Bug 1423822 moved the injected code section before the .text section. When linking with lld, the text section is usually page aligned, and starting a PT_LOAD. We inject code at the beginning of the PT_LOAD, which means the PT_LOAD is going to be extended at least a page downwards. And it means the preceding PT_LOAD can't finish in that same page, so the overhead of the injected code is needs to account for the page alignment.

Comment on attachment 8998469 [details] [diff] [review] Fix an elfhack corner case after bug 1423822 Review of attachment 8998469 [details] [diff] [review]: ----------------------------------------------------------------- ::: build/unix/elfhack/elfhack.cpp @@ +1085,2 @@ > size_t align = first_executable->getSegmentByType(PT_LOAD)->getAlign(); > + size_t new_size = relhack->getSize() + relhackcode->getSize() + relhackcode->getAddr() & (align - 1); This is a slightly odd expression for dealing with page alignment, but I'm not sure how to make it more clearer, short of expanding the comment quite a bit.

Pushed by mh@glandium.org: https://hg.mozilla.org/integration/mozilla-inbound/rev/068870abe0cc Fix an elfhack corner case after bug 1423822. r=froydnj

Pushed by mh@glandium.org: https://hg.mozilla.org/integration/mozilla-inbound/rev/b353e89db751 Fix up for bug 1481727, because operator priorities are fucked up in C/C++. r=bustage on a CLOSED TREE

(In reply to Pulsebot from comment #4) > Pushed by mh@glandium.org: > https://hg.mozilla.org/integration/mozilla-inbound/rev/b353e89db751 > Fix up for bug 1481727, because operator priorities are fucked up in C/C++. > r=bustage on a CLOSED TREE Because who would expect a + b & c to mean (a + b) & c?

https://hg.mozilla.org/mozilla-central/rev/068870abe0cc https://hg.mozilla.org/mozilla-central/rev/b353e89db751