Thank you for the review and your comments @sammccall and @mstorsjo. As I understand there is no strong opinion here either way. I think we can leave things as they are for now. I am going to abandon this patch. If someone else feels strongly about removing them, they can put up a patch for this.

In D155111#4495131, @mstorsjo wrote:

To clarify the issue - the kind of builds that seems to be broken is builds with BUILD_SHARED_LIBS=ON. The reason is that these libraries are needed is because the clangd target includes $<TARGET_OBJECTS:obj.clangDaemonTweaks>, so all the dependencies of clangDaemonTweaks would need to be included here as well. Please include that in the commit message description. (Is there a way to pull in those instead of duplicating the list?)

This looks mostly ok to me, but it does add slightly more libraries than what's needed. As the list of libraries that now are linked into clangdMain is the list of libraries that previously was linked for the two components that now are clangd and clangdMain, so some of the dependencies only need to be moved, not duplicated.

A more minimal set of dependencies, which seems to link successfully with BUILD_SHARED_LIBS=ON, is achieved with this diff on top of current git main:

diff --git a/clang-tools-extra/clangd/tool/CMakeLists.txt b/clang-tools-extra/clangd/tool/CMakeLists.txt
index ddf9c2488819..6c21175d7687 100644
--- a/clang-tools-extra/clangd/tool/CMakeLists.txt
+++ b/clang-tools-extra/clangd/tool/CMakeLists.txt
@@ -26,11 +26,7 @@ clang_target_link_libraries(clangdMain
   clangBasic
   clangFormat
   clangFrontend
-  clangLex
-  clangSema
   clangTooling
-  clangToolingCore
-  clangToolingRefactoring
   clangToolingSyntax
   )
 
@@ -44,7 +40,20 @@ target_link_libraries(clangdMain
   ${CLANGD_XPC_LIBS}
   )
 
+clang_target_link_libraries(clangd
+  PRIVATE
+  clangAST
+  clangBasic
+  clangLex
+  clangSema
+  clangToolingCore
+  clangToolingRefactoring
+  clangToolingSyntax
+  )
+
 target_link_libraries(clangd
   PRIVATE
   clangdMain
+  clangDaemon
+  clangdSupport
   )

Not sure if it's good hygiene to only link specifically to exactly those libraries that are needed and nothing else, or if it's just making things slightly more brittle?

@mstorsjo @ivanmurashko Can you please take a look at this? This is blocking our builds on the PowerPC buildbots. Thanks a lot for your time!

This is to fix failures in the build bot: https://lab.llvm.org/buildbot/#/builders/57/builds/28356

In D140560#4032567, @arsenm wrote:

In D140560#4031989, @arsenm wrote:

In D140560#4030857, @saghir wrote:

Hi, this seems to be causing our bots to fail. One example is https://lab.llvm.org/buildbot/#/builders/231/builds/6915. Can you please take a look? Thanks!

I have a patch ready for this, need to write some more tests for it

Try 40078a6b713730ffc164d4c0733d26352eb1e236

Hi, this seems to be causing our bots to fail. One example is https://lab.llvm.org/buildbot/#/builders/231/builds/6915. Can you please take a look? Thanks!

Hi @Orlando, the test case (remove-undef-fragment.ll) that you added as part of 1d1de7467c32d52926ca56b9167a2c65c451ecfa is breaking one of our bots: https://lab.llvm.org/staging/#/builders/14/builds/1332. Can you please take a look? Thanks!

In D139430#3984689, @1c3t3a wrote:

Oh. I think I know where this comes from - for fixing this: Would I submit to this changelist or open a new one?

Just a heads up that this seems to be breaking one of our bots: https://lab.llvm.org/staging/#/builders/14/builds/1305.

LGTM.

Besides other comments, LGTM. Thanks!

LGTM. Thanks!

update test case output

updated test case.

Address review comments

In D126807#3581621, @lkail wrote:

This fix looks making offset handling more complex. We can make it easier by add LQX_PSEUDO and STQX_PSEUDO to ImmToIdxMap.

ImmToIdxMap[PPC::LQ]  = PPC::LQX_PSEUDO;
ImmToIdxMap[PPC::STQ] = PPC::STQX_PSEUDO;

And expand PPC::LQX_PSEUDO and PPC::STQX_PSEUDO post RA(They are expanded in PPCTargetLowering::EmitInstrWithCustomInserter right now). Something like

$x6 = LQX_PSEUDO $x0, $x1
=>
$x3 = ADD $x0, $x1
$x6 = LQ 0($x3)

Thus we make LQ/STQ fits in how we are handling frame index now, no more code is needed to handling the offset for LQ/STQ(When LQ/STQ are selected by ISEL, alignment is guaranteed to be 16 bytes since they are for atomic operations).

Added LQX_PSEUDO and STQX_PSEUDO to ImmToIdxMap, and rebased.

I think there is a minor comment from Zarko for removing a run line, but looks good other than that.

Handle unaligned offsets and offsets that do not fit in the instruction.

LGTM. Thanks!

Use string substitution blocks for matching register and offset.

In D110552#3528471, @hctim wrote:

Looks like I missed something, probably because 'ppc' isn't a feature recognised by compiler-rt's lit config. Looking.

In D110552#3528471, @hctim wrote:

Looks like I missed something, probably because 'ppc' isn't a feature recognised by compiler-rt's lit config. Looking.

Hi @alexey-katranov, the test case added as part of this commit seems to be failing on our Power PC bots:

LGTM. Thanks for doing this!

LGTM.

LGTM.

LGTM.

Abandoning this since Galina suggested to do the changes locally for the buildbot.

Updates version of gcc build compiler.

Hi,

This change seems to be causing issues with Power PC buildbots. One such bot is https://lab.llvm.org/buildbot/#/builders/52/builds/10906. Can you please take a look to fix/revert the change? If there is anything I could do to help investigate, please let me know.

Thanks!

Stripped RV qualifiers for Sema checking.

Hi,
One of the tests added in this change is failing on PowerPC buildbots. One such bot is https://lab.llvm.org/buildbot/#/builders/121/builds/11903. Please revert and let us know how we can help you figure out what needs to be done to investigate/fix this on the PPC bots.
Thanks!

This was failing most of our buildbots, I reverted this change locally and the errors go away. I am reverting this change for now to get the buildbots back to green and give you a chance to look into it when you can.
Please feel free to reach out if you'd like any help with troubleshooting the issues that this patch was causing on PowerPC buildbots.

Addressed review comments.

Hi, this change seems to be breaking PowerPC buildbots:
One such bot is https://lab.llvm.org/buildbot/#/builders/57/builds/10548
Can you please take a look?

This seems to be breaking one of PPC buildbots: https://lab.llvm.org/buildbot/#/builders/18/builds/2553/steps/7/logs/stdio
Can you please take a look when you get a chance?

Changed approach for Sema checks.

In D109708#3000526, @saghir wrote:

Hi, this patch is likely causing a failure on one of our bots:
https://lab.llvm.org/buildbot/#/builders/76/builds/3841

Can you please take a look at your convenience?
Thanks!

Hi, this patch is likely causing a failure on one of our bots:
https://lab.llvm.org/buildbot/#/builders/76/builds/3841

update test case name to be consistent with existing ones.

In D109624#2998665, @thakis wrote:

In D109624#2998550, @saghir wrote:

Hi, this change also breaks Power PC bots:

1. https://lab.llvm.org/buildbot/#/builders/19/builds/6451/steps/24/logs/stdio
2. https://lab.llvm.org/buildbot/#/builders/18/builds/2443/steps/23/logs/stdio

It fails consistently on our bots. Can you please take a look?

Thanks!

This will hopefully be better after b7bac5a172e51ed065b3b4dc64cc2d8831e8081c.

Hi, this change also breaks Power PC bots:

1. https://lab.llvm.org/buildbot/#/builders/19/builds/6451/steps/24/logs/stdio
2. https://lab.llvm.org/buildbot/#/builders/18/builds/2443/steps/23/logs/stdio

Addressed review comments.

Addressed review comments.

Rebased before commit.

Rebased.

Addressed review comments.

Added comments, and re-organized tests.

Added some more tests.

Change the implementation to add new builtins and keep the depracated builtins.

Addressed review comments. Added another test case.