evm: Allow non-SHA1 digital signatures
Subsystems (5)
PATCH v4 1/1 evm: Allow non-SHA1 digital signatures
SHA1 is reasonable in HMAC constructs, but it's desirable to be able to use stronger hashes in digital signatures. Modify the EVM crypto code so the hash type is imported from the digital signature and passed down to the hash calculation code, and return the digest size to higher layers for validation.
--- security/integrity/evm/evm.h | 10 ++++- security/integrity/evm/evm_crypto.c | 59 +++++++++++++++-------------- security/integrity/evm/evm_main.c | 19 ++++++---- 3 files changed, 51 insertions(+), 37 deletions(-)
security/integrity/evm/evm.h +8 -2
@@ -42,6 +42,11 @@ extern struct crypto_shash *hash_tfm;/* List of EVM protected security xattrs */ extern char *evm_config_xattrnames[]; +struct evm_digest { +struct ima_digest_data hdr; +char digest[IMA_MAX_DIGEST_SIZE ]; +}__packed__ ; +int evm_init_key (void ); int evm_update_evmxattr(struct dentry *dentry, const char *req_xattr_name,
@@ -49,10 +54,11 @@ int evm_update_evmxattr(struct dentry *dentry, size_t req_xattr_value_len);int evm_calc_hmac (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char *digest); +size_t req_xattr_value_len,struct evm_digest *data);int evm_calc_hash (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char type,char *digest); +size_t req_xattr_value_len,char type, +struct evm_digest *data);int evm_init_hmac (struct inode *inode,const struct xattr *xattr,char *hmac_val);int evm_init_secfs (void );
security/integrity/evm/evm_crypto.c +31 -28
149 lines not on this page · raw message
security/integrity/evm/evm_main.c +12 -7
@@ -25,6 +25,7 @@#include <linux/magic.h> #include <crypto/hash.h> +#include <crypto/hash_info.h> #include <crypto/algapi.h> #include "evm.h"
@@ -122,8 +123,9 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, struct integrity_iint_cache *iint) {struct evm_ima_xattr_data *xattr_data =NULL ; -struct evm_ima_xattr_data calc; +struct signature_v2_hdr *hdr;enum integrity_status evm_status =INTEGRITY_PASS ; +struct evm_digest digest;struct inode *inode;int rc, xattr_len;
@@ -159,25 +161,28 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, evm_status =INTEGRITY_FAIL ; goto out; } + + digest.hdr.algo =HASH_ALGO_SHA1 ; rc =evm_calc_hmac (dentry, xattr_name, xattr_value, - xattr_value_len, calc.digest); + xattr_value_len, &digest);if (rc)break ; - rc =crypto_memneq (xattr_data->digest, calc.digest, -sizeof (calc.digest)); + rc =crypto_memneq (xattr_data->digest, digest.digest, +SHA1_DIGEST_SIZE );if (rc) rc = -EINVAL ;break ;case EVM_IMA_XATTR_DIGSIG :case EVM_XATTR_PORTABLE_DIGSIG : + hdr = (struct signature_v2_hdr *)xattr_data; + digest.hdr.algo = hdr->hash_algo; rc =evm_calc_hash (dentry, xattr_name, xattr_value, - xattr_value_len, xattr_data->type, - calc.digest); + xattr_value_len, xattr_data->type, &digest);if (rc)break ; rc =integrity_digsig_verify (INTEGRITY_KEYRING_EVM , (const char *)xattr_data, xattr_len, - calc.digest,sizeof (calc.digest)); + digest.digest, digest.hdr.length);if (!rc) { inode =d_backing_inode (dentry);
Signature · 1 line
2.17.0.441.gb46fe60e1d-goog
Quoted 19 lines · Matthew Garrett
SHA1 is reasonable in HMAC constructs, but it's desirable to be able to use stronger hashes in digital signatures. Modify the EVM crypto code so the hash type is imported from the digital signature and passed down to the hash calculation code, and return the digest size to higher layers for validation.
Signed-off-by Matthew Garrett <[email protected]>--- security/integrity/evm/evm.h | 10 ++++- security/integrity/evm/evm_crypto.c | 59 +++++++++++++++-------------- security/integrity/evm/evm_main.c | 19 ++++++---- 3 files changed, 51 insertions(+), 37 deletions(-)security/integrity/evm/evm.h +0
-0
security/integrity/evm/evm.h extern struct crypto_shash *hash_tfm; · line 48
Quoted 7 lines
security/integrity/evm/evm.h +4
-0@@ -42,6 +42,11 @@ extern struct crypto_shash *hash_tfm;/* List of EVM protected security xattrs */ extern char *evm_config_xattrnames[]; +struct evm_digest { +struct ima_digest_data hdr; +char digest[IMA_MAX_DIGEST_SIZE ]; +}__packed__ ;
security/integrity/evm/evm_crypto.o:(.bss+0x0): multiple definition of `__packed__'
security/integrity/evm/evm_main.o:(.bss+0x20): first defined here security/integrity/evm/evm_secfs.o:(.bss+0x0): multiple definition of `__packed__' security/integrity/evm/evm_main.o:/home/mimi/src/kernel/linux- integrity/security/integrity/evm/evm_main.c:236: first defined here Makefile:1036: recipe for target 'vmlinux' failed
make: *** [vmlinux] Error 1
After removing it, the security.evm hmac's are not properly validating, preventing the EVM/IMA keys from loading. You can imagine how far my system boots without the public keys.
Mimi
Quoted 241 lines
security/integrity/evm/evm.h +4
-2+int evm_init_key (void ); int evm_update_evmxattr(struct dentry *dentry, const char *req_xattr_name,@@ -49,10 +54,11 @@ int evm_update_evmxattr(struct dentry *dentry, size_t req_xattr_value_len);int evm_calc_hmac (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char *digest); +size_t req_xattr_value_len,struct evm_digest *data);int evm_calc_hash (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char type,char *digest); +size_t req_xattr_value_len,char type, +struct evm_digest *data);int evm_init_hmac (struct inode *inode,const struct xattr *xattr,char *hmac_val);int evm_init_secfs (void );security/integrity/evm/evm_crypto.c +31
-28@@ -21,6 +21,7 @@#include <linux/evm.h> #include <keys/encrypted-type.h> #include <crypto/hash.h> +#include <crypto/hash_info.h> #include "evm.h" #define EVMKEY "evm-key"@@ -28,8 +29,7 @@static unsigned char evmkey[MAX_KEY_SIZE ];static int evmkey_len =MAX_KEY_SIZE ; -struct crypto_shash *hmac_tfm; -struct crypto_shash *hash_tfm; +static struct crypto_shash *evm_tfm[HASH_ALGO__LAST ];static DEFINE_MUTEX (mutex);@@ -37,9 +37,6 @@ static DEFINE_MUTEX(mutex);static unsigned long evm_set_key_flags; -static char *const evm_hmac ="hmac(sha1)" ; -static char *const evm_hash ="sha1" ; -/** * evm_set_key() - set EVM HMAC key from the kernel * @key: pointer to a buffer with the key data @@ -74,10 +71,10 @@ int evm_set_key(void *key, size_t keylen) }EXPORT_SYMBOL_GPL (evm_set_key); -static struct shash_desc *init_desc (char type) +static struct shash_desc *init_desc (char type,uint8_t hash_algo) {long rc; -char *algo; +const char *algo;struct crypto_shash **tfm;struct shash_desc *desc;@@ -86,13 +83,11 @@ static struct shash_desc *init_desc(char type)pr_err_once ("HMAC key is not set\n" );return ERR_PTR (-ENOKEY ); } - tfm = &hmac_tfm; - algo = evm_hmac; - }else { - tfm = &hash_tfm; - algo = evm_hash; } + tfm = &evm_tfm[hash_algo]; + algo = hash_algo_name[hash_algo]; +if (*tfm ==NULL ) {mutex_lock (&mutex);if (*tfm)@@ -186,10 +181,10 @@ static void hmac_add_misc(struct shash_desc *desc, struct inode *inode,* each xattr, but attempt to re-use the previously allocated memory. */ static int evm_calc_hmac_or_hash (struct dentry *dentry, -const char *req_xattr_name, -const char *req_xattr_value, -size_t req_xattr_value_len, -char type,char *digest) +const char *req_xattr_name, +const char *req_xattr_value, +size_t req_xattr_value_len, +uint8_t type,struct evm_digest *data) {struct inode *inode =d_backing_inode (dentry);struct shash_desc *desc;@@ -203,10 +198,17 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,if (!(inode->i_opflags &IOP_XATTR ))return -EOPNOTSUPP ; - desc =init_desc (type); + desc =init_desc (type, data->hdr.algo);if (IS_ERR (desc))return PTR_ERR (desc); +/* +* HMACs are always SHA1, but signatures may be of varying sizes - +* make sure the caller knows how long the returned data is +*/ +if (type !=EVM_XATTR_HMAC ) + data->hdr.length =crypto_shash_digestsize (desc->tfm); + error = -ENODATA ;for (xattrname = evm_config_xattrnames; *xattrname !=NULL ; xattrname++) {bool is_ima = false;@@ -238,7 +240,7 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,if (is_ima) ima_present = true; } -hmac_add_misc (desc, inode, type, digest); +hmac_add_misc (desc, inode, type, data->digest);/* Portable EVM signatures must include an IMA hash */ if (type ==EVM_XATTR_PORTABLE_DIGSIG && !ima_present)@@ -251,18 +253,18 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,int evm_calc_hmac (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value,size_t req_xattr_value_len, -char *digest) +struct evm_digest *data) {return evm_calc_hmac_or_hash (dentry, req_xattr_name, req_xattr_value, - req_xattr_value_len,EVM_XATTR_HMAC , digest); + req_xattr_value_len,EVM_XATTR_HMAC , data); }int evm_calc_hash (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value,size_t req_xattr_value_len, -char type,char *digest) +char type,struct evm_digest *data) {return evm_calc_hmac_or_hash (dentry, req_xattr_name, req_xattr_value, - req_xattr_value_len, type, digest); + req_xattr_value_len, type, data); }static int evm_is_immutable (struct dentry *dentry,struct inode *inode)@@ -302,7 +304,7 @@ int evm_update_evmxattr(struct dentry *dentry, const char *xattr_name, const char *xattr_value, size_t xattr_value_len) {struct inode *inode =d_backing_inode (dentry); -struct evm_ima_xattr_data xattr_data; +struct evm_digest data;int rc =0 ;/* @@ -315,13 +317,14 @@ int evm_update_evmxattr(struct dentry *dentry, const char *xattr_name,if (rc)return -EPERM ; + data.hdr.algo =HASH_ALGO_SHA1 ; rc =evm_calc_hmac (dentry, xattr_name, xattr_value, - xattr_value_len, xattr_data.digest); + xattr_value_len, &data);if (rc ==0 ) { - xattr_data.type =EVM_XATTR_HMAC ; + data.hdr.xattr.sha1.type =EVM_XATTR_HMAC ; rc =__vfs_setxattr_noperm (dentry,XATTR_NAME_EVM , - &xattr_data, -sizeof (xattr_data),0 ); + &data.hdr.xattr.data[1 ], +SHA1_DIGEST_SIZE +1 ,0 ); }else if (rc == -ENODATA && (inode->i_opflags &IOP_XATTR )) { rc =__vfs_removexattr (dentry,XATTR_NAME_EVM ); }@@ -333,7 +336,7 @@ int evm_init_hmac(struct inode *inode, const struct xattr *lsm_xattr, {struct shash_desc *desc; - desc =init_desc (EVM_XATTR_HMAC ); + desc =init_desc (EVM_XATTR_HMAC ,HASH_ALGO_SHA1 );if (IS_ERR (desc)) {pr_info ("init_desc failed\n" );return PTR_ERR (desc);security/integrity/evm/evm_main.c +12
-7@@ -25,6 +25,7 @@#include <linux/magic.h> #include <crypto/hash.h> +#include <crypto/hash_info.h> #include <crypto/algapi.h> #include "evm.h" @@ -122,8 +123,9 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, struct integrity_iint_cache *iint) {struct evm_ima_xattr_data *xattr_data =NULL ; -struct evm_ima_xattr_data calc; +struct signature_v2_hdr *hdr;enum integrity_status evm_status =INTEGRITY_PASS ; +struct evm_digest digest;struct inode *inode;int rc, xattr_len;@@ -159,25 +161,28 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, evm_status =INTEGRITY_FAIL ; goto out; } + + digest.hdr.algo =HASH_ALGO_SHA1 ; rc =evm_calc_hmac (dentry, xattr_name, xattr_value, - xattr_value_len, calc.digest); + xattr_value_len, &digest);if (rc)break ; - rc =crypto_memneq (xattr_data->digest, calc.digest, -sizeof (calc.digest)); + rc =crypto_memneq (xattr_data->digest, digest.digest, +SHA1_DIGEST_SIZE );if (rc) rc = -EINVAL ;break ;case EVM_IMA_XATTR_DIGSIG :case EVM_XATTR_PORTABLE_DIGSIG : + hdr = (struct signature_v2_hdr *)xattr_data; + digest.hdr.algo = hdr->hash_algo; rc =evm_calc_hash (dentry, xattr_name, xattr_value, - xattr_value_len, xattr_data->type, - calc.digest); + xattr_value_len, xattr_data->type, &digest);if (rc)break ; rc =integrity_digsig_verify (INTEGRITY_KEYRING_EVM , (const char *)xattr_data, xattr_len, - calc.digest,sizeof (calc.digest)); + digest.digest, digest.hdr.length);if (!rc) { inode =d_backing_inode (dentry);
Oh bother - I think I see what's wrong. Does this version work better? I'm afraid I only tested against signatures rather than HMACs, and I was generating a raw SHA1 rather than an HMAC :(
security/integrity/evm/evm.h +8 -2
@@ -42,6 +42,11 @@ extern struct crypto_shash *hash_tfm;/* List of EVM protected security xattrs */ extern char *evm_config_xattrnames[]; +struct evm_digest { +struct ima_digest_data hdr; +char digest[IMA_MAX_DIGEST_SIZE ]; +} __packed; +int evm_init_key (void ); int evm_update_evmxattr(struct dentry *dentry, const char *req_xattr_name,
@@ -49,10 +54,11 @@ int evm_update_evmxattr(struct dentry *dentry, size_t req_xattr_value_len);int evm_calc_hmac (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char *digest); +size_t req_xattr_value_len,struct evm_digest *data);int evm_calc_hash (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char type,char *digest); +size_t req_xattr_value_len,char type, +struct evm_digest *data);int evm_init_hmac (struct inode *inode,const struct xattr *xattr,char *hmac_val);int evm_init_secfs (void );
security/integrity/evm/evm_crypto.c +25 -22
135 lines not on this page · raw message
security/integrity/evm/evm_main.c +12 -7
@@ -25,6 +25,7 @@#include <linux/magic.h> #include <crypto/hash.h> +#include <crypto/hash_info.h> #include <crypto/algapi.h> #include "evm.h"
@@ -122,8 +123,9 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, struct integrity_iint_cache *iint) {struct evm_ima_xattr_data *xattr_data =NULL ; -struct evm_ima_xattr_data calc; +struct signature_v2_hdr *hdr;enum integrity_status evm_status =INTEGRITY_PASS ; +struct evm_digest digest;struct inode *inode;int rc, xattr_len;
@@ -159,25 +161,28 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, evm_status =INTEGRITY_FAIL ; goto out; } + + digest.hdr.algo =HASH_ALGO_SHA1 ; rc =evm_calc_hmac (dentry, xattr_name, xattr_value, - xattr_value_len, calc.digest); + xattr_value_len, &digest);if (rc)break ; - rc =crypto_memneq (xattr_data->digest, calc.digest, -sizeof (calc.digest)); + rc =crypto_memneq (xattr_data->digest, digest.digest, +SHA1_DIGEST_SIZE );if (rc) rc = -EINVAL ;break ;case EVM_IMA_XATTR_DIGSIG :case EVM_XATTR_PORTABLE_DIGSIG : + hdr = (struct signature_v2_hdr *)xattr_data; + digest.hdr.algo = hdr->hash_algo; rc =evm_calc_hash (dentry, xattr_name, xattr_value, - xattr_value_len, xattr_data->type, - calc.digest); + xattr_value_len, xattr_data->type, &digest);if (rc)break ; rc =integrity_digsig_verify (INTEGRITY_KEYRING_EVM , (const char *)xattr_data, xattr_len, - calc.digest,sizeof (calc.digest)); + digest.digest, digest.hdr.length);if (!rc) { inode =d_backing_inode (dentry);
Signature · 1 line
2.17.0.441.gb46fe60e1d-goog
Quoted 3 lines · Matthew Garrett
Oh bother - I think I see what's wrong. Does this version work better? I'm afraid I only tested against signatures rather than HMACs, and I was generating a raw SHA1 rather than an HMAC :(
That's a lot better!
FYI, Wang Junwen reported a problem with enabling EVM with just the immutable and portable keys. Without trusted keys enabled, SHA1 isn't being built into the kernel. Loading the SHA1 kernel module fails.
Without knowing apriori which hash algorithms need to be builtin is a
problem.
Mimi
Quoted 240 lines
security/integrity/evm/evm.h +8
-2@@ -42,6 +42,11 @@ extern struct crypto_shash *hash_tfm;/* List of EVM protected security xattrs */ extern char *evm_config_xattrnames[]; +struct evm_digest { +struct ima_digest_data hdr; +char digest[IMA_MAX_DIGEST_SIZE ]; +} __packed; +int evm_init_key (void ); int evm_update_evmxattr(struct dentry *dentry, const char *req_xattr_name,@@ -49,10 +54,11 @@ int evm_update_evmxattr(struct dentry *dentry, size_t req_xattr_value_len);int evm_calc_hmac (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char *digest); +size_t req_xattr_value_len,struct evm_digest *data);int evm_calc_hash (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value, -size_t req_xattr_value_len,char type,char *digest); +size_t req_xattr_value_len,char type, +struct evm_digest *data);int evm_init_hmac (struct inode *inode,const struct xattr *xattr,char *hmac_val);int evm_init_secfs (void );security/integrity/evm/evm_crypto.c +25
-22@@ -21,6 +21,7 @@#include <linux/evm.h> #include <keys/encrypted-type.h> #include <crypto/hash.h> +#include <crypto/hash_info.h> #include "evm.h" #define EVMKEY "evm-key"@@ -29,7 +30,7 @@ static unsigned char evmkey[MAX_KEY_SIZE];static int evmkey_len =MAX_KEY_SIZE ;struct crypto_shash *hmac_tfm; -struct crypto_shash *hash_tfm; +static struct crypto_shash *evm_tfm[HASH_ALGO__LAST ];static DEFINE_MUTEX (mutex);@@ -38,7 +39,6 @@ static DEFINE_MUTEX(mutex);static unsigned long evm_set_key_flags;static char *const evm_hmac ="hmac(sha1)" ; -static char *const evm_hash ="sha1" ;/** * evm_set_key() - set EVM HMAC key from the kernel @@ -74,10 +74,10 @@ int evm_set_key(void *key, size_t keylen) }EXPORT_SYMBOL_GPL (evm_set_key); -static struct shash_desc *init_desc (char type) +static struct shash_desc *init_desc (char type,uint8_t hash_algo) {long rc; -char *algo; +const char *algo;struct crypto_shash **tfm;struct shash_desc *desc;@@ -89,8 +89,8 @@ static struct shash_desc *init_desc(char type) tfm = &hmac_tfm; algo = evm_hmac; }else { - tfm = &hash_tfm; - algo = evm_hash; + tfm = &evm_tfm[hash_algo]; + algo = hash_algo_name[hash_algo]; }if (*tfm ==NULL ) {@@ -186,10 +186,10 @@ static void hmac_add_misc(struct shash_desc *desc, struct inode *inode,* each xattr, but attempt to re-use the previously allocated memory. */ static int evm_calc_hmac_or_hash (struct dentry *dentry, -const char *req_xattr_name, -const char *req_xattr_value, -size_t req_xattr_value_len, -char type,char *digest) +const char *req_xattr_name, +const char *req_xattr_value, +size_t req_xattr_value_len, +uint8_t type,struct evm_digest *data) {struct inode *inode =d_backing_inode (dentry);struct shash_desc *desc;@@ -203,10 +203,12 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,if (!(inode->i_opflags &IOP_XATTR ))return -EOPNOTSUPP ; - desc =init_desc (type); + desc =init_desc (type, data->hdr.algo);if (IS_ERR (desc))return PTR_ERR (desc); + data->hdr.length =crypto_shash_digestsize (desc->tfm); + error = -ENODATA ;for (xattrname = evm_config_xattrnames; *xattrname !=NULL ; xattrname++) {bool is_ima = false;@@ -238,7 +240,7 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,if (is_ima) ima_present = true; } -hmac_add_misc (desc, inode, type, digest); +hmac_add_misc (desc, inode, type, data->digest);/* Portable EVM signatures must include an IMA hash */ if (type ==EVM_XATTR_PORTABLE_DIGSIG && !ima_present)@@ -251,18 +253,18 @@ static int evm_calc_hmac_or_hash(struct dentry *dentry,int evm_calc_hmac (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value,size_t req_xattr_value_len, -char *digest) +struct evm_digest *data) {return evm_calc_hmac_or_hash (dentry, req_xattr_name, req_xattr_value, - req_xattr_value_len,EVM_XATTR_HMAC , digest); + req_xattr_value_len,EVM_XATTR_HMAC , data); }int evm_calc_hash (struct dentry *dentry,const char *req_xattr_name,const char *req_xattr_value,size_t req_xattr_value_len, -char type,char *digest) +char type,struct evm_digest *data) {return evm_calc_hmac_or_hash (dentry, req_xattr_name, req_xattr_value, - req_xattr_value_len, type, digest); + req_xattr_value_len, type, data); }static int evm_is_immutable (struct dentry *dentry,struct inode *inode)@@ -302,7 +304,7 @@ int evm_update_evmxattr(struct dentry *dentry, const char *xattr_name, const char *xattr_value, size_t xattr_value_len) {struct inode *inode =d_backing_inode (dentry); -struct evm_ima_xattr_data xattr_data; +struct evm_digest data;int rc =0 ;/* @@ -315,13 +317,14 @@ int evm_update_evmxattr(struct dentry *dentry, const char *xattr_name,if (rc)return -EPERM ; + data.hdr.algo =HASH_ALGO_SHA1 ; rc =evm_calc_hmac (dentry, xattr_name, xattr_value, - xattr_value_len, xattr_data.digest); + xattr_value_len, &data);if (rc ==0 ) { - xattr_data.type =EVM_XATTR_HMAC ; + data.hdr.xattr.sha1.type =EVM_XATTR_HMAC ; rc =__vfs_setxattr_noperm (dentry,XATTR_NAME_EVM , - &xattr_data, -sizeof (xattr_data),0 ); + &data.hdr.xattr.data[1 ], +SHA1_DIGEST_SIZE +1 ,0 ); }else if (rc == -ENODATA && (inode->i_opflags &IOP_XATTR )) { rc =__vfs_removexattr (dentry,XATTR_NAME_EVM ); }@@ -333,7 +336,7 @@ int evm_init_hmac(struct inode *inode, const struct xattr *lsm_xattr, {struct shash_desc *desc; - desc =init_desc (EVM_XATTR_HMAC ); + desc =init_desc (EVM_XATTR_HMAC ,HASH_ALGO_SHA1 );if (IS_ERR (desc)) {pr_info ("init_desc failed\n" );return PTR_ERR (desc);security/integrity/evm/evm_main.c +12
-7@@ -25,6 +25,7 @@#include <linux/magic.h> #include <crypto/hash.h> +#include <crypto/hash_info.h> #include <crypto/algapi.h> #include "evm.h" @@ -122,8 +123,9 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, struct integrity_iint_cache *iint) {struct evm_ima_xattr_data *xattr_data =NULL ; -struct evm_ima_xattr_data calc; +struct signature_v2_hdr *hdr;enum integrity_status evm_status =INTEGRITY_PASS ; +struct evm_digest digest;struct inode *inode;int rc, xattr_len;@@ -159,25 +161,28 @@ static enum integrity_status evm_verify_hmac(struct dentry *dentry, evm_status =INTEGRITY_FAIL ; goto out; } + + digest.hdr.algo =HASH_ALGO_SHA1 ; rc =evm_calc_hmac (dentry, xattr_name, xattr_value, - xattr_value_len, calc.digest); + xattr_value_len, &digest);if (rc)break ; - rc =crypto_memneq (xattr_data->digest, calc.digest, -sizeof (calc.digest)); + rc =crypto_memneq (xattr_data->digest, digest.digest, +SHA1_DIGEST_SIZE );if (rc) rc = -EINVAL ;break ;case EVM_IMA_XATTR_DIGSIG :case EVM_XATTR_PORTABLE_DIGSIG : + hdr = (struct signature_v2_hdr *)xattr_data; + digest.hdr.algo = hdr->hash_algo; rc =evm_calc_hash (dentry, xattr_name, xattr_value, - xattr_value_len, xattr_data->type, - calc.digest); + xattr_value_len, xattr_data->type, &digest);if (rc)break ; rc =integrity_digsig_verify (INTEGRITY_KEYRING_EVM , (const char *)xattr_data, xattr_len, - calc.digest,sizeof (calc.digest)); + digest.digest, digest.hdr.length);if (!rc) { inode =d_backing_inode (dentry);
Quoted 5 lines · Mimi Zohar
Quoted 3 lines · Matthew Garrett
Oh bother - I think I see what's wrong. Does this version work better? I'm afraid I only tested against signatures rather than HMACs, and I was generating a raw SHA1 rather than an HMAC :(
Quoted 7 lines
That's a lot better!
FYI, Wang Junwen reported a problem with enabling EVM with just the immutable and portable keys. Without trusted keys enabled, SHA1 isn't being built into the kernel. Loading the SHA1 kernel module fails.
Without knowing apriori which hash algorithms need to be builtin is a
problem.
It looks like Kconfig is selecting CRYPTO_SHA1 when EVM is enabled, and since that's a bool it should be forcing it to be built-in? I can't see a good way of extending that generally, unfortunately. Is the problem with loading the module that you're enforcing an IMA policy before loading it?
A message that is not in the archive (sent off-list or not yet received): 15252CF8C1B4384C8CE16D7D55C66479011414E7BF@BC-MAIL-M04.internal.baidu.com
Quoted 4 lines · Wang,Junwen
if we need fix this problem
1. load the hash algorithm at initial time instead of runtime OR
2. avoid thecrypto_alloc_shashtry to load modules ininit_desc
The outcome here is presumably going to be failure regardless - if appraisal is required and the hash module is unavailable, failing to load the module won't result in deadlock but will result in an unusable machine? I think the only way this can work is to ensure the crypto modules are available before a policy is enabled, but let me look to see if there's a way to at least make the failure clean and more debuggable.
PATCH 1/1 evm: Don't deadlock if a crypto algorithm is unavailable
Does this help?
Trying to instantiate a non-existent crypto algorithm will cause the kernel to trigger a module load. If EVM appraisal is enabled, this will in turn trigger appraisal of the module, which will fail because the crypto algorithm isn't available. Add a CRYPTO_NOLOAD flag and skip module loading if it's set, and add that flag in the EVM case.
--- crypto/api.c | 2 +- include/linux/crypto.h | 5 +++++ security/integrity/evm/evm_crypto.c | 3 ++- 3 files changed, 8 insertions(+), 2 deletions(-)
crypto/api.c +1 -1
@@ -229,7 +229,7 @@ static struct crypto_alg *crypto_larval_lookup(const char *name, u32 type, mask &= ~(CRYPTO_ALG_LARVAL |CRYPTO_ALG_DEAD ); alg =crypto_alg_lookup (name, type, mask); -if (!alg) { +if (!alg && !(mask &CRYPTO_NOLOAD )) {request_module ("crypto-%s" , name); if (!((type ^ CRYPTO_ALG_NEED_FALLBACK) & mask &
include/linux/crypto.h +5 -0
@@ -112,6 +112,11 @@*/ #define CRYPTO_ALG_OPTIONAL_KEY 0x00004000 +/* +* Don't trigger module loading +*/ +#define CRYPTO_NOLOAD 0x00008000 +/* * Transform masks and values (for crt_flags). */
security/integrity/evm/evm_crypto.c +2 -1
@@ -97,7 +97,8 @@ static struct shash_desc *init_desc(char type, uint8_t hash_algo)mutex_lock (&mutex);if (*tfm) goto out; - *tfm =crypto_alloc_shash (algo,0 ,CRYPTO_ALG_ASYNC ); + *tfm =crypto_alloc_shash (algo,0 , +CRYPTO_ALG_ASYNC |CRYPTO_NOLOAD );if (IS_ERR (*tfm)) { rc =PTR_ERR (*tfm);pr_err ("Can not allocate %s (reason: %ld)\n" , algo, rc);
Signature · 1 line
2.17.1.1185.g55be947832-goog
A message that is not in the archive (sent off-list or not yet received): 15252CF8C1B4384C8CE16D7D55C66479011414E83B@BC-MAIL-M04.internal.baidu.com
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Cool. Mimi, does this satisfy your concerns with the non-SHA1 patch? If so I'll resend with a s-o-b
Quoted 69 lines · Wang,Junwen
this pretty good, now
ima_appraise& evm working well on my machine thank you, Matthew :)
________________________________________
???: Matthew Garrett [[email protected]]
????: 2018?5?31? 4:28
???: [email protected]
??: [email protected]; Matthew Garrett; Wang,Junwen ??: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailableDoes this help?
Trying to instantiate a non-existent crypto algorithm will cause the kernel to trigger a module load. If EVM appraisal is enabled, this will in turn trigger appraisal of the module, which will fail because the crypto algorithm isn't available. Add a
CRYPTO_NOLOADflag and skip module loading if it's set, and add that flag in the EVM case.--- crypto/api.c | 2 +- include/linux/crypto.h | 5 +++++ security/integrity/evm/evm_crypto.c | 3 ++- 3 files changed, 8 insertions(+), 2 deletions(-)crypto/api.c +1
-1@@ -229,7 +229,7 @@ static struct crypto_alg *crypto_larval_lookup(const char *name, u32 type, mask &= ~(CRYPTO_ALG_LARVAL |CRYPTO_ALG_DEAD ); alg =crypto_alg_lookup (name, type, mask); -if (!alg) { +if (!alg && !(mask &CRYPTO_NOLOAD )) {request_module ("crypto-%s" , name); if (!((type ^ CRYPTO_ALG_NEED_FALLBACK) & mask &include/linux/crypto.h +5
-0@@ -112,6 +112,11 @@*/ #define CRYPTO_ALG_OPTIONAL_KEY 0x00004000 +/* +* Don't trigger module loading +*/ +#define CRYPTO_NOLOAD 0x00008000 +/* * Transform masks and values (for crt_flags). */ security/integrity/evm/evm_crypto.c +2
-2@@ -97,7 +97,8 @@ static struct shash_desc *init_desc(char type, uint8_t hash_algo)mutex_lock (&mutex);if (*tfm) goto out; - *tfm =crypto_alloc_shash (algo,0 ,CRYPTO_ALG_ASYNC ); + *tfm =crypto_alloc_shash (algo,0 , +CRYPTO_ALG_ASYNC |CRYPTO_NOLOAD );if (IS_ERR (*tfm)) { rc =PTR_ERR (*tfm);pr_err ("Can not allocate %s (reason: %ld)\n" , algo, rc); --2.17.1.1185.g55be947832-goog
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 1 line · Matthew Garrett
Cool. Mimi, does this satisfy your concerns with the non-SHA1 patch?
This should be posted as a separate patch, with the appropriate "Fixes" commit info. It requires an ack from Herbert Xu.
thanks,
Mimi
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 6 lines · Mimi Zohar
Quoted 1 line · Matthew Garrett
Cool. Mimi, does this satisfy your concerns with the non-SHA1 patch?
This should be posted as a separate patch, with the appropriate "Fixes" commit info. It requires an ack from Herbert Xu.
The non-sha1 patch doesn't seem to be in -next at the moment - should I wait for you to merge that so I can add the fixes tag?
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 10 lines · Matthew Garrett
Quoted 6 lines · Mimi Zohar
Quoted 1 line · Matthew Garrett
Cool. Mimi, does this satisfy your concerns with the non-SHA1 patch?
This should be posted as a separate patch, with the appropriate "Fixes" commit info. It requires an ack from Herbert Xu.
The non-sha1 patch doesn't seem to be in -next at the moment - should I wait for you to merge that so I can add the fixes tag?
The problem exists prior to the non-sha1 patch, but could have been resolved differently for only sha1 (eg. Kconfig requiring sha1). The non-sha1 patch requires a different solution.
Junwen, is this a regression? Did a particular patch introduce this problem?
Mimi
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 10 lines · Mimi Zohar
Quoted 8 lines · Matthew Garrett
Quoted 6 lines · Mimi Zohar
Quoted 1 line · Matthew Garrett
Cool. Mimi, does this satisfy your concerns with the non-SHA1 patch?
This should be posted as a separate patch, with the appropriate "Fixes" commit info. It requires an ack from Herbert Xu.
Quoted 6 lines
Quoted 2 lines
The non-sha1 patch doesn't seem to be in -next at the moment - should I wait for you to merge that so I can add the fixes tag?
The problem exists prior to the non-sha1 patch, but could have been resolved differently for only sha1 (eg.
Kconfigrequiring sha1). The non-sha1 patch requires a different solution.
EVM looks like it SELECTs CONFIG_SHA1, so I /think/ it should be ok before that patch?
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 14 lines · Matthew Garrett
Quoted 13 lines · Mimi Zohar
Quoted 10 lines · Matthew Garrett
Quoted 6 lines · Mimi Zohar
Quoted 1 line · Matthew Garrett
Cool. Mimi, does this satisfy your concerns with the non-SHA1 patch?
This should be posted as a separate patch, with the appropriate "Fixes" commit info. It requires an ack from Herbert Xu.
The non-sha1 patch doesn't seem to be in -next at the moment - should I wait for you to merge that so I can add the fixes tag?
Quoted 6 lines
Quoted 3 lines
The problem exists prior to the non-sha1 patch, but could have been resolved differently for only sha1 (eg.
Kconfigrequiring sha1). The non-sha1 patch requires a different solution.EVM looks like it SELECTs
CONFIG_SHA1, so I /think/ it should be ok before that patch?
According to Junwen, with CONFIG_TRUSTED_KEYS enabled the HMAC and SHA1 are allocated at __init. The locking problem occurs when CONFIG_TRUSTED_KEYS is not enabled. His solution would have been to move the crypto_alloc_shash() in EVM to an __init function.
Mimi
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 8 lines · Mimi Zohar
Quoted 2 lines · Matthew Garrett
EVM looks like it SELECTs
CONFIG_SHA1, so I /think/ it should be ok before that patch?According to Junwen, with
CONFIG_TRUSTED_KEYSenabled the HMAC and SHA1 are allocated at__init. The locking problem occurs whenCONFIG_TRUSTED_KEYSis not enabled. His solution would have been to move thecrypto_alloc_shash()in EVM to an__initfunction.
Ok - I think just allowing it to be deferred is preferable, since otherwise we'd have to build in every hash algorithm that could be used for the signatures (which wasn't a problem before the non-sha1 patch). How would you prefer me to send these two? The non-sha1 patch isn't in -next, so I can't add a fixes: for it at this point.
Re: [PATCH] evm: Don't deadlock if a crypto algorithm is unavailable
Quoted 10 lines · Matthew Garrett
Quoted 8 lines · Mimi Zohar
Quoted 2 lines · Matthew Garrett
EVM looks like it SELECTs
CONFIG_SHA1, so I /think/ it should be ok before that patch?According to Junwen, with
CONFIG_TRUSTED_KEYSenabled the HMAC and SHA1 are allocated at__init. The locking problem occurs whenCONFIG_TRUSTED_KEYSis not enabled. His solution would have been to move thecrypto_alloc_shash()in EVM to an__initfunction.
Quoted 5 lines
Ok - I think just allowing it to be deferred is preferable, since otherwise we'd have to build in every hash algorithm that could be used for the signatures (which wasn't a problem before the non-sha1 patch). How would you prefer me to send these two? The non-sha1 patch isn't in -next, so I can't add a fixes: for it at this point.
Switch the order of the two patches. The bug fix goes first, then the non-sha1 patch. Remember we need an Ack from Herbert Xu for introducing CRYPTO_NOLOAD.
Mimi