CQ shared: Fix off-by-nine error leading to lost messages #14576
Conversation
lhoguin
force-pushed
the
loic-fix-cq-eof-eof
branch
from
5ea57d8 to
9246583
Compare
|
cc @gomoripeti |
|
CQ test for file scanning is now failing. I'll look on Monday. |
lhoguin
force-pushed
the
loic-fix-cq-eof-eof
branch
from
9246583 to
8294c3f
Compare
|
I've changed it to +8, but will need to double check if that's right. |
|
For greater clarity I will change it to |
lhoguin
force-pushed
the
loic-fix-cq-eof-eof
branch
from
8294c3f to
52363c2
Compare
|
I am adding a test case that demonstrates the bug and the fix and can indeed confirm that it happens because of this off-by-nine (ie current code works if we are outside of this range). |
lhoguin
force-pushed
the
loic-fix-cq-eof-eof
branch
from
52363c2 to
1a7e92a
Compare
lhoguin
changed the title
And `eof` crashes. The problem is that we may end up trying to read more data from the file when scanning, despite being at the end of the file. This results in the current Acc to be returned instead of the remaining data being parsed. This results in some messages at the end of the file being truncated off despite still being in memory (and still pointing to the end of the original file, well past the truncation point).
lukebakken
force-pushed
the
loic-fix-cq-eof-eof
branch
from
1a7e92a to
f363854
Compare
|
@lukebakken Did you change something in the commit? |
|
No, sorry, I just clicked the "Update with rebase" button here. |
|
Update - I hit the |
|
I will merge this now. The tests have confirmed it fixes a bug. If we still hit |
CQ shared: Fix off-by-nine error leading to lost messages (backport #14576)
And
eofcrashes.The problem is that we may end up trying to read more data from the file when scanning, despite being at the end of the file. This results in the current Acc to be returned instead of the remaining data being parsed.
This results in some messages at the end of the file being truncated off despite still being in memory (and still pointing to the end of the original file, well past the truncation point).
Tentative fix for #14181 (reply in thread)
Someone please double check that there's no off-by-one error.