<div dir="ltr">It's 'known' (by some number of LLVM developers) that opt -O3 isn't the same as clang -O3. It'd be nice if they were closer - patches welcome, etc, but it hasn't been a priority for anyone. opt -O3 is rarely used - usually opt is used for testing specific optimizations.<br><br>Clang's IR output will differ between -O0 and -O3 (even before running any LLVM optimizations) - things like lifetime intrinsics, etc, are emitted only with optimizations enabled, for instance.<br><br>If you want to reproduce clang's -O3, best to use clang -O3 (with source code, or with LLVM IR generated from clang -O3 (so it has lifetime intrinsics, etc))</div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, Oct 24, 2019 at 9:22 PM Neil Nelson via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org">llvm-dev@lists.llvm.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div bgcolor="#FFFFFF">
<p><font size="-1">Yes, this is another indication that there some
processing or bridge in the clang -O3 compile not so far
evidenced as well when compiling with clang to its IR before the
optimization passes.</font></p>
<p><font size="-1">This may be an issue explained in a yet to be
known documentation page. Or it may be a point at the moment
overlooked by the well informed.</font></p>
<p><font size="-1">An issue being noted here but not well addressed
is that a well stated design of LLVM with its front-ends and
back ends is that the front-ends compile to an IR without
optimization that LLVM uses for optimization and preparation for
various back-ends. But that with clang -O3, given this evidence,
we are not easily seeing how the division between the clang
front end and </font><font size="-1"><font size="-1">LLVM</font>
works, though the assumed design suggests it should be quite
easy.</font></p>
<p><font size="-1">We should be able to compile with clang to the IR
before optimization and then apply the LLVM optimization
separately to obtain the same final IR as a clang -O3 compile
doing both of those. But we are not seeing that.</font></p>
<p><font size="-1">This also bears on the e2e thread in that this
assumed division posits that the separate clang and LLVM debug
sequences can provide a high reliability since the IR
intermediate between the two is not expected to be that error
prone. The errors are expected to be primarily either in clang
in obtaining a correct IR or in opt (LLVM) in optimizing that IR
for the back-end. But since we are not able to identify the IR
between the two under clang -O3 it is a question as to what
debug sequence would handle what we could not identify.</font></p>
<p><font size="-1">Neil Nelson<br>
</font></p>
<div><font size="-1">On 10/24/19 5:04 AM,
hameeza ahmed wrote:</font><br>
</div>
<blockquote type="cite">
<div dir="ltr">
<div><font size="-1">I run matrix multiplication code with both
the approaches o3 at clang and o3 at opt. clang o3 is about
2.97x faster than opt o3.</font></div>
<div><font size="-1"><br>
</font></div>
<div><font size="-1"><br>
</font></div>
</div>
<font size="-1"><br>
</font>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr"><font size="-1">On Mon, Oct
21, 2019 at 8:24 AM Neil Nelson <<a href="mailto:nnelson@infowest.com" target="_blank">nnelson@infowest.com</a>>
wrote:</font><br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div bgcolor="#FFFFFF">
<pre><code>is_sorted.cpp
bool</code> <code>is_sorted(</code><code>int</code> <code>*a, </code><code>int</code> <code>n) {</code><code>
</code><code>for</code> <code>(</code><code>int</code> <code>i = 0; i < n - 1; i++)</code></pre>
<div>
<div><code> </code><code>if</code> <code>(a[i] >
a[i + 1])</code></div>
<div><code> </code><code>return</code> <code>false</code><code>;</code></div>
<div><code> </code><code>return</code> <code>true</code><code>;</code></div>
<div><code>}</code><br>
</div>
</div>
<pre><a href="https://blog.regehr.org/archives/1605" target="_blank">https://blog.regehr.org/archives/1605</a> How Clang Compiles a Function
<a href="https://blog.regehr.org/archives/1603" target="_blank">https://blog.regehr.org/archives/1603</a> How LLVM Optimizes a Function
clang version 10.0.0, Xubuntu 19.04
clang is_sorted.cpp -S -emit-llvm -o is_sorted_.ll
clang is_sorted.cpp -O0 -S -emit-llvm -o is_sorted_O0.ll
clang is_sorted.cpp -O0 -Xclang -disable-llvm-passes -S -emit-llvm -o is_sorted_disable.ll
No difference in the prior three ll files.
clang is_sorted.cpp -O1 -S -emit-llvm -o is_sorted_O1.ll
Many differences between is_sorted_O1.ll and is_sorted_.ll.
opt -O3 -S is_sorted_.ll -o is_sorted_optO3.ll
clang is_sorted.cpp -mllvm -debug-pass=Arguments -O3 -S -emit-llvm -o is_sorted_O3arg.ll
opt <optimization sequence obtained in prior step> -S is_sorted_.ll -o is_sorted_opt_parms.ll
No difference between is_sorted_optO3.ll and is_sorted_opt_parms.ll, the last two opt runs.
Many differences between is_sorted_O3arg.ll and is_sorted_opt_parms.ll, the last two runs,
clang and opt.
Conclusions:
Given my current understanding, the ll files from the first three clang runs
are before any optimizations. Those ll files are from the front-end phase (CFE).
But this is a simple program and it may be that for a more complex program that
the ll files could be different.
Whether or not we use a -O3 optimization or use the parameters provided by clang for a
-03 optimization, we obtain the same result.
The difference in question is why an opt run using the CFE ll before optimization
obtains a different ll than a CFE run that includes optimization. That is, for this case,
it is not the expansion of the -O3 parameters that is the difference.
Initially, it would be interesting to have an ll listing before optimization from the
clang run that includes optimization to compare with the ll from the clang run without
optimization.
Neil Nelson
On 10/19/19 11:48 AM, Mehdi AMINI via llvm-dev wrote:
</pre>
<blockquote type="cite">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr"><br>
</div>
<br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Thu, Oct 17,
2019 at 11:22 AM David Greene via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">hameeza ahmed
via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>>
writes:<br>
<br>
> Hello,<br>
> I want to study the individual O3
optimizations. For this I am using<br>
> following commands, but unable to replicate
O3 behavior.<br>
><br>
> 1.
Documents/clang+llvm-9.0.0-x86_64-linux-gnu-ubuntu-18.04/bin/clang
-O1<br>
> -Xclang -disable-llvm-passes -emit-llvm -S
vecsum.c -o vecsum-noopt.ll<br>
><br>
> 2.
Documents/clang+llvm-9.0.0-x86_64-linux-gnu-ubuntu-18.04/bin/clang
-O3<br>
> -mllvm -debug-pass=Arguments -emit-llvm -S
vecsum.c<br>
><br>
> 3.
Documents/clang+llvm-9.0.0-x86_64-linux-gnu-ubuntu-18.04/bin/opt<br>
> <optimization sequence obtained in step
2> -S vecsum-noopt.ll -S -o<br>
> o3-chk.ll<br>
><br>
> Why the IR obtained by above step i.e
individual O3 sequences, is not same<br>
> when O3 is passed?<br>
><br>
> Where I am doing mistake?<br>
</blockquote>
<div><br>
</div>
<div>If you could provide the full reproducer, it
could help to debug this.</div>
<div> <br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
I think you need to turn off LLVM optimizations
when doing the<br>
-emit-llvm dump. Something like this:<br>
<br>
Documents/clang+llvm-9.0.0-x86_64-linux-gnu-ubuntu-18.04/bin/clang -O3 \<br>
-mllvm -debug-pass=Arguments -Xclang
-disable-llvm-optzns -emit-llvm \<br>
-S vecsum.c<br>
<br>
Otherwise you are effectively running the O3
pipeline twice, as clang<br>
will emit LLVM IR after optimization, not before
(this confused me too<br>
when I first tried it).<br>
</blockquote>
<div><br>
</div>
<div>This is the common pitfall indeed!</div>
<div>I think they are doing it correctly in step 1
though by including: `-Xclang
-disable-llvm-passes`.</div>
<div><br>
</div>
<div><br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">That said, I'm
not sure you will get the same IR out of opt as
with<br>
clang -O3 even with the above. For example, clang
sets<br>
TargetTransformInfo for the pass pipeline and the
detailed information<br>
it uses may or may not be transmitted via the IR
it dumps out. I have<br>
not personally tried to do this kind of thing in a
while.</blockquote>
<div><br>
</div>
<div>I struggled as well to setup TTI and TLI the
same way clang does :(</div>
<div>It'd be nice to revisit our PassManagerBuilder
setup and the opt integration to provide
reproducibility (maybe could be a starter project
for someone?).</div>
<div><br>
</div>
<div>-- </div>
<div>Mehdi</div>
<div><br>
</div>
</div>
</div>
</div>
<br>
<fieldset></fieldset>
<pre>_______________________________________________
LLVM Developers mailing list
<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>
<a href="https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev" target="_blank">https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a>
</pre>
</blockquote>
</div>
</blockquote>
</div>
</blockquote>
</div>
_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a><br>
<a href="https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev" rel="noreferrer" target="_blank">https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a><br>
</blockquote></div>