Before this patch, the maximum size of the GHASH table was 2^31 buckets. However we were storing the bucket index into a TypeIndex which has an input limit of (2^31)-4095 indices, see this link. Any value above that limit will improperly set the TypeIndex's high bit, which is interpreted as DecoratedItemIdMask. This used to cause bad indices on extraction when calling TypeIndex::toArrayIndex().
Differential Revision: https://reviews.llvm.org/D103297
Previously we simply didn't check this. Prereq to make the test suite
pass with ghash enabled by default.
Differential Revision: https://reviews.llvm.org/D102885
Before this patch, when using LLD with /DEBUG:GHASH and MSVC precomp.OBJ files, we had a bunch of:
lld-link: warning: S_[GL]PROC32ID record in blabla.obj refers to PDB item index 0x206ED1 which is not a LF[M]FUNC_ID record
This was caused by LF_FUNC_ID and LF_MFUNC_ID which didn't have correct mapping to the corresponding TPI records. The root issue was that the indexMapStorage was improperly re-assembled in UsePrecompSource::remapTpiWithGHashes.
After this patch, /DEBUG and /DEBUG:GHASH produce exactly the same debug infos in the PDB.
Differential Revision: https://reviews.llvm.org/D93732
This adds the following two new lines to /summary:
21351 Input OBJ files (expanded from all cmd-line inputs)
61 PDB type server dependencies
38 Precomp OBJ dependencies
1420669231 Input type records <<<<
78665073382 Input type records bytes <<<<
8801393 Merged TPI records
3177158 Merged IPI records
59194 Output PDB strings
71576766 Global symbol records
25416935 Module symbol records
2103431 Public symbol records
Differential Revision: https://reviews.llvm.org/D88703
Before this patch /summary was crashing with some .PCH.OBJ files, because tpiMap[srcIdx++] was reading at the wrong location. When the TpiSource depends on a .PCH.OBJ file, the types should be offset by the previously merged PCH.OBJ set of indices.
Differential Revision: https://reviews.llvm.org/D88678
Stored Error objects have to be checked, even if they are success
values.
This reverts commit 8d250ac3cd.
Relands commit 49b3459930655d879b2dc190ff8fe11c38a8be5f..
Original commit message:
-----------------------------------------
This makes type merging much faster (-24% on chrome.dll) when multiple
threads are available, but it slightly increases the time to link (+10%)
when /threads:1 is passed. With only one more thread, the new type
merging is faster (-11%). The output PDB should be identical to what it
was before this change.
To give an idea, here is the /time output placed side by side:
BEFORE | AFTER
Input File Reading: 956 ms | 968 ms
Code Layout: 258 ms | 190 ms
Commit Output File: 6 ms | 7 ms
PDB Emission (Cumulative): 6691 ms | 4253 ms
Add Objects: 4341 ms | 2927 ms
Type Merging: 2814 ms | 1269 ms -55%!
Symbol Merging: 1509 ms | 1645 ms
Publics Stream Layout: 111 ms | 112 ms
TPI Stream Layout: 764 ms | 26 ms trivial
Commit to Disk: 1322 ms | 1036 ms -300ms
----------------------------------------- --------
Total Link Time: 8416 ms 5882 ms -30% overall
The main source of the additional overhead in the single-threaded case
is the need to iterate all .debug$T sections up front to check which
type records should go in the IPI stream. See fillIsItemIndexFromDebugT.
With changes to the .debug$H section, we could pre-calculate this info
and eliminate the need to do this walk up front. That should restore
single-threaded performance back to what it was before this change.
This change will cause LLD to be much more parallel than it used to, and
for users who do multiple links in parallel, it could regress
performance. However, when the user is only doing one link, it's a huge
improvement. In the future, we can use NT worker threads to avoid
oversaturating the machine with work, but for now, this is such an
improvement for the single-link use case that I think we should land
this as is.
Algorithm
----------
Before this change, we essentially used a
DenseMap<GloballyHashedType, TypeIndex> to check if a type has already
been seen, and if it hasn't been seen, insert it now and use the next
available type index for it in the destination type stream. DenseMap
does not support concurrent insertion, and even if it did, the linker
must be deterministic: it cannot produce different PDBs by using
different numbers of threads. The output type stream must be in the same
order regardless of the order of hash table insertions.
In order to create a hash table that supports concurrent insertion, the
table cells must be small enough that they can be updated atomically.
The algorithm I used for updating the table using linear probing is
described in this paper, "Concurrent Hash Tables: Fast and General(?)!":
https://dl.acm.org/doi/10.1145/3309206
The GHashCell in this change is essentially a pair of 32-bit integer
indices: <sourceIndex, typeIndex>. The sourceIndex is the index of the
TpiSource object, and it represents an input type stream. The typeIndex
is the index of the type in the stream. Together, we have something like
a ragged 2D array of ghashes, which can be looked up as:
tpiSources[tpiSrcIndex]->ghashes[typeIndex]
By using these side tables, we can omit the key data from the hash
table, and keep the table cell small. There is a cost to this: resolving
hash table collisions requires many more loads than simply looking at
the key in the same cache line as the insertion position. However, most
supported platforms should have a 64-bit CAS operation to update the
cell atomically.
To make the result of concurrent insertion deterministic, the cell
payloads must have a priority function. Defining one is pretty
straightforward: compare the two 32-bit numbers as a combined 64-bit
number. This means that types coming from inputs earlier on the command
line have a higher priority and are more likely to appear earlier in the
final PDB type stream than types from an input appearing later on the
link line.
After table insertion, the non-empty cells in the table can be copied
out of the main table and sorted by priority to determine the ordering
of the final type index stream. At this point, item and type records
must be separated, either by sorting or by splitting into two arrays,
and I chose sorting. This is why the GHashCell must contain the isItem
bit.
Once the final PDB TPI stream ordering is known, we need to compute a
mapping from source type index to PDB type index. To avoid starting over
from scratch and looking up every type again by its ghash, we save the
insertion position of every hash table insertion during the first
insertion phase. Because the table does not support rehashing, the
insertion position is stable. Using the array of insertion positions
indexed by source type index, we can replace the source type indices in
the ghash table cells with the PDB type indices.
Once the table cells have been updated to contain PDB type indices, the
mapping for each type source can be computed in parallel. Simply iterate
the list of cell positions and replace them with the PDB type index,
since the insertion positions are no longer needed.
Once we have a source to destination type index mapping for every type
source, there are no more data dependencies. We know which type records
are "unique" (not duplicates), and what their final type indices will
be. We can do the remapping in parallel, and accumulate type sizes and
type hashes in parallel by type source.
Lastly, TPI stream layout must be done serially. Accumulate all the type
records, sizes, and hashes, and add them to the PDB.
Differential Revision: https://reviews.llvm.org/D87805
This makes type merging much faster (-24% on chrome.dll) when multiple
threads are available, but it slightly increases the time to link (+10%)
when /threads:1 is passed. With only one more thread, the new type
merging is faster (-11%). The output PDB should be identical to what it
was before this change.
To give an idea, here is the /time output placed side by side:
BEFORE | AFTER
Input File Reading: 956 ms | 968 ms
Code Layout: 258 ms | 190 ms
Commit Output File: 6 ms | 7 ms
PDB Emission (Cumulative): 6691 ms | 4253 ms
Add Objects: 4341 ms | 2927 ms
Type Merging: 2814 ms | 1269 ms -55%!
Symbol Merging: 1509 ms | 1645 ms
Publics Stream Layout: 111 ms | 112 ms
TPI Stream Layout: 764 ms | 26 ms trivial
Commit to Disk: 1322 ms | 1036 ms -300ms
----------------------------------------- --------
Total Link Time: 8416 ms 5882 ms -30% overall
The main source of the additional overhead in the single-threaded case
is the need to iterate all .debug$T sections up front to check which
type records should go in the IPI stream. See fillIsItemIndexFromDebugT.
With changes to the .debug$H section, we could pre-calculate this info
and eliminate the need to do this walk up front. That should restore
single-threaded performance back to what it was before this change.
This change will cause LLD to be much more parallel than it used to, and
for users who do multiple links in parallel, it could regress
performance. However, when the user is only doing one link, it's a huge
improvement. In the future, we can use NT worker threads to avoid
oversaturating the machine with work, but for now, this is such an
improvement for the single-link use case that I think we should land
this as is.
Algorithm
----------
Before this change, we essentially used a
DenseMap<GloballyHashedType, TypeIndex> to check if a type has already
been seen, and if it hasn't been seen, insert it now and use the next
available type index for it in the destination type stream. DenseMap
does not support concurrent insertion, and even if it did, the linker
must be deterministic: it cannot produce different PDBs by using
different numbers of threads. The output type stream must be in the same
order regardless of the order of hash table insertions.
In order to create a hash table that supports concurrent insertion, the
table cells must be small enough that they can be updated atomically.
The algorithm I used for updating the table using linear probing is
described in this paper, "Concurrent Hash Tables: Fast and General(?)!":
https://dl.acm.org/doi/10.1145/3309206
The GHashCell in this change is essentially a pair of 32-bit integer
indices: <sourceIndex, typeIndex>. The sourceIndex is the index of the
TpiSource object, and it represents an input type stream. The typeIndex
is the index of the type in the stream. Together, we have something like
a ragged 2D array of ghashes, which can be looked up as:
tpiSources[tpiSrcIndex]->ghashes[typeIndex]
By using these side tables, we can omit the key data from the hash
table, and keep the table cell small. There is a cost to this: resolving
hash table collisions requires many more loads than simply looking at
the key in the same cache line as the insertion position. However, most
supported platforms should have a 64-bit CAS operation to update the
cell atomically.
To make the result of concurrent insertion deterministic, the cell
payloads must have a priority function. Defining one is pretty
straightforward: compare the two 32-bit numbers as a combined 64-bit
number. This means that types coming from inputs earlier on the command
line have a higher priority and are more likely to appear earlier in the
final PDB type stream than types from an input appearing later on the
link line.
After table insertion, the non-empty cells in the table can be copied
out of the main table and sorted by priority to determine the ordering
of the final type index stream. At this point, item and type records
must be separated, either by sorting or by splitting into two arrays,
and I chose sorting. This is why the GHashCell must contain the isItem
bit.
Once the final PDB TPI stream ordering is known, we need to compute a
mapping from source type index to PDB type index. To avoid starting over
from scratch and looking up every type again by its ghash, we save the
insertion position of every hash table insertion during the first
insertion phase. Because the table does not support rehashing, the
insertion position is stable. Using the array of insertion positions
indexed by source type index, we can replace the source type indices in
the ghash table cells with the PDB type indices.
Once the table cells have been updated to contain PDB type indices, the
mapping for each type source can be computed in parallel. Simply iterate
the list of cell positions and replace them with the PDB type index,
since the insertion positions are no longer needed.
Once we have a source to destination type index mapping for every type
source, there are no more data dependencies. We know which type records
are "unique" (not duplicates), and what their final type indices will
be. We can do the remapping in parallel, and accumulate type sizes and
type hashes in parallel by type source.
Lastly, TPI stream layout must be done serially. Accumulate all the type
records, sizes, and hashes, and add them to the PDB.
Differential Revision: https://reviews.llvm.org/D87805
Extending the lifetime of these type index mappings does increase memory
usage (+2% in my case), but it decouples type merging from symbol
merging. This is a pre-requisite for two changes that I have in mind:
- parallel type merging: speeds up slow type merging
- defered symbol merging: avoid heap allocating (relocating) all symbols
This eliminates CVIndexMap and moves its data into TpiSource. The maps
are also split into a SmallVector and ArrayRef component, so that the
ipiMap can alias the tpiMap for /Z7 object files, and so that both maps
can simply alias the PDB type server maps for /Zi files.
Splitting TypeServerSource establishes that all input types to be merged
can be identified with two 32-bit indices:
- The index of the TpiSource object
- The type index of the record
This is useful, because this information can be stored in a single
64-bit atomic word to enable concurrent hashtable insertion.
One last change is that now all object files with debugChunks get a
TpiSource, even if they have no type info. This avoids some null checks
and special cases.
Differential Revision: https://reviews.llvm.org/D87736
This paves the way to doing more things in parallel, and allows us to
order type sources in dependency order. PDBs and PCH objects have to be
loaded before object files which use them.
This is a rebase of the unapplied remaining changes in
https://reviews.llvm.org/D59226. I found it very challenging to rebase
this across the LLD variable name style change. I recall there was a
tool for that, but I didn't take the time to use it.
Reviewers: aganea, akhuang
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D79672
Instead, use `using namespace lld(::coff)`, and fully qualify the names
of free functions where they are defined in cpp files.
This effectively reverts d79c3be618 to follow the new style guide added
in 236fcbc21a.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D74882
This is how it should've been and brings it more in line with
std::string_view. There should be no functional change here.
This is mostly mechanical from a custom clang-tidy check, with a lot of
manual fixups. It uncovers a lot of minor inefficiencies.
This doesn't actually modify StringRef yet, I'll do that in a follow-up.
In lld we rarely use std::unique_ptr but instead allocate new instances
using lld::make<T>() so that they are deallocated at the end of linking.
This patch changes existing code so that that follows the convention.
Differential Revision: https://reviews.llvm.org/D70420
Similar to D67323, but for COFF. Many lld/COFF/ files already use
`namespace lld { namespace coff {`. Only a few need changing.
Reviewed By: ruiu
Differential Revision: https://reviews.llvm.org/D68772
llvm-svn: 374314
Summary:
This is a re-land of r370487 with a fix for the use-after-free bug
that rev contained.
This implements -start-lib and -end-lib flags for lld-link, analogous
to the similarly named options in ld.lld. Object files after
-start-lib are included in the link only when needed to resolve
undefined symbols. The -end-lib flag goes back to the normal behavior
of always including object files in the link. This mimics the
semantics of static libraries, but without needing to actually create
the archive file.
Reviewers: ruiu, smeenai, MaskRay
Reviewed By: ruiu, MaskRay
Subscribers: akhuang, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66848
llvm-svn: 370816
Summary:
This implements -start-lib and -end-lib flags for lld-link, analogous
to the similarly named options in ld.lld. Object files after
-start-lib are included in the link only when needed to resolve
undefined symbols. The -end-lib flag goes back to the normal behavior
of always including object files in the link. This mimics the
semantics of static libraries, but without needing to actually create
the archive file.
Reviewers: ruiu, smeenai, MaskRay
Reviewed By: ruiu, MaskRay
Subscribers: akhuang, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66848
llvm-svn: 370487
This patch does the same thing as r365595 to other subdirectories,
which completes the naming style change for the entire lld directory.
With this, the naming style conversion is complete for lld.
Differential Revision: https://reviews.llvm.org/D64473
llvm-svn: 365730
The age field is only there to say how many times an OBJ or a PDB was incrementally linked. It shouldn't be used to validate the link between the OBJ and the PDB.
Differential Revision: https://reviews.llvm.org/D62837
llvm-svn: 362572
We need to have all input files ready before doing debuginfo type merging.
This patch is moving the late PDB type server discovery much earlier in the process, when the explicit inputs (OBJs, LIBs) are loaded.
The short term goal is to parallelize type merging.
Differential Revision: https://reviews.llvm.org/D60095
llvm-svn: 362393
We need to have all input files ready before doing debuginfo type merging.
This patch is moving the late PDB type server discovery much earlier in the process, when the explicit inputs (OBJs, LIBs) are loaded.
The short term goal is to parallelize type merging.
Differential Revision: https://reviews.llvm.org/D60095
llvm-svn: 361842
We introduce a new class hierarchy for debug types merging (in DebugTypes.h). The end-goal is to parallelize the type merging - please see the plan in D59226.
Previously, dependency discovery was done on the fly, much later, during the type merging loop. Unfortunately, parallelizing the type merging requires the dependencies to be merged in first, before any dependent ObjFile, thus this early discovery.
The overall intention for this path is to discover debug information dependencies at a much earlier stage, when processing input files. Currently, two types of dependency are supported: PDB type servers (when compiling with MSVC /Zi) and precompiled headers OBJs (when compiling with MSVC /Yc and /Yu). Once discovered, an explicit link is added into the dependent ObjFile, through the new debug types class hierarchy introduced in DebugTypes.h.
Differential Revision: https://reviews.llvm.org/D59053
llvm-svn: 357383
Martin Storsjö