Add EC_POINT_hex2point #7191

kojo1 · 2024-01-30T21:53:15Z

Description

Add EC_POINT_hex2point
Enable HAVE_COMP_KEY for Compressed Key

Fixes zd #17090

Testing

compressed and uncompressed in test_wolfSSL_EC_POINT, api.c

Checklist

added tests
updated/added doxygen
updated appropriate READMEs
Updated manual and documentation

src/pk.c

bandi13 · 2024-01-31T14:23:40Z

retest this please..

ejohnstown · 2024-01-31T19:57:35Z

I ran this locally with Valgrind and the "CI/krb5/Build wolfSSL" test failure is legit.

  ./configure --enable-opensslall --disable-shared && make clean && make
  valgrind --leak-test=full ./tests/unit.test

I got the same leak report for function wolfSSL_EC_POINT_hex2point.

I looked into the "CI/openwrt/Compile container (21.02-SNAPSHOT)" failure. It isn't able to download the openwrt sources from the server.

bandi13 · 2024-01-31T21:56:47Z

Looks like there's actually an issue here with missing macro definitions. The openwrt / Compile container (21.02-SNAPSHOT) issue has already been resolved (see PR#7193)

kojo1 · 2024-02-02T00:38:39Z

retest this please

dgarske · 2024-05-15T17:45:45Z

src/pk.c

+            WOLFSSL_MSG("EEC_KEY_get_byte_size, XMALLOC");
+            goto err;
+        }
+        octGx[0] = 0x03;


Can you use ECC_POINT_COMP_ODD?

dgarske · 2024-05-15T17:50:34Z

src/pk.c

+    int    key_sz;
+    byte   *octGx = NULL;
+
+    #define P_ALLOC 1


Adding P_ALLOC here seems pointless. Just use 1 directly.

dgarske · 2024-05-15T17:53:58Z

src/pk.c

+    }
+    else if (hex[0] == '0' && (hex[1] == '2' || hex[1] == '3')) {
+        /* compressed mode */
+        key_sz = ((wolfSSL_EC_GROUP_get_degree(group) + 7) / 8);


Recommend a cleanup to do key_sz = ((wolfSSL_EC_GROUP_get_degree(group) + 7) / 8); above.

dgarske · 2024-05-15T17:54:12Z

src/pk.c

+    }
+
+    if (hex[0] ==  '0' && hex[1] == '4') { /* uncompressed mode */
+        str_sz = ((wolfSSL_EC_GROUP_get_degree(group) + 7) / 8) * 2;


str_sz = key_sz * 2.

dgarske · 2024-05-15T17:55:28Z

src/pk.c

+
+    if (hex[0] ==  '0' && hex[1] == '4') { /* uncompressed mode */
+        str_sz = ((wolfSSL_EC_GROUP_get_degree(group) + 7) / 8) * 2;
+        strGx = (char *)XMALLOC(str_sz + 1, NULL, DYNAMIC_TYPE_ECC);


It might be easier to just use MAX_ECC_BYTES * 2 for the buffer vs using malloc without a heap hint. Its not really an ECC type, its temp buffer.

dgarske · 2024-05-15T17:55:48Z

src/pk.c

+    else if (hex[0] == '0' && (hex[1] == '2' || hex[1] == '3')) {
+        /* compressed mode */
+        key_sz = ((wolfSSL_EC_GROUP_get_degree(group) + 7) / 8);
+        octGx = (byte *)XMALLOC(key_sz + 1, NULL, DYNAMIC_TYPE_ECC);


Same use MAX_ECC_BYTES and only use a single buffer for either case.

kojo1 · 2024-05-22T21:42:41Z

retest this please

dgarske · 2024-05-22T21:50:18Z

src/pk.c

@@ -9788,12 +9787,12 @@ char* wolfSSL_EC_POINT_point2hex(const WOLFSSL_EC_GROUP* group,
             * odd.
             */
            hex[0] = mp_isodd((mp_int*)point->Y->internal) ?
-                ECC_POINT_COMP_ODD : ECC_POINT_COMP_EVEN;
+                0x03 : 0x02;


Did you switch to using the numeric directly to avoid having the #fndef HAVE_SELFTEST?

I really prefer the ECC_POINT_COMP_ODD, ECC_POINT_COMP_EVEN and ECC_POINT_UNCOMP values. If needed you could add some logic above to ensure these are available.

#ifdef HAVE_SELFTEST /* point compression types */ #define ECC_POINT_COMP_EVEN 0x02 #define ECC_POINT_COMP_ODD 0x03 #define ECC_POINT_UNCOMP 0x04 #endif

dgarske · 2024-05-30T22:08:52Z

Retest this please

tests/api.c

kojo1 · 2024-06-01T00:18:38Z

retest this please

Add EC_POINT_hex2point

kojo1 force-pushed the ecpoint-h2p branch 2 times, most recently from 2dd78b3 to 890f46f Compare January 30, 2024 22:29

SparkiDev requested changes Jan 30, 2024

View reviewed changes

src/pk.c Outdated Show resolved Hide resolved

src/pk.c Outdated Show resolved Hide resolved

kojo1 force-pushed the ecpoint-h2p branch 2 times, most recently from 6ac25c2 to 1259e4a Compare January 31, 2024 04:39

kojo1 force-pushed the ecpoint-h2p branch 8 times, most recently from a80a0c5 to d7e7168 Compare February 1, 2024 08:16

Add EC_POINT_hex2point

791c9e7

kojo1 force-pushed the ecpoint-h2p branch from d7e7168 to 791c9e7 Compare February 1, 2024 22:40

kojo1 requested a review from SparkiDev February 2, 2024 01:06

dgarske assigned wolfSSL-Bot May 15, 2024

dgarske requested changes May 15, 2024

View reviewed changes

dgarske assigned kojo1 May 15, 2024

kojo1 force-pushed the ecpoint-h2p branch from dd66aa7 to a5da2a5 Compare May 21, 2024 00:56

dgarske self-requested a review May 22, 2024 21:47

dgarske requested changes May 22, 2024

View reviewed changes

kojo1 force-pushed the ecpoint-h2p branch 3 times, most recently from e760f12 to cd2818a Compare May 26, 2024 22:30

dgarske assigned kojo1 and unassigned kojo1 and SparkiDev May 29, 2024

SparkiDev previously approved these changes May 29, 2024

View reviewed changes

kojo1 dismissed SparkiDev’s stale review via 5b45fe6 May 30, 2024 04:58

kojo1 force-pushed the ecpoint-h2p branch from 7839f39 to 5b45fe6 Compare May 30, 2024 04:58

dgarske self-assigned this May 30, 2024

dgarske requested review from dgarske and SparkiDev May 30, 2024 13:18

dgarske approved these changes May 30, 2024

View reviewed changes

dgarske requested changes May 30, 2024

View reviewed changes

tests/api.c Show resolved Hide resolved

dgarske removed their assignment May 30, 2024

dgarske removed the request for review from SparkiDev May 30, 2024 22:42

dgarske assigned dgarske and unassigned wolfSSL-Bot May 31, 2024

kojo1 force-pushed the ecpoint-h2p branch 3 times, most recently from 2791ba9 to 73b35f1 Compare June 1, 2024 02:07

Add EC_POINT_hex2point: zd #17090

bc2b184

kojo1 force-pushed the ecpoint-h2p branch from 73b35f1 to bc2b184 Compare June 1, 2024 04:45

dgarske assigned wolfSSL-Bot and SparkiDev and unassigned dgarske Jun 1, 2024

dgarske requested review from SparkiDev and dgarske June 1, 2024 05:08

dgarske approved these changes Jun 1, 2024

View reviewed changes

dgarske merged commit 3975af8 into wolfSSL:master Jun 1, 2024
103 checks passed

jefferyq2 pushed a commit to jefferyq2/wolfssl that referenced this pull request Jun 9, 2024

Merge pull request wolfSSL#7191 from kojo1/ecpoint-h2p

97e42ca

Add EC_POINT_hex2point

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add EC_POINT_hex2point #7191

Add EC_POINT_hex2point #7191

kojo1 commented Jan 30, 2024

bandi13 commented Jan 31, 2024 •

edited

Loading

ejohnstown commented Jan 31, 2024

bandi13 commented Jan 31, 2024

kojo1 commented Feb 2, 2024

dgarske May 15, 2024

kojo1 May 20, 2024

dgarske May 15, 2024

kojo1 May 20, 2024

dgarske May 15, 2024

kojo1 May 20, 2024

dgarske May 15, 2024

kojo1 May 20, 2024

dgarske May 15, 2024

kojo1 May 20, 2024

dgarske May 15, 2024

kojo1 May 20, 2024

kojo1 commented May 22, 2024

dgarske May 22, 2024

dgarske commented May 30, 2024

kojo1 commented Jun 1, 2024

Add EC_POINT_hex2point #7191

Add EC_POINT_hex2point #7191

Conversation

kojo1 commented Jan 30, 2024

Description

Testing

Checklist

bandi13 commented Jan 31, 2024 • edited Loading

ejohnstown commented Jan 31, 2024

bandi13 commented Jan 31, 2024

kojo1 commented Feb 2, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kojo1 commented May 22, 2024

Choose a reason for hiding this comment

dgarske commented May 30, 2024

kojo1 commented Jun 1, 2024

bandi13 commented Jan 31, 2024 •

edited

Loading