Skip to contentSkip to conversation

padlock-aes.c used forward inline function

Patch? Adam J. Richter 2 participants 2 messages Dec 30, 2004 Crypto API

lore
Share & tools
lore
#0 lore raw

gcc-3.4.3 and gcc-3.4.1 do not currently support forward
declaration of inline functions, as is used in linux-2.6.10-rc2/drivers/crypto/padlock-aes.c.

The function is trivial, so it is better to inline it
by defining it earlier than it would be to un-inline it. Here is a proposed patch. If it looks okay, I would appreciate it if someone would bless it and forward it downstream.

                    __     ______________ 

Adam J. Richter \ /
[email protected] | g g d r a s i l

drivers/crypto/padlock-aes.c +13 -15
@@ -58,8 +58,6 @@
 #define AES_EXTENDED_KEY_SIZE	64	/* in uint32_t units */
 #define AES_EXTENDED_KEY_SIZE_B	(AES_EXTENDED_KEY_SIZE * sizeof(uint32_t))
 
-static inline int aes_hw_extkey_available (uint8_t key_len);
-
 struct aes_ctx {
 	uint32_t e_data[AES_EXTENDED_KEY_SIZE+4];
 	uint32_t d_data[AES_EXTENDED_KEY_SIZE+4];
@@ -68,6 +66,19 @@
 	int key_length;
 };
 
+/* Tells whether the ACE is capable to generate
+   the extended key for a given key_len. */
+static inline int
+aes_hw_extkey_available(uint8_t key_len)
+{
+	/* TODO: We should check the actual CPU model/stepping
+	         as it's possible that the capability will be
+	         added in the next CPU revisions. */
+	if (key_len == 16)
+		return 1;
+	return 0;
+}
+
 /* ====== Key management routines ====== */
 
 static inline uint32_t
@@ -356,19 +367,6 @@
 	return 0;
 }
 
-/* Tells whether the ACE is capable to generate
-   the extended key for a given key_len. */
-static inline int
-aes_hw_extkey_available(uint8_t key_len)
-{
-	/* TODO: We should check the actual CPU model/stepping
-	         as it's possible that the capability will be
-	         added in the next CPU revisions. */
-	if (key_len == 16)
-		return 1;
-	return 0;
-}
-
 /* ====== Encryption/decryption routines ====== */
 
 /* This is the real call to PadLock. */
#1 lore raw
Quoted 8 lines · Adam J. Richter

gcc-3.4.3 and gcc-3.4.1 do not currently support forward
declaration of inline functions, as is used in linux-2.6.10-rc2/drivers/crypto/padlock-aes.c.

The function is trivial, so it is better to inline it
by defining it earlier than it would be to un-inline it. Here is a proposed patch. If it looks okay, I would appreciate it if someone would bless it and forward it downstream.

I have already sent exactly the same patch directly to Andrew and Linus some days ago ;-)

Michal Ludvig

Signature · 2 lines
* A mouse is a device used to point at the xterm you want to type in.
* Personal homepage - http://www.logix.cz/michal