Closed Bug 538393 Opened 13 years ago Closed 13 years ago
Assembler allocates two code chunks for each assembly pass even if the code has no side exits
We eagerly allocate code for side exits before we know we need it. In tamarin-redux this results in one extra code chunk being allocated and returned to CodeAlloc each time we jit, and we'd like to eliminate that overhead. It turns out a simple fix is to lazy-allocate memory for side exits.
Assignee: nobody → edwsmith
Attachment #420561 - Flags: review?(graydon)
Seems reasonable, but can you add a comment to Assembler::_nExitIns in Assembler.h mentioning it's not valid if there are no side exits because the underlying memory isn't allocated? Thanks.
yes, and i'll double make sure it's nulled out in that case.
Attachment #420561 - Flags: review?(graydon) → review?(nnethercote)
Attachment #421106 - Flags: review?(nnethercote) → review+
pushed to nanojit-central http://hg.mozilla.org/projects/nanojit-central/rev/97bacc5dff62
Whiteboard: fixed-in-nanojit → fixed-in-nanojit fixed-in-tamarin
Whiteboard: fixed-in-nanojit fixed-in-tamarin → fixed-in-nanojit, fixed-in-tamarin, fixed-in-tracemonkey
Status: NEW → RESOLVED
Closed: 13 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.