RxD Fix by ejohnstown · Pull Request #874 · wolfSSL/wolfssh

ejohnstown · 2026-02-02T19:18:10Z

Forwarding fixes in echoserver. (Really a general channel issue.)
Agent fixes in echoserver.
Added a test for forwarding.

Copilot

Pull request overview

Fixes handling around “channel RX’d” (WS_CHAN_RXD) and agent setup paths, primarily to improve forwarding behavior in the echoserver and make agent-related code compile cleanly.

Changes:

Treat WS_CHAN_RXD as a successful outcome for returning lastRxId from wolfSSH_worker().
Silence unused-parameter warnings in agent stubs.
Fix echoserver agent UNIX-socket setup flow so it proceeds correctly after snprintf() and simplifies the socket() error assignment.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
`src/ssh.c`	Returns `channelId` not only on `WS_SUCCESS` but also on `WS_CHAN_RXD` so callers can identify the channel that received data.
`src/agent.c`	Adds `WOLFSSH_UNUSED(agent)` to avoid unused-parameter warnings when logging is compiled out.
`examples/echoserver/echoserver.c`	Corrects agent local setup flow after `snprintf()` and adjusts UNIX-socket bind/setup logic.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

examples/echoserver/echoserver.c

Copilot

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 2 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

examples/echoserver/echoserver.c

JacobBarthelmeh

Changes look okay. I sent a message asking about expected output and return values from running the new scripts/fwd.test.

JacobBarthelmeh · 2026-02-17T17:46:24Z

Assigning to John for investigating into the test case behavior.

1. When wolfSSH_worker() receives channel data, it should set the channelId for the data. It was not happening. Change the check for WS_SUCCESS to also check for WS_CHAN_RXD.

1. Fix a couple unused variable warnings. 2. In wolfSSH_AGENT_DefaultActions(), fix comparison to the result of snprintf() treating normal result as an error. Reset the return code for the error state of the socket() command. Remove the size variable and just use sizeof() the sockaddr_un. Better cleanup of agent startup failures.

1. Add a test script and expect script for testing forwarding. 2. Update portfwd to have a ready file option. 3. Fix echoserver error string, needed NL.

Copilot

Pull request overview

Copilot reviewed 7 out of 7 changed files in this pull request and generated 5 comments.

Comments suppressed due to low confidence (1)

examples/portfwd/portfwd.c:325

The new -R <readyFile> option is parsed, but ShowUsage() doesn’t mention it. This makes the CLI help inaccurate; please update the usage text to document -R and what it writes to the file.

    while ((ch = mygetopt(argc, argv, "?f:h:p:t:u:F:P:R:T:")) != -1) {
        switch (ch) {
            case 'h':
                host = myoptarg;
                break;

            case 'f':
                if (myoptarg == NULL)
                    err_sys("null argument found");
                fwdFromPort = (word16)atoi(myoptarg);
                break;

            case 'p':
                if (myoptarg == NULL)
                    err_sys("null argument found");
                port = (word16)atoi(myoptarg);
                #if !defined(NO_MAIN_DRIVER) || defined(USE_WINDOWS_API)
                    if (port == 0)
                        err_sys("port number cannot be 0");
                #endif
                break;

            case 't':
                if (myoptarg == NULL)
                    err_sys("null argument found");
                fwdToPort = (word16)atoi(myoptarg);
                break;

            case 'u':
                username = myoptarg;
                break;

            case 'F':
                fwdFromHost = myoptarg;
                break;

            case 'P':
                password = myoptarg;
                break;

            case 'R':
                readyFile = myoptarg;
                break;

            case 'T':
                fwdToHost = myoptarg;
                break;

            case '?':
                ShowUsage();
                exit(EXIT_SUCCESS);

            default:
                ShowUsage();
                exit(MY_EX_USAGE);
        }

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-26T00:11:02Z

scripts/include.am

 endif

-dist_noinst_SCRIPTS+= scripts/external.test
+dist_noinst_SCRIPTS+= scripts/external.test scripts/fwd.test


scripts/fwd.test depends on scripts/fwd.test.expect, but include.am only adds scripts/fwd.test to dist_noinst_SCRIPTS. This will break make distcheck / running tests from a release tarball because the .expect file won’t be distributed. Add scripts/fwd.test.expect to the dist list (or another appropriate dist_* variable) alongside scripts/fwd.test.

Suggested change

dist_noinst_SCRIPTS+= scripts/external.test scripts/fwd.test

dist_noinst_SCRIPTS+= scripts/external.test scripts/fwd.test scripts/fwd.test.expect

Copilot · 2026-02-26T00:11:03Z

examples/portfwd/portfwd.c

+            int l = WSNPRINTF(portStr, sizeof(portStr), "%d\n", (int)port);
+            WFWRITE(NULL, portStr, MIN((size_t)l, sizeof(portStr)), 1, f);


The ready-file writer is currently formatting/writing port (the SSH server port) rather than the local forwarded-from port that clients would need (likely fwdFromPort, which may be updated by tcp_listen() when binding ephemeral ports). Also, WSNPRINTF() can return a negative value on error; casting that to size_t can cause WFWRITE() to write an unintended length. Write the correct port value and guard against l <= 0 (and ideally handle a truncated result).

Suggested change

int l = WSNPRINTF(portStr, sizeof(portStr), "%d\n", (int)port);

WFWRITE(NULL, portStr, MIN((size_t)l, sizeof(portStr)), 1, f);

int l = WSNPRINTF(portStr, sizeof(portStr), "%d\n", (int)fwdFromPort);

if (l > 0) {

size_t writeLen = (size_t)l;

if (writeLen >= sizeof(portStr)) {

writeLen = sizeof(portStr) - 1;

}

WFWRITE(NULL, portStr, writeLen, 1, f);

}

Copilot · 2026-02-26T00:11:03Z

examples/echoserver/echoserver.c

+            if (envSet) {
+                unsetenv(EnvNameAuthPort);
+            }
+            if (ctx->listenFd >= 0) {
+                close(ctx->listenFd);
+                ctx->listenFd = -1;
+            }


If bind() succeeds but setenv() or listen() fails, the setup error path closes the FD and unsets the env var, but it does not remove the already-created UNIX socket pathname. Add an unlink(ctx->name.sun_path) (or unlink(name->sun_path)) in the failure cleanup to avoid leaving stale /tmp/wolfserver.<pid> sockets behind.

Copilot · 2026-02-26T00:11:03Z

examples/echoserver/echoserver.c

        if (ret == 0) {
-            ret = bind(ctx->listenFd,
-                    (struct sockaddr *)name, (socklen_t)size);
+            ret = bind(ctx->listenFd, (struct sockaddr *)name,
+                    (socklen_t)sizeof(struct sockaddr_un));
        }


The bind() length was changed to sizeof(struct sockaddr_un), but elsewhere (e.g. the client-side agent code) the code computes strlen(path) + offsetof(sockaddr_un, sun_path). Using the full struct size can be less portable across UNIX variants; consider restoring the computed length here for consistency and compatibility.

Copilot · 2026-02-26T00:11:03Z

examples/echoserver/echoserver.c

 #ifdef WOLFSSH_TEST_BLOCK
    if (!nonBlock) {
-        ES_ERROR("Use -N when testing forced non blocking");
+        ES_ERROR("Use -N when testing forced non blocking\n");


Hyphenation in this user-facing message is inconsistent with the common term “non-blocking”. Update the wording to “forced non-blocking”.

Suggested change

ES_ERROR("Use -N when testing forced non blocking\n");

ES_ERROR("Use -N when testing forced non-blocking\n");

ejohnstown self-assigned this Feb 2, 2026

ejohnstown force-pushed the rxd-fix branch 2 times, most recently from 69e1d55 to fc39ced Compare February 3, 2026 17:26

ejohnstown requested review from JacobBarthelmeh and Copilot February 4, 2026 03:41

Copilot started reviewing on behalf of ejohnstown February 4, 2026 03:42 View session

Copilot AI reviewed Feb 4, 2026

View reviewed changes

examples/echoserver/echoserver.c Show resolved Hide resolved

ejohnstown force-pushed the rxd-fix branch from fc39ced to f18c692 Compare February 4, 2026 03:52

ejohnstown assigned wolfSSL-Bot and unassigned ejohnstown Feb 5, 2026

ejohnstown force-pushed the rxd-fix branch 4 times, most recently from 676df52 to 773c3bf Compare February 12, 2026 21:48

dgarske assigned JacobBarthelmeh Feb 16, 2026

JacobBarthelmeh requested a review from Copilot February 16, 2026 21:01

Copilot AI reviewed Feb 16, 2026

View reviewed changes

examples/echoserver/echoserver.c Show resolved Hide resolved

examples/echoserver/echoserver.c Show resolved Hide resolved

Copilot started reviewing on behalf of JacobBarthelmeh February 16, 2026 21:38 View session

JacobBarthelmeh reviewed Feb 16, 2026

View reviewed changes

JacobBarthelmeh assigned ejohnstown and unassigned JacobBarthelmeh and wolfSSL-Bot Feb 17, 2026

ejohnstown force-pushed the rxd-fix branch from 773c3bf to cbc787b Compare February 24, 2026 18:38

ejohnstown added 3 commits February 24, 2026 11:12

RxD Fix

e47eb8a

1. When wolfSSH_worker() receives channel data, it should set the channelId for the data. It was not happening. Change the check for WS_SUCCESS to also check for WS_CHAN_RXD.

Forwarding Test

a097eea

1. Add a test script and expect script for testing forwarding. 2. Update portfwd to have a ready file option. 3. Fix echoserver error string, needed NL.

ejohnstown force-pushed the rxd-fix branch from cbc787b to a097eea Compare February 24, 2026 21:34

ejohnstown assigned wolfSSL-Bot and unassigned ejohnstown Feb 25, 2026

ejohnstown requested a review from JacobBarthelmeh February 25, 2026 01:00

JacobBarthelmeh requested a review from Copilot February 26, 2026 00:05

Copilot started reviewing on behalf of JacobBarthelmeh February 26, 2026 00:06 View session

Copilot AI reviewed Feb 26, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

RxD Fix#874

RxD Fix#874
ejohnstown wants to merge 3 commits intowolfSSL:masterfrom
ejohnstown:rxd-fix

ejohnstown commented Feb 2, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

JacobBarthelmeh left a comment

Uh oh!

JacobBarthelmeh commented Feb 17, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Feb 26, 2026

Uh oh!

Copilot AI Feb 26, 2026

Uh oh!

Copilot AI Feb 26, 2026

Uh oh!

Copilot AI Feb 26, 2026

Uh oh!

Copilot AI Feb 26, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

	dist_noinst_SCRIPTS+= scripts/external.test scripts/fwd.test
	dist_noinst_SCRIPTS+= scripts/external.test scripts/fwd.test scripts/fwd.test.expect

		int l = WSNPRINTF(portStr, sizeof(portStr), "%d\n", (int)port);
		WFWRITE(NULL, portStr, MIN((size_t)l, sizeof(portStr)), 1, f);

-            int l = WSNPRINTF(portStr, sizeof(portStr), "%d\n", (int)port);
-            WFWRITE(NULL, portStr, MIN((size_t)l, sizeof(portStr)), 1, f);
+            int l = WSNPRINTF(portStr, sizeof(portStr), "%d\n", (int)fwdFromPort);
+            if (l > 0) {
+                size_t writeLen = (size_t)l;
+                if (writeLen >= sizeof(portStr)) {
+                    writeLen = sizeof(portStr) - 1;
+                }
+                WFWRITE(NULL, portStr, writeLen, 1, f);
+            }

	ES_ERROR("Use -N when testing forced non blocking\n");
	ES_ERROR("Use -N when testing forced non-blocking\n");

Conversation

ejohnstown commented Feb 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

JacobBarthelmeh left a comment

Choose a reason for hiding this comment

Uh oh!

JacobBarthelmeh commented Feb 17, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

ejohnstown commented Feb 2, 2026 •

edited

Loading