The checkin of the patch to bug 34297 caused Linux --disable-bidi build to fail with the following lines: nsFormSubmitter.cpp: In function `nsresult nsFormSubmitter::OnSubmit (nsIForm *, nsIPresContext *, nsIContent *)': nsFormSubmitter.cpp:212: `ctrlsModAtSubmit' undeclared (first use this function) nsFormSubmitter.cpp:212: (Each undeclared identifier is reported only once for each function it appears in.) nsFormSubmitter.cpp:212: `textDirAtSubmit' undeclared (first use this function) gmake[5]: *** [nsFormSubmitter.o] Error 1 gmake[5]: Leaving directory `/home/dark/DISK/mozilla/content/html/content/src' gmake[4]: *** [install] Error 2 Those lines are pasted from R.K.Aa's comment in 34297, he mentioned it also happens after clobbering. I got the same bustage from current trunk, after the checkin (I pulled at 2001-11-02, I guess that was 5am PST - if I can convert CET to PST correctly). Assingning to jkeiser as it was his patch.

CC twalker (smoketester) and reviewers of the patch too (jst, rods).

$ grep -n bidiOptions mozilla/content/html/content/src/* nsFormSubmitter.cpp:145: PRUint32 bidiOptions; nsFormSubmitter.cpp:146: aPresContext->GetBidi(&bidiOptions); nsFormSubmitter.cpp:147: PRUint8 ctrlsModAtSubmit = GET_BIDI_OPTION_CONTROLSTEXTMODE(bidiOptions); nsFormSubmitter.cpp:148: PRUint8 textDirAtSubmit = GET_BIDI_OPTION_DIRECTION(bidiOptions);

attachment 56250 [details] [diff] [review] is good: it builds and runs with --disable-bidi (Also tested on bidi pages, for the fun of it.) jkeiser had to go offline. He requested that i ask for an r, sr, and someone to check this in.

R.K.Aa, you probably shouldn't use today's builds, you just deleted the summary due to bug 108175

QA vanished too

Comment on attachment 56250 [details] [diff] [review] Updated Patch r=sicking

Comment on attachment 56250 [details] [diff] [review] Updated Patch sr=jst

Patch checked in.

someone please verify this, I have no idea what this is.

I've been building this since it was marked fixed, and it still is working OK. Verifying.