	errmsg = "Jump stack overflow";
+	goto error;
+cons_stack_underflow:
+	errmsg = "Cons stack underflow";
+	goto error;
+cons_stack_overflow:
+	errmsg = "Cons stack overflow";
+	goto error;
+cons_length_error:
+	errmsg = "Cons length error";
+	goto error;
+missing_eoc:
+	errmsg = "Missing EOC in indefinite len cons";
+	goto error;
+invalid_eoc:
+	errmsg = "Invalid length EOC";
+	goto error;
+length_too_long:
+	errmsg = "Unsupported length";
+	goto error;
+indefinite_len_primitive:
+	errmsg = "Indefinite len primitive not permitted";
+	goto error;
+tag_mismatch:
+	errmsg = "Unexpected tag";
+	goto error;
+long_tag_not_supported:
+	errmsg = "Long tag not supported";
+error:
+	pr_debug("\nASN1: %s [m=%zu d=%zu ot=%02x t=%02x l=%zu]\n",
+		 errmsg, pc, dp, optag, tag, len);
+	return -EBADMSG;
+}
+EXPORT_SYMBOL_GPL(asn1_ber_decoder);
+
+MODULE_LICENSE("GPL");
#21 lore raw

PATCH 7/11 lib: add oid registry utility

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

---
 include/linux/oid_registry.h | 117 ++++++++++++++++++++
 lib/Kconfig                  |   5 +
 lib/Makefile                 |  16 +++
 lib/build_OID_registry       | 203 +++++++++++++++++++++++++++++++++++
 lib/oid_registry.c           | 179 ++++++++++++++++++++++++++++++
 5 files changed, 520 insertions(+)
 create mode 100644 include/linux/oid_registry.h
 create mode 100755 lib/build_OID_registry
 create mode 100644 lib/oid_registry.c
include/linux/oid_registry.h new file +117 -0
new file mode 100644
index 000000000000..657d6bf2c064
@@ -0,0 +1,117 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 Object identifier (OID) registry
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_OID_REGISTRY_H
+#define _LINUX_OID_REGISTRY_H
+
+#include <linux/types.h>
+
+/*
+ * OIDs are turned into these values if possible, or OID__NR if not held here.
+ *
+ * NOTE!  Do not mess with the format of each line as this is read by
+ *	  build_OID_registry.pl to generate the data for look_up_OID().
+ */
+enum OID {
+	OID_id_dsa_with_sha1,		/* 1.2.840.10030.4.3 */
+	OID_id_dsa,			/* 1.2.840.10040.4.1 */
+	OID_id_ecdsa_with_sha1,		/* 1.2.840.10045.4.1 */
+	OID_id_ecPublicKey,		/* 1.2.840.10045.2.1 */
+
+	/* PKCS#1 {iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs-1(1)} */
+	OID_rsaEncryption,		/* 1.2.840.113549.1.1.1 */
+	OID_md2WithRSAEncryption,	/* 1.2.840.113549.1.1.2 */
+	OID_md3WithRSAEncryption,	/* 1.2.840.113549.1.1.3 */
+	OID_md4WithRSAEncryption,	/* 1.2.840.113549.1.1.4 */
+	OID_sha1WithRSAEncryption,	/* 1.2.840.113549.1.1.5 */
+	OID_sha256WithRSAEncryption,	/* 1.2.840.113549.1.1.11 */
+	OID_sha384WithRSAEncryption,	/* 1.2.840.113549.1.1.12 */
+	OID_sha512WithRSAEncryption,	/* 1.2.840.113549.1.1.13 */
+	OID_sha224WithRSAEncryption,	/* 1.2.840.113549.1.1.14 */
+	/* PKCS#7 {iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs-7(7)} */
+	OID_data,			/* 1.2.840.113549.1.7.1 */
+	OID_signed_data,		/* 1.2.840.113549.1.7.2 */
+	/* PKCS#9 {iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs-9(9)} */
+	OID_email_address,		/* 1.2.840.113549.1.9.1 */
+	OID_contentType,		/* 1.2.840.113549.1.9.3 */
+	OID_messageDigest,		/* 1.2.840.113549.1.9.4 */
+	OID_signingTime,		/* 1.2.840.113549.1.9.5 */
+	OID_smimeCapabilites,		/* 1.2.840.113549.1.9.15 */
+	OID_smimeAuthenticatedAttrs,	/* 1.2.840.113549.1.9.16.2.11 */
+
+	/* {iso(1) member-body(2) us(840) rsadsi(113549) digestAlgorithm(2)} */
+	OID_md2,			/* 1.2.840.113549.2.2 */
+	OID_md4,			/* 1.2.840.113549.2.4 */
+	OID_md5,			/* 1.2.840.113549.2.5 */
+
+	/* Microsoft Authenticode & Software Publishing */
+	OID_msIndirectData,		/* 1.3.6.1.4.1.311.2.1.4 */
+	OID_msStatementType,		/* 1.3.6.1.4.1.311.2.1.11 */
+	OID_msSpOpusInfo,		/* 1.3.6.1.4.1.311.2.1.12 */
+	OID_msPeImageDataObjId,		/* 1.3.6.1.4.1.311.2.1.15 */
+	OID_msIndividualSPKeyPurpose,	/* 1.3.6.1.4.1.311.2.1.21 */
+	OID_msOutlookExpress,		/* 1.3.6.1.4.1.311.16.4 */
+
+	OID_certAuthInfoAccess,		/* 1.3.6.1.5.5.7.1.1 */
+	OID_sha1,			/* 1.3.14.3.2.26 */
+	OID_sha256,			/* 2.16.840.1.101.3.4.2.1 */
+	OID_sha384,			/* 2.16.840.1.101.3.4.2.2 */
+	OID_sha512,			/* 2.16.840.1.101.3.4.2.3 */
+	OID_sha224,			/* 2.16.840.1.101.3.4.2.4 */
+
+	/* Distinguished Name attribute IDs [RFC 2256] */
+	OID_commonName,			/* 2.5.4.3 */
+	OID_surname,			/* 2.5.4.4 */
+	OID_countryName,		/* 2.5.4.6 */
+	OID_locality,			/* 2.5.4.7 */
+	OID_stateOrProvinceName,	/* 2.5.4.8 */
+	OID_organizationName,		/* 2.5.4.10 */
+	OID_organizationUnitName,	/* 2.5.4.11 */
+	OID_title,			/* 2.5.4.12 */
+	OID_description,		/* 2.5.4.13 */
+	OID_name,			/* 2.5.4.41 */
+	OID_givenName,			/* 2.5.4.42 */
+	OID_initials,			/* 2.5.4.43 */
+	OID_generationalQualifier,	/* 2.5.4.44 */
+
+	/* Certificate extension IDs */
+	OID_subjectKeyIdentifier,	/* 2.5.29.14 */
+	OID_keyUsage,			/* 2.5.29.15 */
+	OID_subjectAltName,		/* 2.5.29.17 */
+	OID_issuerAltName,		/* 2.5.29.18 */
+	OID_basicConstraints,		/* 2.5.29.19 */
+	OID_crlDistributionPoints,	/* 2.5.29.31 */
+	OID_certPolicies,		/* 2.5.29.32 */
+	OID_authorityKeyIdentifier,	/* 2.5.29.35 */
+	OID_extKeyUsage,		/* 2.5.29.37 */
+
+	/* EC-RDSA */
+	OID_gostCPSignA,		/* 1.2.643.2.2.35.1 */
+	OID_gostCPSignB,		/* 1.2.643.2.2.35.2 */
+	OID_gostCPSignC,		/* 1.2.643.2.2.35.3 */
+	OID_gost2012PKey256,		/* 1.2.643.7.1.1.1.1 */
+	OID_gost2012PKey512,		/* 1.2.643.7.1.1.1.2 */
+	OID_gost2012Digest256,		/* 1.2.643.7.1.1.2.2 */
+	OID_gost2012Digest512,		/* 1.2.643.7.1.1.2.3 */
+	OID_gost2012Signature256,	/* 1.2.643.7.1.1.3.2 */
+	OID_gost2012Signature512,	/* 1.2.643.7.1.1.3.3 */
+	OID_gostTC26Sign256A,		/* 1.2.643.7.1.2.1.1.1 */
+	OID_gostTC26Sign256B,		/* 1.2.643.7.1.2.1.1.2 */
+	OID_gostTC26Sign256C,		/* 1.2.643.7.1.2.1.1.3 */
+	OID_gostTC26Sign256D,		/* 1.2.643.7.1.2.1.1.4 */
+	OID_gostTC26Sign512A,		/* 1.2.643.7.1.2.1.2.1 */
+	OID_gostTC26Sign512B,		/* 1.2.643.7.1.2.1.2.2 */
+	OID_gostTC26Sign512C,		/* 1.2.643.7.1.2.1.2.3 */
+
+	OID__NR
+};
+
+extern enum OID look_up_OID(const void *data, size_t datasize);
+extern int sprint_oid(const void *, size_t, char *, size_t);
+extern int sprint_OID(enum OID, char *, size_t);
+
+#endif /* _LINUX_OID_REGISTRY_H */
lib/Kconfig +5 -0
index 26c94f49ecd2..d2955c8feb0e 100644
@@ -527,6 +527,11 @@ config ASN1
 	help
 	  Enable asn1 decoder library.
 
+config OID_REGISTRY
+	bool
+	help
+	  Enable fast lookup object identifier registry.
+
 source lib/efi/Kconfig
 source lib/efi_loader/Kconfig
 source lib/optee/Kconfig
lib/Makefile +16 -0
index eb3a675fb8c2..677ec33a1ce1 100644
@@ -118,4 +118,20 @@ else
 obj-y += vsprintf.o strto.o strmhz.o
 endif
 
+#
+# Build a fast OID lookip registry from include/linux/oid_registry.h
+#
+obj-$(CONFIG_OID_REGISTRY) += oid_registry.o
+
+$(obj)/oid_registry.o: $(obj)/oid_registry_data.c
+
+$(obj)/oid_registry_data.c: $(srctree)/include/linux/oid_registry.h \
+			    $(src)/build_OID_registry
+	$(call cmd,build_OID_registry)
+
+quiet_cmd_build_OID_registry = GEN     $@
+      cmd_build_OID_registry = perl $(srctree)/$(src)/build_OID_registry $< $@
+
+clean-files     += oid_registry_data.c
+
 subdir-ccflags-$(CONFIG_CC_OPTIMIZE_LIBS_FOR_SPEED) += -O2
lib/build_OID_registry new file +203 -0

203 lines not on this page · raw message

lib/oid_registry.c new file +179 -0

179 lines not on this page · raw message

Signature · 1 line
2.21.0
#22 lore raw

PATCH 7/11 lib: add oid registry utility

Quoted 1 line · AKASHI Takahiro

Imported from linux kernel v5.3.

Please, indicate in the commit message which files are verbatim copies and which files had to be adjusted. E.g. in lib/oid_registry.c you adjusted the includes.

Quoted 7 lines
---
  include/linux/oid_registry.h | 117 ++++++++++++++++++++
  lib/Kconfig                  |   5 +
  lib/Makefile                 |  16 +++
  lib/build_OID_registry       | 203 +++++++++++++++++++++++++++++++++++

Shouldn't this file go into /scripts or /tools?

Best regards

Heinrich

Quoted 565 lines
  lib/oid_registry.c           | 179 ++++++++++++++++++++++++++++++
  5 files changed, 520 insertions(+)

create mode 100644 include/linux/oid_registry.h
create mode 100755 lib/build_OID_registry
create mode 100644 lib/oid_registry.c

include/linux/oid_registry.h new file +117 -0
new file mode 100644
index 000000000000..657d6bf2c064
@@ -0,0 +1,117 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ASN.1 Object identifier (OID) registry
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_OID_REGISTRY_H
+#define _LINUX_OID_REGISTRY_H
+
+#include <linux/types.h>
+
+/*
+ * OIDs are turned into these values if possible, or OID__NR if not held here.
+ *
+ * NOTE!  Do not mess with the format of each line as this is read by
+ *	  build_OID_registry.pl to generate the data for look_up_OID().
+ */
+enum OID {
+	OID_id_dsa_with_sha1,		/* 1.2.840.10030.4.3 */
+	OID_id_dsa,			/* 1.2.840.10040.4.1 */
+	OID_id_ecdsa_with_sha1,		/* 1.2.840.10045.4.1 */
+	OID_id_ecPublicKey,		/* 1.2.840.10045.2.1 */
+
+	/* PKCS#1 {iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs-1(1)} */
+	OID_rsaEncryption,		/* 1.2.840.113549.1.1.1 */
+	OID_md2WithRSAEncryption,	/* 1.2.840.113549.1.1.2 */
+	OID_md3WithRSAEncryption,	/* 1.2.840.113549.1.1.3 */
+	OID_md4WithRSAEncryption,	/* 1.2.840.113549.1.1.4 */
+	OID_sha1WithRSAEncryption,	/* 1.2.840.113549.1.1.5 */
+	OID_sha256WithRSAEncryption,	/* 1.2.840.113549.1.1.11 */
+	OID_sha384WithRSAEncryption,	/* 1.2.840.113549.1.1.12 */
+	OID_sha512WithRSAEncryption,	/* 1.2.840.113549.1.1.13 */
+	OID_sha224WithRSAEncryption,	/* 1.2.840.113549.1.1.14 */
+	/* PKCS#7 {iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs-7(7)} */
+	OID_data,			/* 1.2.840.113549.1.7.1 */
+	OID_signed_data,		/* 1.2.840.113549.1.7.2 */
+	/* PKCS#9 {iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs-9(9)} */
+	OID_email_address,		/* 1.2.840.113549.1.9.1 */
+	OID_contentType,		/* 1.2.840.113549.1.9.3 */
+	OID_messageDigest,		/* 1.2.840.113549.1.9.4 */
+	OID_signingTime,		/* 1.2.840.113549.1.9.5 */
+	OID_smimeCapabilites,		/* 1.2.840.113549.1.9.15 */
+	OID_smimeAuthenticatedAttrs,	/* 1.2.840.113549.1.9.16.2.11 */
+
+	/* {iso(1) member-body(2) us(840) rsadsi(113549) digestAlgorithm(2)} */
+	OID_md2,			/* 1.2.840.113549.2.2 */
+	OID_md4,			/* 1.2.840.113549.2.4 */
+	OID_md5,			/* 1.2.840.113549.2.5 */
+
+	/* Microsoft Authenticode & Software Publishing */
+	OID_msIndirectData,		/* 1.3.6.1.4.1.311.2.1.4 */
+	OID_msStatementType,		/* 1.3.6.1.4.1.311.2.1.11 */
+	OID_msSpOpusInfo,		/* 1.3.6.1.4.1.311.2.1.12 */
+	OID_msPeImageDataObjId,		/* 1.3.6.1.4.1.311.2.1.15 */
+	OID_msIndividualSPKeyPurpose,	/* 1.3.6.1.4.1.311.2.1.21 */
+	OID_msOutlookExpress,		/* 1.3.6.1.4.1.311.16.4 */
+
+	OID_certAuthInfoAccess,		/* 1.3.6.1.5.5.7.1.1 */
+	OID_sha1,			/* 1.3.14.3.2.26 */
+	OID_sha256,			/* 2.16.840.1.101.3.4.2.1 */
+	OID_sha384,			/* 2.16.840.1.101.3.4.2.2 */
+	OID_sha512,			/* 2.16.840.1.101.3.4.2.3 */
+	OID_sha224,			/* 2.16.840.1.101.3.4.2.4 */
+
+	/* Distinguished Name attribute IDs [RFC 2256] */
+	OID_commonName,			/* 2.5.4.3 */
+	OID_surname,			/* 2.5.4.4 */
+	OID_countryName,		/* 2.5.4.6 */
+	OID_locality,			/* 2.5.4.7 */
+	OID_stateOrProvinceName,	/* 2.5.4.8 */
+	OID_organizationName,		/* 2.5.4.10 */
+	OID_organizationUnitName,	/* 2.5.4.11 */
+	OID_title,			/* 2.5.4.12 */
+	OID_description,		/* 2.5.4.13 */
+	OID_name,			/* 2.5.4.41 */
+	OID_givenName,			/* 2.5.4.42 */
+	OID_initials,			/* 2.5.4.43 */
+	OID_generationalQualifier,	/* 2.5.4.44 */
+
+	/* Certificate extension IDs */
+	OID_subjectKeyIdentifier,	/* 2.5.29.14 */
+	OID_keyUsage,			/* 2.5.29.15 */
+	OID_subjectAltName,		/* 2.5.29.17 */
+	OID_issuerAltName,		/* 2.5.29.18 */
+	OID_basicConstraints,		/* 2.5.29.19 */
+	OID_crlDistributionPoints,	/* 2.5.29.31 */
+	OID_certPolicies,		/* 2.5.29.32 */
+	OID_authorityKeyIdentifier,	/* 2.5.29.35 */
+	OID_extKeyUsage,		/* 2.5.29.37 */
+
+	/* EC-RDSA */
+	OID_gostCPSignA,		/* 1.2.643.2.2.35.1 */
+	OID_gostCPSignB,		/* 1.2.643.2.2.35.2 */
+	OID_gostCPSignC,		/* 1.2.643.2.2.35.3 */
+	OID_gost2012PKey256,		/* 1.2.643.7.1.1.1.1 */
+	OID_gost2012PKey512,		/* 1.2.643.7.1.1.1.2 */
+	OID_gost2012Digest256,		/* 1.2.643.7.1.1.2.2 */
+	OID_gost2012Digest512,		/* 1.2.643.7.1.1.2.3 */
+	OID_gost2012Signature256,	/* 1.2.643.7.1.1.3.2 */
+	OID_gost2012Signature512,	/* 1.2.643.7.1.1.3.3 */
+	OID_gostTC26Sign256A,		/* 1.2.643.7.1.2.1.1.1 */
+	OID_gostTC26Sign256B,		/* 1.2.643.7.1.2.1.1.2 */
+	OID_gostTC26Sign256C,		/* 1.2.643.7.1.2.1.1.3 */
+	OID_gostTC26Sign256D,		/* 1.2.643.7.1.2.1.1.4 */
+	OID_gostTC26Sign512A,		/* 1.2.643.7.1.2.1.2.1 */
+	OID_gostTC26Sign512B,		/* 1.2.643.7.1.2.1.2.2 */
+	OID_gostTC26Sign512C,		/* 1.2.643.7.1.2.1.2.3 */
+
+	OID__NR
+};
+
+extern enum OID look_up_OID(const void *data, size_t datasize);
+extern int sprint_oid(const void *, size_t, char *, size_t);
+extern int sprint_OID(enum OID, char *, size_t);
+
+#endif /* _LINUX_OID_REGISTRY_H */
lib/Kconfig +5 -0
index 26c94f49ecd2..d2955c8feb0e 100644
@@ -527,6 +527,11 @@ config ASN1
  	help
  	  Enable asn1 decoder library.
 
+config OID_REGISTRY
+	bool
+	help
+	  Enable fast lookup object identifier registry.
+
  source lib/efi/Kconfig
  source lib/efi_loader/Kconfig
  source lib/optee/Kconfig
lib/Makefile +16 -0
index eb3a675fb8c2..677ec33a1ce1 100644
@@ -118,4 +118,20 @@ else
  obj-y += vsprintf.o strto.o strmhz.o
  endif
 
+#
+# Build a fast OID lookip registry from include/linux/oid_registry.h
+#
+obj-$(CONFIG_OID_REGISTRY) += oid_registry.o
+
+$(obj)/oid_registry.o: $(obj)/oid_registry_data.c
+
+$(obj)/oid_registry_data.c: $(srctree)/include/linux/oid_registry.h \
+			    $(src)/build_OID_registry
+	$(call cmd,build_OID_registry)
+
+quiet_cmd_build_OID_registry = GEN     $@
+      cmd_build_OID_registry = perl $(srctree)/$(src)/build_OID_registry $< $@
+
+clean-files     += oid_registry_data.c
+
  subdir-ccflags-$(CONFIG_CC_OPTIMIZE_LIBS_FOR_SPEED) += -O2
lib/build_OID_registry new file +203 -0
new file mode 100755
index 000000000000..d7fc32ea8ac2
@@ -0,0 +1,203 @@
+#!/usr/bin/perl -w
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# Build a static ASN.1 Object Identified (OID) registry
+#
+# Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+# Written by David Howells (dhowells at redhat.com)
+#
+
+use strict;
+
+my @names = ();
+my @oids = ();
+
+if ($#ARGV != 1) {
+    print STDERR "Format: ", $0, " <in-h-file> <out-c-file>\n";
+    exit(2);
+}
+
+#
+# Open the file to read from
+#
+open IN_FILE, "<$ARGV[0]" || die;
+while (<IN_FILE>) {
+    chomp;
+    if (m!\s+OID_([a-zA-z][a-zA-Z0-9_]+),\s+/[*]\s+([012][.0-9]*)\s+[*]/!) {
+	push @names, $1;
+	push @oids, $2;
+    }
+}
+close IN_FILE || die;
+
+#
+# Open the files to write into
+#
+open C_FILE, ">$ARGV[1]" or die;
+print C_FILE "/*\n";
+print C_FILE " * Automatically generated by ", $0, ".  Do not edit\n";
+print C_FILE " */\n";
+
+#
+# Split the data up into separate lists and also determine the lengths of the
+# encoded data arrays.
+#
+my @indices = ();
+my @lengths = ();
+my $total_length = 0;
+
+for (my $i = 0; $i <= $#names; $i++) {
+    my $name = $names[$i];
+    my $oid = $oids[$i];
+
+    my @components = split(/[.]/, $oid);
+
+    # Determine the encoded length of this OID
+    my $size = $#components;
+    for (my $loop = 2; $loop <= $#components; $loop++) {
+	my $c = $components[$loop];
+
+	# We will base128 encode the number
+	my $tmp = ($c == 0) ? 0 : int(log($c)/log(2));
+	$tmp = int($tmp / 7);
+	$size += $tmp;
+    }
+    push @lengths, $size;
+    push @indices, $total_length;
+    $total_length += $size;
+}
+
+#
+# Emit the look-up-by-OID index table
+#
+print C_FILE "\n";
+if ($total_length <= 255) {
+    print C_FILE "static const unsigned char oid_index[OID__NR + 1] = {\n";
+} else {
+    print C_FILE "static const unsigned short oid_index[OID__NR + 1] = {\n";
+}
+for (my $i = 0; $i <= $#names; $i++) {
+    print C_FILE "\t[OID_", $names[$i], "] = ", $indices[$i], ",\n"
+}
+print C_FILE "\t[OID__NR] = ", $total_length, "\n";
+print C_FILE "};\n";
+
+#
+# Encode the OIDs
+#
+my @encoded_oids = ();
+
+for (my $i = 0; $i <= $#names; $i++) {
+    my @octets = ();
+
+    my @components = split(/[.]/, $oids[$i]);
+
+    push @octets, $components[0] * 40 + $components[1];
+
+    for (my $loop = 2; $loop <= $#components; $loop++) {
+	my $c = $components[$loop];
+
+	# Base128 encode the number
+	my $tmp = ($c == 0) ? 0 : int(log($c)/log(2));
+	$tmp = int($tmp / 7);
+
+	for (; $tmp > 0; $tmp--) {
+	    push @octets, (($c >> $tmp * 7) & 0x7f) | 0x80;
+	}
+	push @octets, $c & 0x7f;
+    }
+
+    push @encoded_oids, \@octets;
+}
+
+#
+# Create a hash value for each OID
+#
+my @hash_values = ();
+for (my $i = 0; $i <= $#names; $i++) {
+    my @octets = @{$encoded_oids[$i]};
+
+    my $hash = $#octets;
+    foreach (@octets) {
+	$hash += $_ * 33;
+    }
+
+    $hash = ($hash >> 24) ^ ($hash >> 16) ^ ($hash >> 8) ^ ($hash);
+
+    push @hash_values, $hash & 0xff;
+}
+
+#
+# Emit the OID data
+#
+print C_FILE "\n";
+print C_FILE "static const unsigned char oid_data[", $total_length, "] = {\n";
+for (my $i = 0; $i <= $#names; $i++) {
+    my @octets = @{$encoded_oids[$i]};
+    print C_FILE "\t";
+    print C_FILE $_, ", " foreach (@octets);
+    print C_FILE "\t// ", $names[$i];
+    print C_FILE "\n";
+}
+print C_FILE "};\n";
+
+#
+# Build the search index table (ordered by length then hash then content)
+#
+my @index_table = ( 0 .. $#names );
+
+ at index_table = sort {
+    my @octets_a = @{$encoded_oids[$a]};
+    my @octets_b = @{$encoded_oids[$b]};
+
+    return $hash_values[$a] <=> $hash_values[$b]
+	if ($hash_values[$a] != $hash_values[$b]);
+    return $#octets_a <=> $#octets_b
+	if ($#octets_a != $#octets_b);
+    for (my $i = $#octets_a; $i >= 0; $i--) {
+	return $octets_a[$i] <=> $octets_b[$i]
+	    if ($octets_a[$i] != $octets_b[$i]);
+    }
+    return 0;
+
+} @index_table;
+
+#
+# Emit the search index and hash value table
+#
+print C_FILE "\n";
+print C_FILE "static const struct {\n";
+print C_FILE "\tunsigned char hash;\n";
+if ($#names <= 255) {
+    print C_FILE "\tenum OID oid : 8;\n";
+} else {
+    print C_FILE "\tenum OID oid : 16;\n";
+}
+print C_FILE "} oid_search_table[OID__NR] = {\n";
+for (my $i = 0; $i <= $#names; $i++) {
+    my @octets = @{$encoded_oids[$index_table[$i]]};
+    printf(C_FILE "\t[%3u] = { %3u, OID_%-35s }, // ",
+	   $i,
+	   $hash_values[$index_table[$i]],
+	   $names[$index_table[$i]]);
+    printf C_FILE "%02x", $_ foreach (@octets);
+    print C_FILE "\n";
+}
+print C_FILE "};\n";
+
+#
+# Emit the OID debugging name table
+#
+#print C_FILE "\n";
+#print C_FILE "const char *const oid_name_table[OID__NR + 1] = {\n";
+#
+#for (my $i = 0; $i <= $#names; $i++) {
+#    print C_FILE "\t\"", $names[$i], "\",\n"
+#}
+#print C_FILE "\t\"Unknown-OID\"\n";
+#print C_FILE "};\n";
+
+#
+# Polish off
+#
+close C_FILE or die;
lib/oid_registry.c new file +179 -0
new file mode 100644
index 000000000000..209edc73b99f
@@ -0,0 +1,179 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* ASN.1 Object identifier (OID) registry
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifdef __UBOOT__
+#include <linux/compat.h>
+#else
+#include <linux/module.h>
+#include <linux/export.h>
+#endif
+#include <linux/oid_registry.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/bug.h>
+#include "oid_registry_data.c"
+
+MODULE_DESCRIPTION("OID Registry");
+MODULE_AUTHOR("Red Hat, Inc.");
+MODULE_LICENSE("GPL");
+
+/**
+ * look_up_OID - Find an OID registration for the specified data
+ * @data: Binary representation of the OID
+ * @datasize: Size of the binary representation
+ */
+enum OID look_up_OID(const void *data, size_t datasize)
+{
+	const unsigned char *octets = data;
+	enum OID oid;
+	unsigned char xhash;
+	unsigned i, j, k, hash;
+	size_t len;
+
+	/* Hash the OID data */
+	hash = datasize - 1;
+
+	for (i = 0; i < datasize; i++)
+		hash += octets[i] * 33;
+	hash = (hash >> 24) ^ (hash >> 16) ^ (hash >> 8) ^ hash;
+	hash &= 0xff;
+
+	/* Binary search the OID registry.  OIDs are stored in ascending order
+	 * of hash value then ascending order of size and then in ascending
+	 * order of reverse value.
+	 */
+	i = 0;
+	k = OID__NR;
+	while (i < k) {
+		j = (i + k) / 2;
+
+		xhash = oid_search_table[j].hash;
+		if (xhash > hash) {
+			k = j;
+			continue;
+		}
+		if (xhash < hash) {
+			i = j + 1;
+			continue;
+		}
+
+		oid = oid_search_table[j].oid;
+		len = oid_index[oid + 1] - oid_index[oid];
+		if (len > datasize) {
+			k = j;
+			continue;
+		}
+		if (len < datasize) {
+			i = j + 1;
+			continue;
+		}
+
+		/* Variation is most likely to be at the tail end of the
+		 * OID, so do the comparison in reverse.
+		 */
+		while (len > 0) {
+			unsigned char a = oid_data[oid_index[oid] + --len];
+			unsigned char b = octets[len];
+			if (a > b) {
+				k = j;
+				goto next;
+			}
+			if (a < b) {
+				i = j + 1;
+				goto next;
+			}
+		}
+		return oid;
+	next:
+		;
+	}
+
+	return OID__NR;
+}
+EXPORT_SYMBOL_GPL(look_up_OID);
+
+/*
+ * sprint_OID - Print an Object Identifier into a buffer
+ * @data: The encoded OID to print
+ * @datasize: The size of the encoded OID
+ * @buffer: The buffer to render into
+ * @bufsize: The size of the buffer
+ *
+ * The OID is rendered into the buffer in "a.b.c.d" format and the number of
+ * bytes is returned.  -EBADMSG is returned if the data could not be intepreted
+ * and -ENOBUFS if the buffer was too small.
+ */
+int sprint_oid(const void *data, size_t datasize, char *buffer, size_t bufsize)
+{
+	const unsigned char *v = data, *end = v + datasize;
+	unsigned long num;
+	unsigned char n;
+	size_t ret;
+	int count;
+
+	if (v >= end)
+		goto bad;
+
+	n = *v++;
+	ret = count = snprintf(buffer, bufsize, "%u.%u", n / 40, n % 40);
+	if (count >= bufsize)
+		return -ENOBUFS;
+	buffer += count;
+	bufsize -= count;
+
+	while (v < end) {
+		num = 0;
+		n = *v++;
+		if (!(n & 0x80)) {
+			num = n;
+		} else {
+			num = n & 0x7f;
+			do {
+				if (v >= end)
+					goto bad;
+				n = *v++;
+				num <<= 7;
+				num |= n & 0x7f;
+			} while (n & 0x80);
+		}
+		ret += count = snprintf(buffer, bufsize, ".%lu", num);
+		if (count >= bufsize)
+			return -ENOBUFS;
+		buffer += count;
+		bufsize -= count;
+	}
+
+	return ret;
+
+bad:
+	snprintf(buffer, bufsize, "(bad)");
+	return -EBADMSG;
+}
+EXPORT_SYMBOL_GPL(sprint_oid);
+
+/**
+ * sprint_OID - Print an Object Identifier into a buffer
+ * @oid: The OID to print
+ * @buffer: The buffer to render into
+ * @bufsize: The size of the buffer
+ *
+ * The OID is rendered into the buffer in "a.b.c.d" format and the number of
+ * bytes is returned.
+ */
+int sprint_OID(enum OID oid, char *buffer, size_t bufsize)
+{
+	int ret;
+
+	BUG_ON(oid >= OID__NR);
+
+	ret = sprint_oid(oid_data + oid_index[oid],
+			 oid_index[oid + 1] - oid_index[oid],
+			 buffer, bufsize);
+	BUG_ON(ret == -EBADMSG);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(sprint_OID);
#23 lore raw

PATCH 8/11 lib: crypto: add public key utility

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

---
 include/crypto/public_key.h    |  90 +++++
 include/keys/asymmetric-type.h |  88 +++++
 lib/crypto/Kconfig             |  19 +
 lib/crypto/Makefile            |  10 +
 lib/crypto/asymmetric_type.c   | 668 +++++++++++++++++++++++++++++++++
 lib/crypto/public_key.c        | 376 +++++++++++++++++++
 6 files changed, 1251 insertions(+)
 create mode 100644 include/crypto/public_key.h
 create mode 100644 include/keys/asymmetric-type.h
 create mode 100644 lib/crypto/Kconfig
 create mode 100644 lib/crypto/Makefile
 create mode 100644 lib/crypto/asymmetric_type.c
 create mode 100644 lib/crypto/public_key.c
include/crypto/public_key.h new file +90 -0
new file mode 100644
index 000000000000..436a1ee1ee64
@@ -0,0 +1,90 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* Asymmetric public-key algorithm definitions
+ *
+ * See Documentation/crypto/asymmetric-keys.txt
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _LINUX_PUBLIC_KEY_H
+#define _LINUX_PUBLIC_KEY_H
+
+#ifdef __UBOOT__
+#include <linux/types.h>
+#else
+#include <linux/keyctl.h>
+#endif
+#include <linux/oid_registry.h>
+
+/*
+ * Cryptographic data for the public-key subtype of the asymmetric key type.
+ *
+ * Note that this may include private part of the key as well as the public
+ * part.
+ */
+struct public_key {
+	void *key;
+	u32 keylen;
+	enum OID algo;
+	void *params;
+	u32 paramlen;
+	bool key_is_private;
+	const char *id_type;
+	const char *pkey_algo;
+};
+
+extern void public_key_free(struct public_key *key);
+
+/*
+ * Public key cryptography signature data
+ */
+struct public_key_signature {
+	struct asymmetric_key_id *auth_ids[2];
+	u8 *s;			/* Signature */
+	u32 s_size;		/* Number of bytes in signature */
+	u8 *digest;
+	u8 digest_size;		/* Number of bytes in digest */
+	const char *pkey_algo;
+	const char *hash_algo;
+	const char *encoding;
+};
+
+extern void public_key_signature_free(struct public_key_signature *sig);
+
+#ifndef __UBOOT__
+extern struct asymmetric_key_subtype public_key_subtype;
+
+struct key;
+struct key_type;
+union key_payload;
+
+extern int restrict_link_by_signature(struct key *dest_keyring,
+				      const struct key_type *type,
+				      const union key_payload *payload,
+				      struct key *trust_keyring);
+
+extern int restrict_link_by_key_or_keyring(struct key *dest_keyring,
+					   const struct key_type *type,
+					   const union key_payload *payload,
+					   struct key *trusted);
+
+extern int restrict_link_by_key_or_keyring_chain(struct key *trust_keyring,
+						 const struct key_type *type,
+						 const union key_payload *payload,
+						 struct key *trusted);
+
+extern int query_asymmetric_key(const struct kernel_pkey_params *,
+				struct kernel_pkey_query *);
+
+extern int encrypt_blob(struct kernel_pkey_params *, const void *, void *);
+extern int decrypt_blob(struct kernel_pkey_params *, const void *, void *);
+extern int create_signature(struct kernel_pkey_params *, const void *, void *);
+extern int verify_signature(const struct key *,
+			    const struct public_key_signature *);
+
+int public_key_verify_signature(const struct public_key *pkey,
+				const struct public_key_signature *sig);
+#endif /* !__UBOOT__ */
+
+#endif /* _LINUX_PUBLIC_KEY_H */
include/keys/asymmetric-type.h new file +88 -0

88 lines not on this page · raw message

lib/crypto/Kconfig new file +19 -0
new file mode 100644
index 000000000000..b8e8288d2f80
@@ -0,0 +1,19 @@
+menuconfig ASYMMETRIC_KEY_TYPE
+	bool "Asymmetric (public-key cryptographic) key Support"
+	help
+	  This option provides support for a key type that holds the data for
+	  the asymmetric keys used for public key cryptographic operations such
+	  as encryption, decryption, signature generation and signature
+	  verification.
+
+if ASYMMETRIC_KEY_TYPE
+
+config ASYMMETRIC_PUBLIC_KEY_SUBTYPE
+	bool "Asymmetric public-key crypto algorithm subtype"
+	help
+	  This option provides support for asymmetric public key type handling.
+	  If signature generation and/or verification are to be used,
+	  appropriate hash algorithms (such as SHA-1) must be available.
+	  ENOPKG will be reported if the requisite algorithm is unavailable.
+
+endif # ASYMMETRIC_KEY_TYPE
lib/crypto/Makefile new file +10 -0
new file mode 100644
index 000000000000..a284de9e0411
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0+
+#
+# Makefile for asymmetric cryptographic keys
+#
+
+obj-$(CONFIG_ASYMMETRIC_KEY_TYPE) += asymmetric_keys.o
+
+asymmetric_keys-y := asymmetric_type.o
+
+obj-$(CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE) += public_key.o
lib/crypto/asymmetric_type.c new file +668 -0

668 lines not on this page · raw message

lib/crypto/public_key.c new file +376 -0

376 lines not on this page · raw message

Signature · 1 line
2.21.0
#24 lore raw

PATCH 9/11 lib: crypto: add x509 parser

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

---
 lib/Kconfig                   |   1 +
 lib/Makefile                  |   1 +
 lib/crypto/Kconfig            |  10 +
 lib/crypto/Makefile           |  17 +
 lib/crypto/x509.asn1          |  60 +++
 lib/crypto/x509_akid.asn1     |  35 ++
 lib/crypto/x509_cert_parser.c | 697 ++++++++++++++++++++++++++++++++++
 lib/crypto/x509_parser.h      |  57 +++
 lib/crypto/x509_public_key.c  | 292 ++++++++++++++
 9 files changed, 1170 insertions(+)
 create mode 100644 lib/crypto/x509.asn1
 create mode 100644 lib/crypto/x509_akid.asn1
 create mode 100644 lib/crypto/x509_cert_parser.c
 create mode 100644 lib/crypto/x509_parser.h
 create mode 100644 lib/crypto/x509_public_key.c
lib/Kconfig +1 -0
index d2955c8feb0e..aa44f52d9181 100644
@@ -253,6 +253,7 @@ config AES
 	  present.
 
 source lib/rsa/Kconfig
+source lib/crypto/Kconfig
 
 config TPM
 	bool "Trusted Platform Module (TPM) Support"
lib/Makefile +1 -0
index 677ec33a1ce1..64ff105945b3 100644
@@ -18,6 +18,7 @@ obj-$(CONFIG_CMD_DHRYSTONE) += dhry/
 obj-$(CONFIG_ARCH_AT91) += at91/
 obj-$(CONFIG_OPTEE) += optee/
 obj-$(CONFIG_BUILD_ASN1) += asn1_decoder.o
+obj-y += crypto/
 
 obj-$(CONFIG_AES) += aes.o
 
lib/crypto/Kconfig +10 -0
index b8e8288d2f80..a87cf6107fc1 100644
@@ -16,4 +16,14 @@ config ASYMMETRIC_PUBLIC_KEY_SUBTYPE
 	  appropriate hash algorithms (such as SHA-1) must be available.
 	  ENOPKG will be reported if the requisite algorithm is unavailable.
 
+config X509_CERTIFICATE_PARSER
+	bool "X.509 certificate parser"
+	depends on ASYMMETRIC_PUBLIC_KEY_SUBTYPE
+	select ASN1
+	select OID_REGISTRY
+	help
+	  This option provides support for parsing X.509 format blobs for key
+	  data and provides the ability to instantiate a crypto key from a
+	  public key packet found inside the certificate.
+
 endif # ASYMMETRIC_KEY_TYPE
lib/crypto/Makefile +17 -0
index a284de9e0411..2ec2d81a1f3c 100644
@@ -8,3 +8,20 @@ obj-$(CONFIG_ASYMMETRIC_KEY_TYPE) += asymmetric_keys.o
 asymmetric_keys-y := asymmetric_type.o
 
 obj-$(CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE) += public_key.o
+
+#
+# X.509 Certificate handling
+#
+obj-$(CONFIG_X509_CERTIFICATE_PARSER) += x509_key_parser.o
+x509_key_parser-y := \
+	x509.asn1.o \
+	x509_akid.asn1.o \
+	x509_cert_parser.o \
+	x509_public_key.o
+
+$(obj)/x509_cert_parser.o: \
+	$(obj)/x509.asn1.h \
+	$(obj)/x509_akid.asn1.h
+
+$(obj)/x509.asn1.o: $(obj)/x509.asn1.c $(obj)/x509.asn1.h
+$(obj)/x509_akid.asn1.o: $(obj)/x509_akid.asn1.c $(obj)/x509_akid.asn1.h
lib/crypto/x509.asn1 new file +60 -0

60 lines not on this page · raw message

lib/crypto/x509_akid.asn1 new file +35 -0
new file mode 100644
index 000000000000..1a33231a75a8
@@ -0,0 +1,35 @@
+-- X.509 AuthorityKeyIdentifier
+-- rfc5280 section 4.2.1.1
+
+AuthorityKeyIdentifier ::= SEQUENCE {
+	keyIdentifier			[0] IMPLICIT KeyIdentifier		OPTIONAL,
+	authorityCertIssuer		[1] IMPLICIT GeneralNames		OPTIONAL,
+	authorityCertSerialNumber	[2] IMPLICIT CertificateSerialNumber	OPTIONAL
+	}
+
+KeyIdentifier ::= OCTET STRING ({ x509_akid_note_kid })
+
+CertificateSerialNumber ::= INTEGER ({ x509_akid_note_serial })
+
+GeneralNames ::= SEQUENCE OF GeneralName
+
+GeneralName ::= CHOICE {
+	otherName			[0] ANY,
+	rfc822Name			[1] IA5String,
+	dNSName				[2] IA5String,
+	x400Address			[3] ANY,
+	directoryName			[4] Name ({ x509_akid_note_name }),
+	ediPartyName			[5] ANY,
+	uniformResourceIdentifier	[6] IA5String,
+	iPAddress			[7] OCTET STRING,
+	registeredID			[8] OBJECT IDENTIFIER
+	}
+
+Name ::= SEQUENCE OF RelativeDistinguishedName
+
+RelativeDistinguishedName ::= SET OF AttributeValueAssertion
+
+AttributeValueAssertion ::= SEQUENCE {
+	attributeType		OBJECT IDENTIFIER ({ x509_note_OID }),
+	attributeValue		ANY ({ x509_extract_name_segment })
+	}
lib/crypto/x509_cert_parser.c new file +697 -0

697 lines not on this page · raw message

lib/crypto/x509_parser.h new file +57 -0

57 lines not on this page · raw message

lib/crypto/x509_public_key.c new file +292 -0

292 lines not on this page · raw message

Signature · 1 line
2.21.0
#25 lore raw

PATCH 10/11 lib: crypto: add pkcs7 message parser

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

---
 include/crypto/pkcs7.h    |  47 +++
 lib/crypto/Kconfig        |   9 +
 lib/crypto/Makefile       |  11 +
 lib/crypto/pkcs7.asn1     | 135 ++++++++
 lib/crypto/pkcs7_parser.c | 693 ++++++++++++++++++++++++++++++++++++++
 lib/crypto/pkcs7_parser.h |  65 ++++
 scripts/Makefile.build    |   4 +-
 7 files changed, 961 insertions(+), 3 deletions(-)
 create mode 100644 include/crypto/pkcs7.h
 create mode 100644 lib/crypto/pkcs7.asn1
 create mode 100644 lib/crypto/pkcs7_parser.c
 create mode 100644 lib/crypto/pkcs7_parser.h
include/crypto/pkcs7.h new file +47 -0
new file mode 100644
index 000000000000..8f5c8a7ee3b9
@@ -0,0 +1,47 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* PKCS#7 crypto data parser
+ *
+ * Copyright (C) 2012 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells at redhat.com)
+ */
+
+#ifndef _CRYPTO_PKCS7_H
+#define _CRYPTO_PKCS7_H
+
+#ifndef __UBOOT__
+#include <linux/verification.h>
+#include <crypto/public_key.h>
+#endif
+
+struct key;
+struct pkcs7_message;
+
+/*
+ * pkcs7_parser.c
+ */
+extern struct pkcs7_message *pkcs7_parse_message(const void *data,
+						 size_t datalen);
+extern void pkcs7_free_message(struct pkcs7_message *pkcs7);
+
+extern int pkcs7_get_content_data(const struct pkcs7_message *pkcs7,
+				  const void **_data, size_t *_datalen,
+				  size_t *_headerlen);
+
+#ifndef __UBOOT__
+/*
+ * pkcs7_trust.c
+ */
+extern int pkcs7_validate_trust(struct pkcs7_message *pkcs7,
+				struct key *trust_keyring);
+
+/*
+ * pkcs7_verify.c
+ */
+extern int pkcs7_verify(struct pkcs7_message *pkcs7,
+			enum key_being_used_for usage);
+
+extern int pkcs7_supply_detached_data(struct pkcs7_message *pkcs7,
+				      const void *data, size_t datalen);
+#endif
+
+#endif /* _CRYPTO_PKCS7_H */
lib/crypto/Kconfig +9 -0
index a87cf6107fc1..86b676515405 100644
@@ -26,4 +26,13 @@ config X509_CERTIFICATE_PARSER
 	  data and provides the ability to instantiate a crypto key from a
 	  public key packet found inside the certificate.
 
+config PKCS7_MESSAGE_PARSER
+	bool "PKCS#7 message parser"
+	depends on X509_CERTIFICATE_PARSER
+	select ASN1
+	select OID_REGISTRY
+	help
+	  This option provides support for parsing PKCS#7 format messages for
+	  signature data and provides the ability to verify the signature.
+
 endif # ASYMMETRIC_KEY_TYPE
lib/crypto/Makefile +11 -0
index 2ec2d81a1f3c..a02ad51e698c 100644
@@ -25,3 +25,14 @@ $(obj)/x509_cert_parser.o: \
 
 $(obj)/x509.asn1.o: $(obj)/x509.asn1.c $(obj)/x509.asn1.h
 $(obj)/x509_akid.asn1.o: $(obj)/x509_akid.asn1.c $(obj)/x509_akid.asn1.h
+
+#
+# PKCS#7 message handling
+#
+obj-$(CONFIG_PKCS7_MESSAGE_PARSER) += pkcs7_message.o
+pkcs7_message-y := \
+	pkcs7.asn1.o \
+	pkcs7_parser.o
+
+$(obj)/pkcs7_parser.o: $(obj)/pkcs7.asn1.h
+$(obj)/pkcs7.asn1.o: $(obj)/pkcs7.asn1.c $(obj)/pkcs7.asn1.h
lib/crypto/pkcs7.asn1 new file +135 -0

135 lines not on this page · raw message

lib/crypto/pkcs7_parser.c new file +693 -0

693 lines not on this page · raw message

lib/crypto/pkcs7_parser.h new file +65 -0

65 lines not on this page · raw message

scripts/Makefile.build +1 -3
index 9357d310e50b..71458e089cd6 100644
@@ -419,11 +419,9 @@ targets += $(multi-used-y) $(multi-used-m)
 intermediate_targets = $(foreach sfx, $(2), \
 				$(patsubst %$(strip $(1)),%$(sfx), \
 					$(filter %$(strip $(1)), $(targets))))
-# %.asn1.o <- %.asn1.[ch] <- %.asn1
 # %.lex.o <- %.lex.c <- %.l
 # %.tab.o <- %.tab.[ch] <- %.y
-targets += $(call intermediate_targets, .asn1.o, .asn1.c .asn1.h) \
-	   $(call intermediate_targets, .lex.o, .lex.c) \
+targets += $(call intermediate_targets, .lex.o, .lex.c) \
 	   $(call intermediate_targets, .tab.o, .tab.c .tab.h)
 
 # Descending
Signature · 1 line
2.21.0
#26 lore raw

PATCH 11/11 lib: crypto: add rsa public key parser

In reply to AKASHI Takahiro

Imported from linux kernel v5.3.

---
 include/crypto/internal/rsa.h |  57 ++++++++++
 lib/crypto/Makefile           |  10 +-
 lib/crypto/rsa_helper.c       | 198 ++++++++++++++++++++++++++++++++++
 lib/crypto/rsapubkey.asn1     |   4 +
 4 files changed, 268 insertions(+), 1 deletion(-)
 create mode 100644 include/crypto/internal/rsa.h
 create mode 100644 lib/crypto/rsa_helper.c
 create mode 100644 lib/crypto/rsapubkey.asn1
include/crypto/internal/rsa.h new file +57 -0
new file mode 100644
index 000000000000..e870133f4b77
@@ -0,0 +1,57 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * RSA internal helpers
+ *
+ * Copyright (c) 2015, Intel Corporation
+ * Authors: Tadeusz Struk <[email protected]>
+ */
+#ifndef _RSA_HELPER_
+#define _RSA_HELPER_
+#include <linux/types.h>
+
+/**
+ * rsa_key - RSA key structure
+ * @n           : RSA modulus raw byte stream
+ * @e           : RSA public exponent raw byte stream
+ * @d           : RSA private exponent raw byte stream
+ * @p           : RSA prime factor p of n raw byte stream
+ * @q           : RSA prime factor q of n raw byte stream
+ * @dp          : RSA exponent d mod (p - 1) raw byte stream
+ * @dq          : RSA exponent d mod (q - 1) raw byte stream
+ * @qinv        : RSA CRT coefficient q^(-1) mod p raw byte stream
+ * @n_sz        : length in bytes of RSA modulus n
+ * @e_sz        : length in bytes of RSA public exponent
+ * @d_sz        : length in bytes of RSA private exponent
+ * @p_sz        : length in bytes of p field
+ * @q_sz        : length in bytes of q field
+ * @dp_sz       : length in bytes of dp field
+ * @dq_sz       : length in bytes of dq field
+ * @qinv_sz     : length in bytes of qinv field
+ */
+struct rsa_key {
+	const u8 *n;
+	const u8 *e;
+	const u8 *d;
+	const u8 *p;
+	const u8 *q;
+	const u8 *dp;
+	const u8 *dq;
+	const u8 *qinv;
+	size_t n_sz;
+	size_t e_sz;
+	size_t d_sz;
+	size_t p_sz;
+	size_t q_sz;
+	size_t dp_sz;
+	size_t dq_sz;
+	size_t qinv_sz;
+};
+
+int rsa_parse_pub_key(struct rsa_key *rsa_key, const void *key,
+		      unsigned int key_len);
+
+int rsa_parse_priv_key(struct rsa_key *rsa_key, const void *key,
+		       unsigned int key_len);
+
+extern struct crypto_template rsa_pkcs1pad_tmpl;
+#endif
lib/crypto/Makefile +9 -1
index a02ad51e698c..ac0ca05298e2 100644
@@ -17,7 +17,9 @@ x509_key_parser-y := \
 	x509.asn1.o \
 	x509_akid.asn1.o \
 	x509_cert_parser.o \
-	x509_public_key.o
+	x509_public_key.o \
+	rsapubkey.asn1.o \
+	rsa_helper.o
 
 $(obj)/x509_cert_parser.o: \
 	$(obj)/x509.asn1.h \
@@ -36,3 +38,9 @@ pkcs7_message-y := \
 
 $(obj)/pkcs7_parser.o: $(obj)/pkcs7.asn1.h
 $(obj)/pkcs7.asn1.o: $(obj)/pkcs7.asn1.c $(obj)/pkcs7.asn1.h
+
+#
+# RSA public key parser
+#
+$(obj)/rsapubkey.asn1.o: $(obj)/rsapubkey.asn1.c $(obj)/rsapubkey.asn1.h
+$(obj)/rsa_helper.o: $(obj)/rsapubkey.asn1.h
lib/crypto/rsa_helper.c new file +198 -0

198 lines not on this page · raw message

lib/crypto/rsapubkey.asn1 new file +4 -0
new file mode 100644
index 000000000000..725498e461d2
@@ -0,0 +1,4 @@
+RsaPubKey ::= SEQUENCE {
+	n INTEGER ({ rsa_get_n }),
+	e INTEGER ({ rsa_get_e })
+}
Signature · 1 line
2.21.0
#27 lore raw

PATCH 11/11 lib: crypto: add rsa public key parser

Quoted 1 line · AKASHI Takahiro

Imported from linux kernel v5.3.

It is unclear from your commit message which files are verbatime copies and which have been adjusted to fit into U-Boot.

Where is the unit test?

Best regards

Heinrich

Quoted 11 lines
---
  include/crypto/internal/rsa.h |  57 ++++++++++
  lib/crypto/Makefile           |  10 +-
  lib/crypto/rsa_helper.c       | 198 ++++++++++++++++++++++++++++++++++
  lib/crypto/rsapubkey.asn1     |   4 +
  4 files changed, 268 insertions(+), 1 deletion(-)

create mode 100644 include/crypto/internal/rsa.h create mode 100644 lib/crypto/rsa_helper.c create mode 100644 lib/crypto/rsapubkey.asn1

#28 lore raw

cover import x509/pkcs7 parsers from linux

In reply to AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 24 lines · AKASHI Takahiro

Asn1 parsers of x509 certificates and pkcs7 messages are required to implement image authentication and variable authentication as part of UEFI secure boot feature.

As we discussed before in the thread[1], most people insisted that we should re-use corresponding source code from Linux repository for this purpose.

Here is my attempt to import all the necessary files from Linux; Those will eventually be part of UEFI secure boot implementation, but I'd like to get early feedback from other peoples before submitting the whole patchset so that they will be better formatted for merging.

My approach here is
* files from the latest Linux
* modify files as little as possible
* mark/protect unavoidable changes with "#if(n)def __UBOOT__" so that future fixes/differences in Linux repository will easily be applied to U-Boot.

Please note that checkpatch.pl will complain with a bunch of warnings/errors but I intentionally left them unchanged for the sake of better maintainability I said above.

Quoted 7 lines

Any comments will be appreciated.
-Takahiro Akashi

[1] https://lists.denx.de/pipermail/u-boot/2019-April/366423.html

Changes in v1 (Oct 11, 2019) from RFC
* change the kernel code base from v5.0 to v5.3

* add preparatory patches (#1, #2 and #3)

-Takahiro Akashi

Quoted 84 lines

* comment off x509_check_for_self_signed() which is not useful for UEFI secure boot (patch#9)
* improve usages of "#if(n)def __UBOOT__* to minimize differences between U-Boot and linux kernel

AKASHI Takahiro (11):
linux_compat: add kmemdup()
include: time.h: define time64_t
include: kernel.h: include printk.h
cmd: add asn1_compiler
Makefile: add build script for asn1 parsers
lib: add asn1 decoder
lib: add oid registry utility
lib: crypto: add public key utility
lib: crypto: add x509 parser
lib: crypto: add pkcs7 message parser
lib: crypto: add rsa public key parser

 cmd/Kconfig                       |    3 +
 include/crypto/internal/rsa.h     |   57 +
 include/crypto/pkcs7.h            |   47 +
 include/crypto/public_key.h       |   90 ++
 include/keys/asymmetric-type.h    |   88 ++
 include/linux/asn1.h              |   65 ++
 include/linux/asn1_ber_bytecode.h |   89 ++
 include/linux/asn1_decoder.h      |   20 +
 include/linux/compat.h            |    4 +-
 include/linux/kernel.h            |    2 +
 include/linux/oid_registry.h      |  117 +++
 include/linux/time.h              |   24 +
 lib/Kconfig                       |   12 +
 lib/Makefile                      |   18 +
 lib/asn1_decoder.c                |  527 ++++++++++
 lib/build_OID_registry            |  203 ++++
 lib/crypto/Kconfig                |   38 +
 lib/crypto/Makefile               |   46 +
 lib/crypto/asymmetric_type.c      |  668 ++++++++++++
 lib/crypto/pkcs7.asn1             |  135 +++
 lib/crypto/pkcs7_parser.c         |  693 +++++++++++++
 lib/crypto/pkcs7_parser.h         |   65 ++
 lib/crypto/public_key.c           |  376 +++++++
 lib/crypto/rsa_helper.c           |  198 ++++
 lib/crypto/rsapubkey.asn1         |    4 +
 lib/crypto/x509.asn1              |   60 ++
 lib/crypto/x509_akid.asn1         |   35 +
 lib/crypto/x509_cert_parser.c     |  697 +++++++++++++
 lib/crypto/x509_parser.h          |   57 +
 lib/crypto/x509_public_key.c      |  292 ++++++
 lib/linux_compat.c                |   11 +
 lib/oid_registry.c                |  179 ++++
 scripts/Makefile                  |    3 +
 scripts/Makefile.build            |    2 +-
 scripts/asn1_compiler.c           | 1611 +++++++++++++++++++++++++++++
 35 files changed, 6533 insertions(+), 3 deletions(-)
 create mode 100644 include/crypto/internal/rsa.h
 create mode 100644 include/crypto/pkcs7.h
 create mode 100644 include/crypto/public_key.h
 create mode 100644 include/keys/asymmetric-type.h
 create mode 100644 include/linux/asn1.h
 create mode 100644 include/linux/asn1_ber_bytecode.h
 create mode 100644 include/linux/asn1_decoder.h
 create mode 100644 include/linux/oid_registry.h
 create mode 100644 lib/asn1_decoder.c
 create mode 100755 lib/build_OID_registry
 create mode 100644 lib/crypto/Kconfig
 create mode 100644 lib/crypto/Makefile
 create mode 100644 lib/crypto/asymmetric_type.c
 create mode 100644 lib/crypto/pkcs7.asn1
 create mode 100644 lib/crypto/pkcs7_parser.c
 create mode 100644 lib/crypto/pkcs7_parser.h
 create mode 100644 lib/crypto/public_key.c
 create mode 100644 lib/crypto/rsa_helper.c
 create mode 100644 lib/crypto/rsapubkey.asn1
 create mode 100644 lib/crypto/x509.asn1
 create mode 100644 lib/crypto/x509_akid.asn1
 create mode 100644 lib/crypto/x509_cert_parser.c
 create mode 100644 lib/crypto/x509_parser.h
 create mode 100644 lib/crypto/x509_public_key.c
 create mode 100644 lib/oid_registry.c
 create mode 100644 scripts/asn1_compiler.c
Signature · 1 line
2.21.0
#29 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

How can I make all of these new files being built to check for build warnings?

Please, provide unit tests for the patch series.

Please, provide a documentation how these different tools and files work together.

Best regards

Heinrich

#30 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 7 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Please note that almost of all the code here come from the latest linux code without any changes. A few changes that I made are quite U-Boot/UEFI-secure-boot specific and is *best* tested through UEFI secure boot test.

That said, you can at least build the code by enabling
Library routines
Security support
Asymmetric Key Support
CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE CONFIG_X509_CERTIFICATE_PARSER CONFIG_PKCS7_MESSAGE_PARSER

Quoted 1 line

Please, provide unit tests for the patch series.

As I said above, it will be exercised and tested under UEFI secure boot test.

Quoted 2 lines

Please, provide a documentation how these different tools and files work together.

What do you mean by different tools?
Asn1 compiler and what?
Do you want to have doc/README.asn1compiler?

Thanks,
-Takahiro Akashi

Quoted 3 lines

Best regards

Heinrich

#31 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

Quoted 14 lines

Please note that almost of all the code here come from the latest linux code without any changes. A few changes that I made are quite U-Boot/UEFI-secure-boot specific and is *best* tested through UEFI secure boot test.

That said, you can at least build the code by enabling
Library routines
Security support
Asymmetric Key Support
CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE CONFIG_X509_CERTIFICATE_PARSER CONFIG_PKCS7_MESSAGE_PARSER

Quoted 4 lines
Quoted 1 line

Please, provide unit tests for the patch series.

As I said above, it will be exercised and tested under UEFI secure boot test.

If there is nothing I can test now, I would not know how to evaluate these patches.

There is good reason that we have unit tests and don't simply say U-Boot can be tested by booting Linux.

Quoted 7 lines
Quoted 3 lines

Please, provide a documentation how these different tools and files work together.

What do you mean by different tools?
Asn1 compiler and what?
Do you want to have doc/README.asn1compiler?

This patch series provides some puzzle pieces but doesn't tell me how they fit together. Maybe a README describing the different elements provided for UEFI secure boot would be most appropriate.

Best regards

Heinrich

Quoted 9 lines

Thanks,
-Takahiro Akashi

Quoted 3 lines

Best regards

Heinrich

#32 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 30 lines · Heinrich Schuchardt
Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

Quoted 14 lines

Please note that almost of all the code here come from the latest linux code without any changes. A few changes that I made are quite U-Boot/UEFI-secure-boot specific and is *best* tested through UEFI secure boot test.

That said, you can at least build the code by enabling
Library routines
Security support
Asymmetric Key Support
CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE CONFIG_X509_CERTIFICATE_PARSER CONFIG_PKCS7_MESSAGE_PARSER

Quoted 7 lines
Quoted 4 lines
Quoted 1 line

Please, provide unit tests for the patch series.

As I said above, it will be exercised and tested under UEFI secure boot test.

If there is nothing I can test now, I would not know how to evaluate these patches.

Do you ask me to write "unit tests" to test all the aspects of asn1 compiler and parsers that I have not developed any part of and that are not changed from the original?
Doesn't make sense.

Quoted 2 lines

There is good reason that we have unit tests and don't simply say U-Boot can be tested by booting Linux.

There are lots of examples, one is the original RSA routines, as I said, which have not direct-linked tests and are only tested by vboot.py.

Quoted 11 lines
Quoted 7 lines
Quoted 3 lines

Please, provide a documentation how these different tools and files work together.

What do you mean by different tools?
Asn1 compiler and what?
Do you want to have doc/README.asn1compiler?

This patch series provides some puzzle pieces but doesn't tell me how they fit together. Maybe a README describing the different elements provided for UEFI secure boot would be most appropriate.

All what you need to know is that the patch set will generate and provide x509 parser and pkcs7 parser as a result of build process.

I will a few lines of README.asn1compiler to describe that.

-Takahiro Akashi

Quoted 14 lines

Best regards

Heinrich

Quoted 9 lines

Thanks,
-Takahiro Akashi

Quoted 3 lines

Best regards

Heinrich

#33 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 36 lines · AKASHI Takahiro
Quoted 35 lines · Heinrich Schuchardt
Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

Quoted 18 lines

Please note that almost of all the code here come from the latest linux code without any changes. A few changes that I made are quite U-Boot/UEFI-secure-boot specific and is *best* tested through UEFI secure boot test.

That said, you can at least build the code by enabling
Library routines
Security support
Asymmetric Key Support
CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE CONFIG_X509_CERTIFICATE_PARSER CONFIG_PKCS7_MESSAGE_PARSER

Quoted 1 line

Please, provide unit tests for the patch series.

As I said above, it will be exercised and tested under UEFI secure boot test.

Quoted 7 lines
Quoted 2 lines

If there is nothing I can test now, I would not know how to evaluate these patches.

Do you ask me to write "unit tests" to test all the aspects of asn1 compiler and parsers that I have not developed any part of and that are not changed from the original?
Doesn't make sense.

Wouldn't it be enough to ASN1-compile one file and check the MD5 hash of the result file?

Regards

Heinrich

Quoted 41 lines
Quoted 3 lines

There is good reason that we have unit tests and don't simply say U-Boot can be tested by booting Linux.

There are lots of examples, one is the original RSA routines, as I said, which have not direct-linked tests and are only tested by vboot.py.

Quoted 11 lines
Quoted 7 lines
Quoted 3 lines

Please, provide a documentation how these different tools and files work together.

What do you mean by different tools?
Asn1 compiler and what?
Do you want to have doc/README.asn1compiler?

This patch series provides some puzzle pieces but doesn't tell me how they fit together. Maybe a README describing the different elements provided for UEFI secure boot would be most appropriate.

All what you need to know is that the patch set will generate and provide x509 parser and pkcs7 parser as a result of build process.

I will a few lines of README.asn1compiler to describe that.

-Takahiro Akashi

Quoted 14 lines

Best regards

Heinrich

Quoted 9 lines

Thanks,
-Takahiro Akashi

Quoted 3 lines

Best regards

Heinrich

#34 lore raw

cover import x509/pkcs7 parsers from linux

In reply to Heinrich Schuchardt
Quoted 7 lines · Heinrich Schuchardt
Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

As far as this patch set is concerned, no configuration enables any of kconfig options listed below and running Travis doesn't make sense.

-Takahiro Akashi

Quoted 51 lines
Quoted 14 lines

Please note that almost of all the code here come from the latest linux code without any changes. A few changes that I made are quite U-Boot/UEFI-secure-boot specific and is *best* tested through UEFI secure boot test.

That said, you can at least build the code by enabling
Library routines
Security support
Asymmetric Key Support
CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE CONFIG_X509_CERTIFICATE_PARSER CONFIG_PKCS7_MESSAGE_PARSER

Quoted 4 lines
Quoted 1 line

Please, provide unit tests for the patch series.

As I said above, it will be exercised and tested under UEFI secure boot test.

If there is nothing I can test now, I would not know how to evaluate these patches.

There is good reason that we have unit tests and don't simply say U-Boot can be tested by booting Linux.

Quoted 7 lines
Quoted 3 lines

Please, provide a documentation how these different tools and files work together.

What do you mean by different tools?
Asn1 compiler and what?
Do you want to have doc/README.asn1compiler?

This patch series provides some puzzle pieces but doesn't tell me how they fit together. Maybe a README describing the different elements provided for UEFI secure boot would be most appropriate.

Best regards

Heinrich

Quoted 9 lines

Thanks,
-Takahiro Akashi

Quoted 3 lines

Best regards

Heinrich

#35 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 11 lines · AKASHI Takahiro
Quoted 10 lines · Heinrich Schuchardt
Quoted 9 lines · AKASHI Takahiro
Quoted 7 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

How can I make all of these new files being built to check for build warnings?

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

As far as this patch set is concerned, no configuration enables any of kconfig options listed below and running Travis doesn't make sense.

That's a problem that needs to be fixed. I am not OK with the idea of adding a new feature that will not be put through our CI and so when things break it (which will happen, inadvertently) it won't be caught until much later. Figuring out how to extend our CI to test this is a must. Thanks!

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20191017/6a6458df/attachment.sig>
#36 lore raw

cover import x509/pkcs7 parsers from linux

Hi Tom,

Quoted 18 lines · Tom Rini
Quoted 17 lines · AKASHI Takahiro
Quoted 15 lines · Heinrich Schuchardt
Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

Quoted 8 lines
Quoted 2 lines

As far as this patch set is concerned, no configuration enables any of kconfig options listed below and running Travis doesn't make sense.

That's a problem that needs to be fixed. I am not OK with the idea of adding a new feature that will not be put through our CI and so when things break it (which will happen, inadvertently) it won't be caught until much later. Figuring out how to extend our CI to test this is a must. Thanks!

I added a simple *unit* test under "test/lib."
As I said before, however, no existing configuration enables either
CONFIG_X509_CERTIFICATE_PARSER, nor CONFIG_PKCS7_MESSAGE_PARSER

and the related code won't be built or exercised in any way. So I made a small trick to Kconfig:

=== 8< ===
config UT_LIB
bool "Unit tests for library functions"
depends on UNIT_TEST
default y
help
Enables the 'ut lib' command which tests library functions like memcat(), memcyp(), memmove().

if UT_LIB

config UT_LIB_ASN1
bool "Unit test for asn1 compiler and decoder function" default y
imply ASYMMETRIC_KEY_TYPE
imply ASYMMETRIC_PUBLIC_KEY_SUBTYPE
imply X509_CERTIFICATE_PARSER
imply PKCS7_MESSAGE_PARSER
imply RSA_PUBLIC_KEY_PARSER
help
Enables a test which exercises asn1 compiler and decoder function via various parsers.

endif
=== >8 ===

So as long as UT_LIB is enabled and run by one of Travis CI test cases, a new test for ASN1 will also be exercised.
(I don't know which one will invoke "ut" command.)

Do you agree to this approach?

Thanks,
-Takahiro Akashi

Quoted 2 lines
Signature · 1 line
Tom
#37 lore raw

cover import x509/pkcs7 parsers from linux

Quoted 61 lines · AKASHI Takahiro

Hi Tom,

Quoted 18 lines · Tom Rini
Quoted 17 lines · AKASHI Takahiro
Quoted 15 lines · Heinrich Schuchardt
Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

Quoted 8 lines
Quoted 2 lines

As far as this patch set is concerned, no configuration enables any of kconfig options listed below and running Travis doesn't make sense.

That's a problem that needs to be fixed. I am not OK with the idea of adding a new feature that will not be put through our CI and so when things break it (which will happen, inadvertently) it won't be caught until much later. Figuring out how to extend our CI to test this is a must. Thanks!

I added a simple *unit* test under "test/lib."
As I said before, however, no existing configuration enables either
CONFIG_X509_CERTIFICATE_PARSER, nor CONFIG_PKCS7_MESSAGE_PARSER

and the related code won't be built or exercised in any way. So I made a small trick to Kconfig:

=== 8< ===
config UT_LIB
bool "Unit tests for library functions"
depends on UNIT_TEST
default y
help
Enables the 'ut lib' command which tests library functions like memcat(), memcyp(), memmove().

if UT_LIB

config UT_LIB_ASN1
bool "Unit test for asn1 compiler and decoder function" default y
imply ASYMMETRIC_KEY_TYPE
imply ASYMMETRIC_PUBLIC_KEY_SUBTYPE
imply X509_CERTIFICATE_PARSER
imply PKCS7_MESSAGE_PARSER
imply RSA_PUBLIC_KEY_PARSER
help
Enables a test which exercises asn1 compiler and decoder function via various parsers.

Quoted 8 lines

endif
=== >8 ===

So as long as UT_LIB is enabled and run by one of Travis CI test cases, a new test for ASN1 will also be exercised.
(I don't know which one will invoke "ut" command.)

Do you agree to this approach?

I think you're going to need to get a bit more familiar with some aspects of testing and building. Today, CONFIG_UNIT_TEST is enabled for sandbox. And we want as much as possible enabled on sandbox as that's also where coverity scan is performed. So that will get us part of the way forward here longer term. I think however you're going to also need to enable some tests on the QEMU platforms so that we can have all of this new secure boot code put through CI there. Thanks!

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20191018/69c91ac0/attachment.sig>
#38 lore raw

cover import x509/pkcs7 parsers from linux

Tom,

Quoted 62 lines · Tom Rini
Quoted 61 lines · AKASHI Takahiro

Hi Tom,

Quoted 18 lines · Tom Rini
Quoted 17 lines · AKASHI Takahiro
Quoted 15 lines · Heinrich Schuchardt
Quoted 6 lines · AKASHI Takahiro
Quoted 5 lines · Heinrich Schuchardt
Quoted 3 lines · AKASHI Takahiro

I hope this patch set will be reviewed promptly as I'm aiming to push my "UEFI secure boot" patch for v2020.01.

Quoted 6 lines
Quoted 2 lines

How can I make all of these new files being built to check for build warnings?

As always in my case of UEFI secure boot, they have gone through build and run/tests as part of UEFI secure boot. This is also true for RSA extension as UEFI secure boot is the only user of those features.

Did you run them through Travis?

Quoted 8 lines
Quoted 2 lines

As far as this patch set is concerned, no configuration enables any of kconfig options listed below and running Travis doesn't make sense.

That's a problem that needs to be fixed. I am not OK with the idea of adding a new feature that will not be put through our CI and so when things break it (which will happen, inadvertently) it won't be caught until much later. Figuring out how to extend our CI to test this is a must. Thanks!

I added a simple *unit* test under "test/lib."
As I said before, however, no existing configuration enables either
CONFIG_X509_CERTIFICATE_PARSER, nor CONFIG_PKCS7_MESSAGE_PARSER

and the related code won't be built or exercised in any way. So I made a small trick to Kconfig:

=== 8< ===
config UT_LIB
bool "Unit tests for library functions"
depends on UNIT_TEST
default y
help
Enables the 'ut lib' command which tests library functions like memcat(), memcyp(), memmove().

if UT_LIB

config UT_LIB_ASN1
bool "Unit test for asn1 compiler and decoder function" default y
imply ASYMMETRIC_KEY_TYPE
imply ASYMMETRIC_PUBLIC_KEY_SUBTYPE
imply X509_CERTIFICATE_PARSER
imply PKCS7_MESSAGE_PARSER
imply RSA_PUBLIC_KEY_PARSER
help
Enables a test which exercises asn1 compiler and decoder function via various parsers.

Quoted 8 lines
Quoted 8 lines

endif
=== >8 ===

So as long as UT_LIB is enabled and run by one of Travis CI test cases, a new test for ASN1 will also be exercised.
(I don't know which one will invoke "ut" command.)

Do you agree to this approach?

I'd like to confirm exactly what you suggested here:

Quoted 4 lines

I think you're going to need to get a bit more familiar with some aspects of testing and building. Today, CONFIG_UNIT_TEST is enabled for sandbox. And we want as much as possible enabled on sandbox as that's also where coverity scan is performed.

Is this ("as much as possible") true? I wonder how it should be achieved.

As far as my RSA test approach above is concerned,
1) If UT_TEST is enabled, UT_LIB and then UT_LIB_ASN1 are also enabled *by default* and expected to be run automatically through Travis's sandbox build with test/py even if, say, X509_CERTIRFFICATE_PARSER is *not* enabled in any of sandbox_*_defconfig.
Is this approach is OK for you?

Or,
2) Should we add ASYMMETRIC_*/X509_CERTIFICATE_PARSER/PKCS7_MESSAGE_PARSER to one (or all) of sandbox_*_defconfig (even though there is no explicit user of these features before my secure boot patch will be merged)?
Or,
3) Should we create a new sandbox_*_defconfig for any further tests? Or,
4) Would we better set ASYMMETRIC_*/X509_CERTIFICATE_PARSER/PKCS7_MESSAGE_PARSER enabled by default on Sandbox? I mean,
=== 8< ===
menuconfig ASYMMETRIC_KEY_TYPE

	bool "Asymmetric (public-key cryptographic) key Support"
	default y if SANDBOX	<== added

if ASYMMETRIC_KEY_TYPE
...
config X509_CERTIFICATE_PARSER
bool "RSA public key parser"
depends on ASYMMETRIC_PUBLIC_KEY_SUBTYPE
default y if SANDBOX <== added
...
=== >8 ===

Quoted 4 lines

So that will get us part of the
way forward here longer term. I think however you're going to also need to enable some tests on the QEMU platforms so that we can have all of this new secure boot code put through CI there. Thanks!

Are you talking about this patch set(asn1 parsers) or is this your general comment on my overall "UEFI secure boot" patch?

Please note that my secure boot patch is architecture agnostic and will perfectly work on Sandbox and all the related py tests will also be done on Sandbox.

Thanks,
-Takahiro Akashi

Quoted 2 lines
Signature · 1 line
Tom