This patch adds support for parsing the proposed non-instruction debug
info ("RemoveDIs") from textual IR, and adds a test for the parser as well
as a set of verifier tests that are dependent on parsing to fire.
An important detail of this patch is the fact that although we can now
parse in the RemoveDIs (new) and Intrinsic (old) debug info formats, we
will always convert back to the old format at the end of parsing - this
is done for two reasons: firstly to ensure that every tool is able to
process IR printed in the new format, regardless of whether that tool
has had RemoveDIs support added, and secondly to maintain the effect of
the existing flags: for the tools where support for the new format has
been added, we will run LLVM passes in the new format iff
`--try-experimental-debuginfo-iterators=true`, and we will print in the
new format iff `--write-experimental-debuginfo-iterators=true`; the
format of the textual IR input should have no effect on either of these
features.
`DPLabel`, the RemoveDI version of `llvm.dbg.label`, was landed recently
at the same time the RemoveDIs IR printing and verifier patches were
landing. The patches were updated to not miscompile, but did not have
full-featured and correct support for DPLabel built in; this patch makes
the remaining set of changes to enable DPLabel support.
As part of the RemoveDIs project, this patch adds support for checking
DPValues in the verifier. Although this is not strictly parsing-related,
and we currently automatically convert back to the old debug info format
immediately after parsing, we are approaching the point where the we can
operate end-to-end in the new debug info format, at which point it is
appropriate that we can actually validate modules in the new format.
This patch also contains some changes that aren't strictly
parsing-related, but are necessary class refactors for parsing support,
and are used in the verifier checks (i.e. changing the DILocalVariable
field to be a tracking MD reference, and adding a Verifier check to
confirm that it is a DILocalVariable).
Patch 2 of 3 to add llvm.dbg.label support to the RemoveDIs project. The
patch stack adds the DPLabel class, which is the RemoveDIs
llvm.dbg.label
equivalent.
1. Add DbgRecord base class for DPValue and the not-yet-added
DPLabel class.
-> 2. Add the DPLabel class.
3. Enable dbg.label conversion and add support to passes.
This will be used (and tested) in the final patch(es), coming next.
Patch 1 of 3 to add llvm.dbg.label support to the RemoveDIs project. The
patch stack adds a new base class
-> 1. Add DbgRecord base class for DPValue and the not-yet-added
DPLabel class.
2. Add the DPLabel class.
3. Enable dbg.label conversion and add support to passes.
Patches 1 and 2 are NFC.
In the near future we also will rename DPValue to DbgVariableRecord and
DPLabel to DbgLabelRecord, at which point we'll overhaul the function
names too. The name DPLabel keeps things consistent for now.
Ensure that when a constant value dies, all ValueAsMetadata uses of that
constant in a DebugValueUser will be replaced with a PoisonValue instead
of a nullptr. This has little visible effect currently, as any nullptr metadata
uses in a DebugValueUser would be converted to an empty MDNode in
the end anyway, but this patch adds an assert that would be triggered by
two existing tests without the functional component of this patch:
llvm/test/Transforms/Inline/delete-function-with-metadata-use.ll
llvm/test/DebugInfo/X86/array2.ll
This is an optimisation patch that shouldn't have any functional effect.
There's no need for all instructions to have a DPMarker attached to them,
because not all instructions have adjacent DPValues (aka dbg.values).
This patch inserts the appropriate conditionals into functions like
BasicBlock::spliceDebugInfo to ensure we don't step on a null pointer when
there isn't a DPMarker allocated. Mostly, this is a case of calling
createMarker occasionally, which will create a marker on an instruction
if there isn't one there already.
Also folded into this is the use of adoptDbgValues, which is a natural
extension: if we have a sequence of instructions and debug records:
%foo = add i32 %0,...
# dbg_value { %foo, ...
# dbg_value { %bar, ...
%baz = add i32 %...
%qux = add i32 %...
and delete, for example, the %baz instruction, then the dbg_value records
would naturally be transferred onto the %qux instruction (they "fall down"
onto it). There's no point in creating and splicing DPMarkers in the case
shown when %qux doesn't have a DPMarker already, we can instead just change
the owner of %baz's DPMarker from %baz to %qux. This also avoids calling
setParent on every DPValue.
Update LoopRotationUtils: it was relying on each instruction having it's
own distinct end(), so that we could express ranges and lack-of-ranges.
That's no longer true though: so switch to storing the range of DPValues on
the next instruction when we want to consider it's range next time around
the loop (see the nearby comment).
In instcombine, when we sink an instruction into a successor block, we try
to clone and salvage all the variable assignments that use that Value. This
is a behaviour that's (IMO) flawed, but there are important use cases where
we want to avoid regressions, thus we're implementing this for the
non-instruction debug-info representation.
This patch refactors the dbg.value sinking code into it's own function, and
installs a parallel implementation for DPValues, the non-instruction
debug-info container. This is mostly identical to the dbg.value
implementation, except that we don't have an easy-to-access ordering
between DPValues, and have to jump through extra hoops to establish one in
the (rare) cases where that ordering is required.
The test added represents a common use-case in LLVM where these behaviours
are important: a loop has been completely optimised away, leaving several
dbg.values in a row referring to an instruction that's going to sink. The
dbg.values should sink in both dbg.value and RemoveDIs mode, and
additionally only the last assignment should sink.
This implements the DbgAssignIntrinsic class as a variant of DPValues -
unfortunately this involves increasing the size of the `DebugValueUser`
storage by 3x, but this is necessary to enable assigns to be
represented, and can be offset in a future patch by splitting DPValue
into subclasses such that each variant can store only the fields it
needs. This patch does not actually create DPVAssigns in any case;
future patches will handle this variant in all cases where generic
DPValue handling does not. This patch also does not implement tracking
support for DIAssignIDs, which is necessary to find DPVAssigns that
reference a given DIAssignID; that is added in a subsequent patch.
This patch adds a set of functions to the DPValue class that
conveniently perform some common operations, and some that replicate
existing functions on `DbgVariableIntrinsic` and its subclasses.
Here's a problem for the RemoveDIs project to make debug-info not be
stored in instructions -- in the following sequence:
dbg.value(foo
%bar = add i32 ...
dbg.value(baz
It's possible for rare passes (only CodeGenPrepare) to remove the add
instruction, and then re-insert it back in the same place. When
debug-info is stored in instructions and there's a total order on "when"
things happen this is easy, but by moving that information out of the
instruction stream we start having to do manual maintenance.
This patch adds some utilities for re-inserting an instruction into a
sequence of DPValue objects. Someday we hope to design this away, but
for now it's necessary to support all the things you can do with
dbg.values. The two unit tests show how DPValues get shuffled around
using the relevant function calls. A follow-up patch adds
instrumentation to CodeGenPrepare.
This is the "central" patch to the removing-debug-intrinsics project: it
changes the instruction movement APIs (insert, move, splice) to interpret
the "Head" bits we're attaching to BasicBlock::iterators, and updates
debug-info records in the background to preserve the ordering of debug-info
(which is in DPValue objects instead of dbg.values). The cost is the
complexity of this patch, plus memory. The benefit is that LLVM developers
can cease thinking about whether they're moving debug-info or not, because
it'll happen behind the scenes.
All that complexity appears in BasicBlock::spliceDebugInfo, see the diagram
there for how we now manually shuffle debug-info around. Each potential
splice configuration gets tested in the added unit tests.
The rest of this patch applies the same reasoning in a variety of
scenarios. When moveBefore (and it's siblings) are used to move
instructions around, the caller has to indicate whether they intend for
debug-info to move too (is it a "Preserving" call or not), and then the
"Head" bits used to determine where debug-info moves to. Similar reasoning
is needed for insertBefore.
Differential Revision: https://reviews.llvm.org/D154353
BasicBlock.h and Instruction.h will eventually need to include
DebugProgramInstruction.h so that debug-info attached to instructions can
be enumerated and cloned. Originally including it made compiling clang
much slower, I think I've pinned that down as being the inclusion of
DebugInfoMetadata.h causing ~every LLVM translation unit to parse
all the debug-info classes.
This patch avoids that by shifting some functions into the cpp file rather
than the header, and restores the inclusion of DebugProgramInstruction.h in
BasicBlock.h so that the rest of the RemoveDIs functionality can land.
This reverts commit 957efa4ce4f0391147cec62746e997226ee2b836.
Original commit message below -- in this follow up, I've shifted
un-necessary inclusions of DebugProgramInstruction.h into being forward
declarations (fixes clang-compile time I hope), and a memory leak in the
DebugInfoTest.cpp IR unittests.
I also tracked a compile-time regression in D154080, more explanation
there, but the result of which is hiding some of the changes behind the
EXPERIMENTAL_DEBUGINFO_ITERATORS compile-time flag. This is tested by the
"new-debug-iterators" buildbot.
[DebugInfo][RemoveDIs] Add prototype storage classes for "new" debug-info
This patch adds a variety of classes needed to record variable location
debug-info without using the existing intrinsic approach, see the rationale
at [0].
The two added files and corresponding unit tests are the majority of the
plumbing required for this, but at this point isn't accessible from the
rest of LLVM as we need to stage it into the repo gently. An overview is
that classes are added for recording variable information attached to Real
(TM) instructions, in the form of DPValues and DPMarker objects. The
metadata-uses of DPValues is plumbed into the metadata hierachy, and a
field added to class Instruction, which are all stimulated in the unit
tests. The next few patches in this series add utilities to convert to/from
this new debug-info format and add instruction/block utilities to have
debug-info automatically updated in the background when various operations
occur.
This patch was reviewed in Phab in D153990 and D154080, I've squashed them
together into this commit as there are dependencies between the two
patches, and there's little profit in landing them separately.
[0] https://discourse.llvm.org/t/rfc-instruction-api-changes-needed-to-eliminate-debug-intrinsics-from-ir/68939
And some intervening fixups. There are two remaining problems:
* A memory leak via https://lab.llvm.org/buildbot/#/builders/236/builds/7120/steps/10/logs/stdio
* A performance slowdown with -g where I'm not completely sure what the cause it
These might be fairly straightforwards to fix, but it's the end of the day
hear, so I figure I'll clear the buildbots til tomorrow.
This reverts commit 7d77bbef4ad9230f6f427649373fe46a668aa909.
This reverts commit 9026f35afe6ffdc5e55b6615efcbd36f25b11558.
This reverts commit d97b2b389a0e511c65af6845119eb08b8a2cb473.
This patch adds a variety of classes needed to record variable location
debug-info without using the existing intrinsic approach, see the rationale
at [0].
The two added files and corresponding unit tests are the majority of the
plumbing required for this, but at this point isn't accessible from the
rest of LLVM as we need to stage it into the repo gently. An overview is
that classes are added for recording variable information attached to Real
(TM) instructions, in the form of DPValues and DPMarker objects. The
metadata-uses of DPValues is plumbed into the metadata hierachy, and a
field added to class Instruction, which are all stimulated in the unit
tests. The next few patches in this series add utilities to convert to/from
this new debug-info format and add instruction/block utilities to have
debug-info automatically updated in the background when various operations
occur.
This patch was reviewed in Phab in D153990 and D154080, I've squashed them
together into this commit as there are dependencies between the two
patches, and there's little profit in landing them separately.
[0] https://discourse.llvm.org/t/rfc-instruction-api-changes-needed-to-eliminate-debug-intrinsics-from-ir/68939
