Hi Dan,
On Tue, 2018-05-29 at 12:05 +0300, Dan Carpenter wrote:
> Not really related to this patch except I was looking at the function:
>
> security/integrity/evm/evm_secfs.c
> 191 ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_INTEGRITY_EVM_XATTR);
> 192 if (IS_ERR(ab))
> 193 return PTR_ERR(ab);
> 194
> 195 xattr = kmalloc(sizeof(struct xattr_list), GFP_KERNEL);
> 196 if (!xattr) {
> 197 err = -ENOMEM;
> 198 goto out;
> 199 }
> 200
> 201 xattr->name = memdup_user_nul(buf, count);
> 202 if (IS_ERR(xattr->name)) {
> 203 err = PTR_ERR(xattr->name);
> 204 xattr->name = NULL;
> 205 goto out;
> 206 }
> 207
> 208 /* Remove any trailing newline */
> 209 len = strlen(xattr->name);
> 210 if (xattr->name[len-1] == '\n')
>
> strlen() could be zero, leading to a read underflow here.
Thanks! Could you modify the maximum xattr size check (before this
code snippet) to check for underflow?
Mimi
>
> 211 xattr->name[len-1] = '\0';
> 212
> 213 if (strcmp(xattr->name, ".") == 0) {
> 214 evm_xattrs_locked = 1;
> 215 newattrs.ia_mode = S_IFREG | 0440;
> 216 newattrs.ia_valid = ATTR_MODE;
> 217 inode = evm_xattrs->d_inode;
> 218 inode_lock(inode);
> 219 err = simple_setattr(evm_xattrs, &newattrs);
> 220 inode_unlock(inode);
> 221 audit_log_format(ab, "locked");
> 222 if (!err)
> 223 err = count;
> 224 goto out;
> 225 }
>
> regards,
> dan carpenter
>
--
To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
