-
Notifications
You must be signed in to change notification settings - Fork 11.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[LLVM][DebugInfo] Refactor some code for easier sharing. #82153
Conversation
Refactor the code that calculates the offsets for the various pieces of the DWARF .debug_names index section, to make it easier to share the code with other tools, such as LLD.
@llvm/pr-subscribers-debuginfo Author: None (cmtice) ChangesRefactor the code that calculates the offsets for the various pieces of the DWARF .debug_names index section, to make it easier to share the code with other tools, such as LLD. Full diff: https://github.com/llvm/llvm-project/pull/82153.diff 2 Files Affected:
diff --git a/llvm/include/llvm/DebugInfo/DWARF/DWARFAcceleratorTable.h b/llvm/include/llvm/DebugInfo/DWARF/DWARFAcceleratorTable.h
index a26c44bf7e9c29..0dd5ed55efe702 100644
--- a/llvm/include/llvm/DebugInfo/DWARF/DWARFAcceleratorTable.h
+++ b/llvm/include/llvm/DebugInfo/DWARF/DWARFAcceleratorTable.h
@@ -562,6 +562,17 @@ class DWARFDebugNames : public DWARFAcceleratorTable {
uint64_t getEntryOffset() const { return EntryOffset; }
};
+ // Offsets for the start of various important tables from the start of the
+ // section.
+ struct DWARFDebugNamesOffsets {
+ uint64_t CUsBase;
+ uint64_t BucketsBase;
+ uint64_t HashesBase;
+ uint64_t StringOffsetsBase;
+ uint64_t EntryOffsetsBase;
+ uint64_t EntriesBase;
+ };
+
/// Represents a single accelerator table within the DWARF v5 .debug_names
/// section.
class NameIndex {
@@ -572,12 +583,7 @@ class DWARFDebugNames : public DWARFAcceleratorTable {
// Base of the whole unit and of various important tables, as offsets from
// the start of the section.
uint64_t Base;
- uint64_t CUsBase;
- uint64_t BucketsBase;
- uint64_t HashesBase;
- uint64_t StringOffsetsBase;
- uint64_t EntryOffsetsBase;
- uint64_t EntriesBase;
+ DWARFDebugNamesOffsets Offsets;
void dumpCUs(ScopedPrinter &W) const;
void dumpLocalTUs(ScopedPrinter &W) const;
@@ -638,7 +644,7 @@ class DWARFDebugNames : public DWARFAcceleratorTable {
/// Returns the Entry at the relative `Offset` from the start of the Entry
/// pool.
Expected<Entry> getEntryAtRelativeOffset(uint64_t Offset) const {
- auto OffsetFromSection = Offset + this->EntriesBase;
+ auto OffsetFromSection = Offset + this->Offsets.EntriesBase;
return getEntry(&OffsetFromSection);
}
@@ -793,6 +799,14 @@ class DWARFDebugNames : public DWARFAcceleratorTable {
const NameIndex *getCUNameIndex(uint64_t CUOffset);
};
+// Calculates the starting offsets for various sections within the
+// debug_names section. This has been pulled out of DWARFDebugNames, to
+// allow easier sharing of the code with other tools, such as LLD.
+uint64_t FindDebugNamesOffsets(DWARFDebugNames::DWARFDebugNamesOffsets &Offsets,
+ uint64_t HdrSize,
+ const dwarf::DwarfFormat Format,
+ const DWARFDebugNames::Header &Hdr);
+
/// If `Name` is the name of a templated function that includes template
/// parameters, returns a substring of `Name` containing no template
/// parameters.
diff --git a/llvm/lib/DebugInfo/DWARF/DWARFAcceleratorTable.cpp b/llvm/lib/DebugInfo/DWARF/DWARFAcceleratorTable.cpp
index 78f819dd052aac..68adad517f9364 100644
--- a/llvm/lib/DebugInfo/DWARF/DWARFAcceleratorTable.cpp
+++ b/llvm/lib/DebugInfo/DWARF/DWARFAcceleratorTable.cpp
@@ -510,7 +510,7 @@ DWARFDebugNames::Abbrev DWARFDebugNames::AbbrevMapInfo::getTombstoneKey() {
Expected<DWARFDebugNames::AttributeEncoding>
DWARFDebugNames::NameIndex::extractAttributeEncoding(uint64_t *Offset) {
- if (*Offset >= EntriesBase) {
+ if (*Offset >= Offsets.EntriesBase) {
return createStringError(errc::illegal_byte_sequence,
"Incorrectly terminated abbreviation table.");
}
@@ -536,7 +536,7 @@ DWARFDebugNames::NameIndex::extractAttributeEncodings(uint64_t *Offset) {
Expected<DWARFDebugNames::Abbrev>
DWARFDebugNames::NameIndex::extractAbbrev(uint64_t *Offset) {
- if (*Offset >= EntriesBase) {
+ if (*Offset >= Offsets.EntriesBase) {
return createStringError(errc::illegal_byte_sequence,
"Incorrectly terminated abbreviation table.");
}
@@ -552,6 +552,37 @@ DWARFDebugNames::NameIndex::extractAbbrev(uint64_t *Offset) {
return Abbrev(Code, dwarf::Tag(Tag), AbbrevOffset, std::move(*AttrEncOr));
}
+uint64_t llvm::FindDebugNamesOffsets(
+ DWARFDebugNames::DWARFDebugNamesOffsets &Offsets,
+ uint64_t HdrSize,
+ dwarf::DwarfFormat Format,
+ const DWARFDebugNames::Header &Hdr) {
+ uint32_t DwarfSize = (Format == llvm::dwarf::DwarfFormat::DWARF32) ? 4 : 8;
+ uint64_t Offset = HdrSize;
+ Offsets.CUsBase = Offset;
+ Offset += Hdr.CompUnitCount * DwarfSize;
+ Offset += Hdr.LocalTypeUnitCount * DwarfSize;
+ Offset += Hdr.ForeignTypeUnitCount * 8;
+
+ Offsets.BucketsBase = Offset;
+ Offset += Hdr.BucketCount * 4;
+
+ Offsets.HashesBase = Offset;
+ if (Hdr.BucketCount > 0)
+ Offset += Hdr.NameCount * 4;
+
+ Offsets.StringOffsetsBase = Offset;
+ Offset += Hdr.NameCount * DwarfSize;
+
+ Offsets.EntryOffsetsBase = Offset;
+ Offset += Hdr.NameCount * DwarfSize;
+
+ Offset += Hdr.AbbrevTableSize;
+ Offsets.EntriesBase = Offset;
+
+ return Offset;
+}
+
Error DWARFDebugNames::NameIndex::extract() {
const DWARFDataExtractor &AS = Section.AccelSection;
uint64_t Offset = Base;
@@ -559,25 +590,15 @@ Error DWARFDebugNames::NameIndex::extract() {
return E;
const unsigned SectionOffsetSize = dwarf::getDwarfOffsetByteSize(Hdr.Format);
- CUsBase = Offset;
- Offset += Hdr.CompUnitCount * SectionOffsetSize;
- Offset += Hdr.LocalTypeUnitCount * SectionOffsetSize;
- Offset += Hdr.ForeignTypeUnitCount * 8;
- BucketsBase = Offset;
- Offset += Hdr.BucketCount * 4;
- HashesBase = Offset;
- if (Hdr.BucketCount > 0)
- Offset += Hdr.NameCount * 4;
- StringOffsetsBase = Offset;
- Offset += Hdr.NameCount * SectionOffsetSize;
- EntryOffsetsBase = Offset;
- Offset += Hdr.NameCount * SectionOffsetSize;
+ Offset = FindDebugNamesOffsets(Offsets, Offset, Hdr.Format, Hdr);
+
+ Offset = Offsets.EntryOffsetsBase + (Hdr.NameCount * SectionOffsetSize);
if (!AS.isValidOffsetForDataOfSize(Offset, Hdr.AbbrevTableSize))
return createStringError(errc::illegal_byte_sequence,
"Section too small: cannot read abbreviations.");
- EntriesBase = Offset + Hdr.AbbrevTableSize;
+ Offsets.EntriesBase = Offset + Hdr.AbbrevTableSize;
for (;;) {
auto AbbrevOr = extractAbbrev(&Offset);
@@ -679,7 +700,7 @@ void DWARFDebugNames::Entry::dumpParentIdx(
return;
}
- auto AbsoluteOffset = NameIdx->EntriesBase + FormValue.getRawUValue();
+ auto AbsoluteOffset = NameIdx->Offsets.EntriesBase + FormValue.getRawUValue();
W.getOStream() << "Entry @ 0x" + Twine::utohexstr(AbsoluteOffset);
}
@@ -708,14 +729,14 @@ std::error_code DWARFDebugNames::SentinelError::convertToErrorCode() const {
uint64_t DWARFDebugNames::NameIndex::getCUOffset(uint32_t CU) const {
assert(CU < Hdr.CompUnitCount);
const unsigned SectionOffsetSize = dwarf::getDwarfOffsetByteSize(Hdr.Format);
- uint64_t Offset = CUsBase + SectionOffsetSize * CU;
+ uint64_t Offset = Offsets.CUsBase + SectionOffsetSize * CU;
return Section.AccelSection.getRelocatedValue(SectionOffsetSize, &Offset);
}
uint64_t DWARFDebugNames::NameIndex::getLocalTUOffset(uint32_t TU) const {
assert(TU < Hdr.LocalTypeUnitCount);
const unsigned SectionOffsetSize = dwarf::getDwarfOffsetByteSize(Hdr.Format);
- uint64_t Offset = CUsBase + SectionOffsetSize * (Hdr.CompUnitCount + TU);
+ uint64_t Offset = Offsets.CUsBase + SectionOffsetSize * (Hdr.CompUnitCount + TU);
return Section.AccelSection.getRelocatedValue(SectionOffsetSize, &Offset);
}
@@ -723,7 +744,7 @@ uint64_t DWARFDebugNames::NameIndex::getForeignTUSignature(uint32_t TU) const {
assert(TU < Hdr.ForeignTypeUnitCount);
const unsigned SectionOffsetSize = dwarf::getDwarfOffsetByteSize(Hdr.Format);
uint64_t Offset =
- CUsBase +
+ Offsets.CUsBase +
SectionOffsetSize * (Hdr.CompUnitCount + Hdr.LocalTypeUnitCount) + 8 * TU;
return Section.AccelSection.getU64(&Offset);
}
@@ -759,28 +780,28 @@ DWARFDebugNames::NameIndex::getNameTableEntry(uint32_t Index) const {
assert(0 < Index && Index <= Hdr.NameCount);
const unsigned SectionOffsetSize = dwarf::getDwarfOffsetByteSize(Hdr.Format);
uint64_t StringOffsetOffset =
- StringOffsetsBase + SectionOffsetSize * (Index - 1);
+ Offsets.StringOffsetsBase + SectionOffsetSize * (Index - 1);
uint64_t EntryOffsetOffset =
- EntryOffsetsBase + SectionOffsetSize * (Index - 1);
+ Offsets.EntryOffsetsBase + SectionOffsetSize * (Index - 1);
const DWARFDataExtractor &AS = Section.AccelSection;
uint64_t StringOffset =
AS.getRelocatedValue(SectionOffsetSize, &StringOffsetOffset);
uint64_t EntryOffset = AS.getUnsigned(&EntryOffsetOffset, SectionOffsetSize);
- EntryOffset += EntriesBase;
+ EntryOffset += Offsets.EntriesBase;
return {Section.StringSection, Index, StringOffset, EntryOffset};
}
uint32_t
DWARFDebugNames::NameIndex::getBucketArrayEntry(uint32_t Bucket) const {
assert(Bucket < Hdr.BucketCount);
- uint64_t BucketOffset = BucketsBase + 4 * Bucket;
+ uint64_t BucketOffset = Offsets.BucketsBase + 4 * Bucket;
return Section.AccelSection.getU32(&BucketOffset);
}
uint32_t DWARFDebugNames::NameIndex::getHashArrayEntry(uint32_t Index) const {
assert(0 < Index && Index <= Hdr.NameCount);
- uint64_t HashOffset = HashesBase + 4 * (Index - 1);
+ uint64_t HashOffset = Offsets.HashesBase + 4 * (Index - 1);
return Section.AccelSection.getU32(&HashOffset);
}
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
Clang format fixes.
@@ -562,6 +562,17 @@ class DWARFDebugNames : public DWARFAcceleratorTable { | |||
uint64_t getEntryOffset() const { return EntryOffset; } | |||
}; | |||
|
|||
// Offsets for the start of various important tables from the start of the | |||
// section. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you use Doxygen cmments ///
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done.
@@ -793,6 +799,14 @@ class DWARFDebugNames : public DWARFAcceleratorTable { | |||
const NameIndex *getCUNameIndex(uint64_t CUOffset); | |||
}; | |||
|
|||
// Calculates the starting offsets for various sections within the | |||
// debug_names section. This has been pulled out of DWARFDebugNames, to | |||
// allow easier sharing of the code with other tools, such as LLD. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
///
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This has been pulled out of DWARFDebugNames, to
// allow easier sharing of the code with other tools, such as LLD.
This is more of a review comment than documentation, I suggest removing this
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done.
Offset += Hdr.NameCount * SectionOffsetSize; | ||
EntryOffsetsBase = Offset; | ||
Offset += Hdr.NameCount * SectionOffsetSize; | ||
Offset = FindDebugNamesOffsets(Offsets, Offset, Hdr.Format, Hdr); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is strange that we assign to Offset
, but then overwrite it on the next line?
So what's the return value for? Perhaps if it's needed by the future callers but not right now - maybe it should be in the Offsets
structure, and that Offsets
structure should be returned by this function, rather than as an input parameter?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah, something feels off here. Suggestion:
- rename
Offset
toEndOfHeaderOffset
- remove the return value from
FindDebugNamesOffset
- declare
Offset
in the line below
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Did something very similar to what felipe recommended.
Change 'FindDebugNamesOffsets' function: make the return type 'void' (return nothing); update the name to start with lowercase letter. Break 'offset' variable in NameIndex::extract into two parts: hdrSize (to be passed to Header::extract and findDebugNamesOffsets) and 'Offset', to be used for offsets after call to findDebugNamesOffsets. Change comments in header file to doxygen comments.
I think I have addressed all the review comments; if there's something else I need to do please let me know. |
Clang format fixes.
Ping? Is this OK to commit now? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A few minor things to address here in follow-up commits, please. Excuse the delay.
void llvm::findDebugNamesOffsets( | ||
DWARFDebugNames::DWARFDebugNamesOffsets &Offsets, uint64_t HdrSize, | ||
dwarf::DwarfFormat Format, const DWARFDebugNames::Header &Hdr) { | ||
uint32_t DwarfSize = (Format == llvm::dwarf::DwarfFormat::DWARF64) ? 8 : 4; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This should use dwarf::getDwarfOffsetByteSize
@@ -552,32 +552,50 @@ DWARFDebugNames::NameIndex::extractAbbrev(uint64_t *Offset) { | |||
return Abbrev(Code, dwarf::Tag(Tag), AbbrevOffset, std::move(*AttrEncOr)); | |||
} | |||
|
|||
void llvm::findDebugNamesOffsets( | |||
DWARFDebugNames::DWARFDebugNamesOffsets &Offsets, uint64_t HdrSize, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
HdrSize
isn't an accurate name for this parameter, I think @felipepiovezan mentioned EndOfHeaderOffset
as a name in #82153 (comment) which seems accurate.
(the issue here is that while HdrSize
and EndOfHeaderOffset
would be identical for the first debug names table in a .debug_names
section - they aren't identical for the second/other subsequent contributions - EndOfHeaderOffset
is the offset in the .debug_names section after the header we care about here, which might be the Nth header, not necessarily the first)
@@ -552,32 +552,50 @@ DWARFDebugNames::NameIndex::extractAbbrev(uint64_t *Offset) { | |||
return Abbrev(Code, dwarf::Tag(Tag), AbbrevOffset, std::move(*AttrEncOr)); | |||
} | |||
|
|||
void llvm::findDebugNamesOffsets( | |||
DWARFDebugNames::DWARFDebugNamesOffsets &Offsets, uint64_t HdrSize, | |||
dwarf::DwarfFormat Format, const DWARFDebugNames::Header &Hdr) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Probably remove the Format
parameter, since it's accessible from the header?
@@ -552,32 +552,50 @@ DWARFDebugNames::NameIndex::extractAbbrev(uint64_t *Offset) { | |||
return Abbrev(Code, dwarf::Tag(Tag), AbbrevOffset, std::move(*AttrEncOr)); | |||
} | |||
|
|||
void llvm::findDebugNamesOffsets( |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Probably return the offsets from this function, rather than taking them as an out parameter? It's then clearer that they aren't in/out parameters, or something else more interesting.
Refactor the code that calculates the offsets for the various pieces of the DWARF .debug_names index section, to make it easier to share the code with other tools, such as LLD.