Reduce ASCII checks in makeInlineImage(). #5187
Conversation
| state = 0; | ||
| break; // some binary stuff found, resetting the state |
There was a problem hiding this comment.
I wonder about this break (and one below) removal, does it improve the performance?
There was a problem hiding this comment.
with the break removed it would always scan the 5 chars, wouldnt it?
There was a problem hiding this comment.
.. but it might open road for some jit optimizations, that's why I'm wondering. Will it be beneficial to add a constant 5 instead of "followingBytes.length" to the for-loop then?
There was a problem hiding this comment.
I'll add the break back in. I don't know what the JITs will do, but I can put the 5 in its own variable.
makeInlineImage() has a "are the next five chars ASCII?" check which is run after an "EI" sequence has been found. This check involves the creation of a new object because peekBytes() calls subarray(). Unfortunately, the check is currently run on whitespace chars even when an "EI" sequence has not yet been found, i.e. when it's not needed. For the PDF in mozilla#2618, there are over 820,000 such checks. This change reworks the relevant loop so that the check is only done once an "EI" sequence has been seen. This reduces the number of checks to 157,000, and speeds up rendering by somewhere between 2% and 7% (the measurements are noisy).
|
/botio test |
From: Bot.io (Windows)ReceivedCommand cmd_test from @Snuffleupagus received. Current queue size: 0 Live output at: http://107.22.172.223:8877/cdb4df9f9e7f675/output.txt |
From: Bot.io (Linux)ReceivedCommand cmd_test from @Snuffleupagus received. Current queue size: 0 Live output at: http://107.21.233.14:8877/b53e64276c6bc14/output.txt |
From: Bot.io (Windows)SuccessFull output at http://107.22.172.223:8877/cdb4df9f9e7f675/output.txt Total script time: 19.76 mins
|
From: Bot.io (Linux)SuccessFull output at http://107.21.233.14:8877/b53e64276c6bc14/output.txt Total script time: 22.52 mins
|
|
looks good to me. |
Reduce ASCII checks in makeInlineImage().
|
Thank you for the patch |
makeInlineImage() has a "are the next five chars ASCII?" check which is
run after an "EI" sequence has been found. This check involves the
creation of a new object because peekBytes() calls subarray().
Unfortunately, the check is currently run on whitespace chars even when
an "EI" sequence has not yet been found, i.e. when it's not needed. For
the PDF in #2618, there are over 820,000 such checks.
This change reworks the relevant loop so that the check is only done
once an "EI" sequence has been seen. This reduces the number of checks
to 157,000, and speeds up rendering by somewhere between 2% and 7% (the
measurements are noisy).