STR: 1. Open attached "testcase 1" 2. Click in the second line, press End to move cursor to the end of the line [you will get "asdf\nghjk[caret]\nzxcv" in textarea] 3. Press Shift+Right to select "\n" 4. Press Right key to move cursor to the end of text selection 5. Type "x" Result: After Step 4 cursor is displayed at the end of the 2nd line After Step 5 the typed letter ("x") appears in the beginning of the 3rd line Expectations: After Step 4 cursor should be displayed at the start of the 3rd line

Ah right, the fix in bug 1237236 is using GetPrimaryFrame so it only really works on the first line. We need to lookup the actual continuation that renders aParentNode/aOffset...

(In reply to Mats Palmgren (:mats) from comment #3) > Ah right, the fix in bug 1237236 is using GetPrimaryFrame so it only really > works on the first line. We need to lookup the actual continuation that > renders aParentNode/aOffset... Can you rename this bug with the correct diagnosis? We've now got multiple bugs with similar symptoms in the name. Thanks!

This is just a variation that my fix in bug 1237236 didn't catch.

I guess we could skip the first do_QueryFrame now, but nsCaret::GetFrameAndOffset looks relatively expensive so it's probably worth avoid calling it unless we know we have a text frame.

Comment on attachment 8733409 [details] [diff] [review] tests Review of attachment 8733409 [details] [diff] [review]: ----------------------------------------------------------------- ::: layout/base/tests/test_reftests_with_caret.html @@ +170,5 @@ > tests.push([ 'bug966992-1.html' , 'bug966992-1-ref.html' ]); > tests.push([ 'bug966992-2.html' , 'bug966992-2-ref.html' ]); > tests.push([ 'bug966992-3.html' , 'bug966992-3-ref.html' ]); > tests.push(function() {SpecialPowers.pushPrefEnv({'clear': [['layout.css.overflow-clip-box.enabled']]}, nextTest);}); > + tests.push([ 'bug1258308-1.html' , 'bug1258308-1-ref.html' ]); // it seems VK_END doesn't work on Android Hmm. bug664087-*.html seem to use this just fine. But this is fine if you want to land it as is.

Odd... the failures looked identical on OSX and Android -- after I changed VK_END to Accel+VK_RIGHT for OSX the test worked fine so I assumed it was the same problem on Android, only I couldn't figure out what key combo to use there... Anyway, I'll land it without Android testing for now. (and I added a "maybe?" to the test comment quoted above)

https://hg.mozilla.org/integration/mozilla-inbound/rev/82c36cd54626 https://hg.mozilla.org/integration/mozilla-inbound/rev/99a181de0c57

https://hg.mozilla.org/mozilla-central/rev/82c36cd54626 https://hg.mozilla.org/mozilla-central/rev/99a181de0c57

STR in comment 0 is Fixed on: Win7_64, Nightly 48, 32bit, ID 20160324030447 But I still can reproduce the same issue if I select the last \n if it's the last symbol in textarea. I.e. if I press Shift+Right then Right in textarea "A\nB[caret]\n" I already filed that case as bug 1237286 (Or should I file a separate bug not involving Shift+Enter?)

It may be the same underlying bug as bug 1237286, but please file a separate bug just in case, with a testcase and STR. Mark it as blocking 1237286 to make a link between the two. Thanks for verifying these fixes on Windows.