<div dir="ltr">+dvyukov</div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Aug 17, 2015 at 8:37 AM, Renato Golin <span dir="ltr"><<a href="mailto:renato.golin@linaro.org" target="_blank">renato.golin@linaro.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Folks,<br>
<br>
The review of patch <a href="http://reviews.llvm.org/D11532" rel="noreferrer" target="_blank">http://reviews.llvm.org/D11532</a> is extremely slow<br>
due to the number of hacks, left-overs and general undesired changes<br>
and style that the submission has. That happens, and it's ok when the<br>
overall direction the patch is going was agreed, and is acceptable as<br>
generally good.<br>
<br>
But this is not the case.<br>
<br>
To wake up the elephant in the room, do we really think that adding<br>
such a hack to an LLVM project for the sake of saving Android the<br>
burden of carrying such hack is *acceptable*?<br>
<br>
The only "reason" given to add such a hack was that, and I quote:<br>
* "what are we to do for NOW when the "proper" fix is maybe months<br>
off, or possibly longer?"<br>
* "This patch will allow people to experiment with TSAN on their<br>
android devices"<br>
* "don't let the perfect be the enemy of "limping along for a bit"<br>
<br>
So, in order to let *some* people *experiment* with TSAN on *Android*,<br>
we're going to consciously make TSAN *limp along* for the foreseeable<br>
future? Is that a wise price to pay for such a far fetched goal?<br>
<br>
The "proper" solution seems to be to fix TLS support, which:<br>
* "Ideally, this should be supported with the __thread keyword, but<br>
that is not yet supported, and it is a much bigger chunk of work to<br>
get it integrated into the default android cross compiler."<br>
<br>
So, to avoid a larger amount of work on another compiler, we're going<br>
to cripple LLVM?<br>
<br>
I cannot see why this would *ever* be a wise decision... Please,<br>
someone enlighten me.<br>
<br>
cheers,<br>
--renato<br>
</blockquote></div><br></div>