<div dir="ltr"><div class="gmail_extra"><br><div class="gmail_quote">On Fri, May 30, 2014 at 3:44 PM, Adrian Prantl <span dir="ltr"><<a href="mailto:aprantl@apple.com" target="_blank">aprantl@apple.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5"><br>
> On May 30, 2014, at 3:40 PM, David Blaikie <<a href="mailto:dblaikie@gmail.com">dblaikie@gmail.com</a>> wrote:<br>
><br>
> On Fri, May 30, 2014 at 3:37 PM, Adrian Prantl <<a href="mailto:aprantl@apple.com">aprantl@apple.com</a>> wrote:<br>
>><br>
>>> On May 30, 2014, at 3:31 PM, David Blaikie <<a href="mailto:dblaikie@gmail.com">dblaikie@gmail.com</a>> wrote:<br>
>>><br>
>>> On Fri, May 30, 2014 at 3:12 PM, Eric Christopher <<a href="mailto:echristo@gmail.com">echristo@gmail.com</a>> wrote:<br>
>>>> On Fri, May 30, 2014 at 3:03 PM, David Blaikie <<a href="mailto:dblaikie@gmail.com">dblaikie@gmail.com</a>> wrote:<br>
>>>>> On Fri, May 30, 2014 at 3:00 PM, Alexey Samsonov <<a href="mailto:vonosmas@gmail.com">vonosmas@gmail.com</a>> wrote:<br>
>>>>>><br>
>>>>>> On Fri, May 30, 2014 at 12:45 PM, Eric Christopher <<a href="mailto:echristo@gmail.com">echristo@gmail.com</a>><br>
>>>>>> wrote:<br>
>>>>>>><br>
>>>>>>> On Fri, May 30, 2014 at 10:26 AM, Adrian Prantl <<a href="mailto:aprantl@apple.com">aprantl@apple.com</a>> wrote:<br>
>>>>>>>><br>
>>>>>>>>> On May 29, 2014, at 10:03 PM, Alexey Samsonov <<a href="mailto:vonosmas@gmail.com">vonosmas@gmail.com</a>><br>
>>>>>>>>> wrote:<br>
>>>>>>>>><br>
>>>>>>>>> Yeah, well, we can add "FrameTeardown" machine instruction flag in<br>
>>>>>>>>> addition to "FrameSetup", and annotate machine instructions added in frame<br>
>>>>>>>>> lowering code. But do we expect many users of this flag (assuming that this<br>
>>>>>>>>> usage is also somewhat questionable)?<br>
>>>>>>>><br>
>>>>>>>> The FrameSetup flag is also only used by the debug info, so at least it<br>
>>>>>>>> wouldn’t be any more questionable than that :-)<br>
>>>>>><br>
>>>>>><br>
>>>>>> OK. Why don't we land the patch as-is now (don't terminate the register at<br>
>>>>>> the end of MBB if this register is only modified in frame setup and the last<br>
>>>>>> MBB - it should be safe), and then replace "the last MBB" condition with<br>
>>>>>> checking a "FrameTeardown" flag, attached to certain machine instructions?<br>
>>>>>> In this way we won't have to special-case certain registers and make<br>
>>>>>> assumptions about their constant-ness.<br>
>>>>><br>
>>>>> Given the limited nature of this fix (both in the fix's possible scope<br>
>>>>> and intent/purpose) I'm personally sort of inclined to do this by<br>
>>>>> whitelisting certain registers instead, if that were/is possible...<br>
>>>>> seems simpler and still sufficient for the required purpose here.<br>
>>>>><br>
>>>>> Can we easily just test whether a register is the stack/frame pointer?<br>
>>>>> (I actually know so little about this that the difference between<br>
>>>>> those two things isn't straight in my head and I'm not sure which, or<br>
>>>>> both, we need for the common/intended case here)<br>
>>>>><br>
>>>><br>
>>>> If we need to stack realign we might not be able to do this as easily.<br>
>>>> I'm not a huge fan of either method, but I think that the assumption<br>
>>>> that Alexey has might make more sense than whitelisting.<br>
>>><br>
>>> Fair enough.<br>
>>><br>
>>>> My logic:<br>
>>>> Whitelisting involves assuming that various registers won't be used in<br>
>>>> ways that surprise us and I can think of a few ways where it could<br>
>>>> happen: stack realignment, inline asm, a smarter stack coloring<br>
>>>> algorithm.<br>
>>><br>
>>> Can inline asm clobber the frame pointer & we actually tolerate that,<br>
>>> find the right variables, etc, when that happens? (how?)<br>
>>><br>
>>> Can't quite picture the stack coloring you have in mind & Alexey<br>
>>> commented on the stack realignment issue.<br>
>>><br>
>>> The only other idea I'd float would be another simpler solution,<br>
>>> though a bit more expressive: just a check to see that the particular<br>
>>> register(s) (stack/frame pointers, whichever things it is we actually<br>
>>> use to describe the location of stack variables) don't change beyond<br>
>>> the prologue, and if not, describe any variables that are relative to<br>
>>> that register as having that location for the whole program.<br>
>><br>
>> That won’t work for values that are spilled to the stack, because they only live on the stack for part of the function.<br>
><br>
> Do we get those right currently?<br>
<br>
</div></div>Partially. We get them right until the end of the current basic block.<br></blockquote><div><br></div><div>The patch should improve this situation by the way:</div><div><bbX>:</div><div> DBG_VALUE $rsp, 32 !"var" // "var is spilled on stack.</div>
<div>// ... more code ..</div><div><bbY>:</div><div> DBG_VALUE $r9, 0, !"var" // "var" is now loaded to "r9".</div><div><br></div><div>Earlier we used to terminate the location of "var" at the end of bbX, but now we won't do</div>
<div>this, as we know "rsp" doesn't change in the function.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
But there is no code that would insert DBG_VALUES after they are reloaded again.</blockquote><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class=""><br>
><br>
>> [And for (-O0) allocas that are described in the MMI table, we already do what you described].<br>
><br>
> Right - but the problem Alexey is trying to address is what ASan does,<br>
> where it makes one big alloca and describes variables as being within<br>
</div>> that alloca - so the MMI sidetable handling doesn't fire…<br>
<br>
Awright.<br></blockquote><div><br></div><div>Yeah, MMI side table doesn't really work for indirect descriptions, when we need to deref the memory stored at reg+offset.</div><div><br></div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="HOEnZb"><div class="h5">><br>
>><br>
>> -- adrian<br>
>><br>
>>><br>
>>> But, yeah, that's not /lots/ easier than finding any registers that<br>
>>> happen to meet that property.<br>
>>><br>
>>> - David<br>
>>><br>
>>>><br>
>>>> -eric<br>
>>>><br>
>>>>>><br>
>>>>>>><br>
>>>>>>>><br>
>>>>>>><br>
>>>>>>> :)<br>
>>>>>>><br>
>>>>>>>>> I can also make this patch even less general and just special-case the<br>
>>>>>>>>> frame pointer and the stack pointer registers, so that the code wouldn't<br>
>>>>>>>>> pretend to solve a general problem we're facing.<br>
>>>>>>>><br>
>>>>>>>> The frame pointer should be stable over the entire function, special<br>
>>>>>>>> casing it seems to be appropriate. I don’t know whether, e.g, stack coloring<br>
>>>>>>>> would cause the stack pointer to be modified in the middle of a function.<br>
>>>>>>>><br>
>>>>>>><br>
>>>>>>> FWIW it may not _now_ but there's nothing from stopping it either. :\<br>
>>>>>>><br>
>>>>>>> -eric<br>
>>>>>><br>
>>>>>><br>
>>>>>> --<br>
>>>>>> Alexey Samsonov<br>
>>>>>> <a href="mailto:vonosmas@gmail.com">vonosmas@gmail.com</a><br>
<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div dir="ltr">Alexey Samsonov<br><a href="mailto:vonosmas@gmail.com" target="_blank">vonosmas@gmail.com</a></div>
</div></div>