[subset] Raise the bar in new vs old table size #1930
Conversation
This is actually interesting thing that {h,v}mtx allocated as
much as a font pretends have glyphs but the solution is not
that obvious. This change is enough to consider this hmtx case.
Initially we wanted to just find things making crazy amount of
memory but having the assert has led to interesting findings also
so let's don't remove the assert and see what we can find elsewhere.
|
OR JUST REMOVE THE ARBITRARY LIMITATION!!! |
|
Done |
|
Can you explain why |
You are right that I should've explained it more and I am sorry. https://github.com/ebraminio/harfbuzz/blob/2023f0da44ba492740676b5fcd0fb19f4d883979/src/hb-ot-glyf-table.hh#L575 wasn't considering instruction size itself and wasn't failing when there was no space for that. However the logic here |
|
Yeah we should rewrite that bound checking in terms of |
That's why I went for rewriting most of it after painfully understanding of what is going on it. |
https://crbug.com/oss-fuzz/16740
This is actually an interesting thing that {h,v}mtx allocates as
much as a font pretends to have glyphs but the solution is not
that obvious as regular fonts can have less than actually containing
metrics in their {h,v}mtx. This change raises the bar to consider this
hmtx 4 byte for every glyph case.
Initially we wanted to just find things allocating crazy amount of
memory but having the assert has led to interesting findings also
so let's don't remove the assert and see what we can find elsewhere.