Propagate only WANT_WRITE in DoKexInit - #1278
Open
ejohnstown wants to merge 1 commit into
Open
ejohnstown wants to merge 1 commit into
ejohnstown wants to merge 1 commit into
Conversation
DoKexInit() reports ssh->error once it has handled the KEXINIT, so a blocked reply reaches the caller. Report it only when it is WS_WANT_WRITE. Anything else in ssh->error is left by an earlier call, and a WS_WINDOW_FULL from SendChannelData() failed the peer's rekey. - tests/regress.c runs DoKexInit() with WS_SUCCESS, WS_WINDOW_FULL and WS_CHAN_RXD left in ssh->error, and with the reply send blocked. Issue: F-14394
ejohnstown
requested review from
wolfSSL-Fenrir-bot
and
a balanced review from Copilot
September 29, 2026 00:24
wolfSSL-Fenrir-bot
left a comment
wolfSSL-Fenrir-bot
left a comment
There was a problem hiding this comment.
Fenrir Automated Review — PR #1278
Scan targets checked: wolfssh-src, wolfssh-bugs
Coverage: 1 of 2 in-scope changed file(s) opened by the reviewer; not opened: tests/regress.c
Fenrir result: Approved ✅
No new issues found in the changed files.
Advisory only — this automated result does not count as a GitHub approval.
Review tier: Lite
Contributor
There was a problem hiding this comment.
Warning
Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.
Copilot review overview
Review effort: Lite
Findings: 1
Open (5)
retunconditionally based onssh->error. Ifretalready contains a real… · Newssh->errorbeing specifically attributable toSendKexInit(). A… · NewKEXDH_REPLY_REGRESS_KEX_ALGO. If this macro isn’t enabled in… · NewBuildKexInitPayload()succeeded. If it returns 0 (or an otherwise… · NewWS_WANT_WRITE, it would be valuable to add a case where… · New
What changed in this PR
Updates DoKexInit() to only propagate WS_WANT_WRITE back to the caller (avoiding failures due to unrelated stale ssh->error values) and adds regression coverage around stale errors and blocked reply sends.
Changes:
- Limit
DoKexInit()error propagation toWS_WANT_WRITEonly. - Add a regression helper and test cases to validate stale
ssh->errorhandling and blocked send behavior. - Wire the new regression test into
main()under an existing compile-time guard.
| File | Description |
|---|---|
| tests/regress.c | Adds a regression harness/test to ensure stale ssh->error values don’t break peer rekey, while still reporting blocked reply sends. |
| src/internal.c | Changes DoKexInit() to propagate only WS_WANT_WRITE instead of any non-zero ssh->error. |
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
Comment on lines
+6938
to
+6943
| /* Propagate a want write from SendKexInit(). Only that: other | ||
| * codes in ssh->error are left by earlier calls, such as a | ||
| * WS_WINDOW_FULL from SendChannelData(), and would fail the | ||
| * key exchange. */ | ||
| if (ssh->error == WS_WANT_WRITE) | ||
| ret = WS_WANT_WRITE; |
Comment on lines
+6938
to
+6943
| /* Propagate a want write from SendKexInit(). Only that: other | ||
| * codes in ssh->error are left by earlier calls, such as a | ||
| * WS_WINDOW_FULL from SendChannelData(), and would fail the | ||
| * key exchange. */ | ||
| if (ssh->error == WS_WANT_WRITE) | ||
| ret = WS_WANT_WRITE; |
| @@ -14859,6 +14859,75 @@ static void TestKexInitReservedNonZeroRejected(void) | |||
| wolfSSH_CTX_free(ctx); | |||
| } | |||
|
|
|||
| AssertIntEQ(wolfSSH_SetAlgoListKey(ssh, FPF_KEY_GOOD), WS_SUCCESS); | ||
|
|
||
| payloadSz = BuildKexInitPayload(ssh, FPF_KEX_GOOD, FPF_KEY_GOOD, | ||
| 0, payload, (word32)sizeof(payload)); |
Comment on lines
+14920
to
+14928
| static void TestKexInitIgnoresStaleError(void) | ||
| { | ||
| AssertIntEQ(RunKexInitWithError(WS_SUCCESS, 0), WS_SUCCESS); | ||
| AssertIntEQ(RunKexInitWithError(WS_WINDOW_FULL, 0), WS_SUCCESS); | ||
| AssertIntEQ(RunKexInitWithError(WS_CHAN_RXD, 0), WS_SUCCESS); | ||
| /* A blocked reply is still reported. */ | ||
| AssertIntEQ(RunKexInitWithError(WS_SUCCESS, 1), WS_WANT_WRITE); | ||
| AssertIntEQ(RunKexInitWithError(WS_WINDOW_FULL, 1), WS_WANT_WRITE); | ||
| } |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
DoKexInit() now reports ssh->error to its caller only when it is
WS_WANT_WRITE. A stale WS_WINDOW_FULL left by SendChannelData() no
longer fails the peer's rekey.
WS_CHAN_RXD, and a blocked reply send.
Issue: F-14394