-
Notifications
You must be signed in to change notification settings - Fork 13.3k
Adapt table sizes to the contents, accommodating u64 rmeta offsets #113542
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
Merged
Merged
Changes from all commits
Commits
Show all changes
3 commits
Select commit
Hold shift + click to select a range
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -38,6 +38,12 @@ impl IsDefault for u32 { | |
} | ||
} | ||
|
||
impl IsDefault for u64 { | ||
fn is_default(&self) -> bool { | ||
*self == 0 | ||
} | ||
} | ||
|
||
impl<T> IsDefault for LazyArray<T> { | ||
fn is_default(&self) -> bool { | ||
self.num_elems == 0 | ||
|
@@ -89,6 +95,20 @@ impl FixedSizeEncoding for u32 { | |
} | ||
} | ||
|
||
impl FixedSizeEncoding for u64 { | ||
type ByteArray = [u8; 8]; | ||
|
||
#[inline] | ||
fn from_bytes(b: &[u8; 8]) -> Self { | ||
Self::from_le_bytes(*b) | ||
} | ||
|
||
#[inline] | ||
fn write_to_bytes(self, b: &mut [u8; 8]) { | ||
*b = self.to_le_bytes(); | ||
} | ||
} | ||
|
||
macro_rules! fixed_size_enum { | ||
($ty:ty { $(($($pat:tt)*))* }) => { | ||
impl FixedSizeEncoding for Option<$ty> { | ||
|
@@ -299,21 +319,21 @@ impl FixedSizeEncoding for UnusedGenericParams { | |
// generic `LazyValue<T>` impl, but in the general case we might not need / want | ||
// to fit every `usize` in `u32`. | ||
impl<T> FixedSizeEncoding for Option<LazyValue<T>> { | ||
type ByteArray = [u8; 4]; | ||
type ByteArray = [u8; 8]; | ||
|
||
#[inline] | ||
fn from_bytes(b: &[u8; 4]) -> Self { | ||
let position = NonZeroUsize::new(u32::from_bytes(b) as usize)?; | ||
fn from_bytes(b: &[u8; 8]) -> Self { | ||
let position = NonZeroUsize::new(u64::from_bytes(b) as usize)?; | ||
Some(LazyValue::from_position(position)) | ||
} | ||
|
||
#[inline] | ||
fn write_to_bytes(self, b: &mut [u8; 4]) { | ||
fn write_to_bytes(self, b: &mut [u8; 8]) { | ||
match self { | ||
None => unreachable!(), | ||
Some(lazy) => { | ||
let position = lazy.position.get(); | ||
let position: u32 = position.try_into().unwrap(); | ||
let position: u64 = position.try_into().unwrap(); | ||
position.write_to_bytes(b) | ||
} | ||
} | ||
|
@@ -322,55 +342,75 @@ impl<T> FixedSizeEncoding for Option<LazyValue<T>> { | |
|
||
impl<T> LazyArray<T> { | ||
#[inline] | ||
fn write_to_bytes_impl(self, b: &mut [u8; 8]) { | ||
let ([position_bytes, meta_bytes], []) = b.as_chunks_mut::<4>() else { panic!() }; | ||
|
||
let position = self.position.get(); | ||
let position: u32 = position.try_into().unwrap(); | ||
position.write_to_bytes(position_bytes); | ||
|
||
let len = self.num_elems; | ||
let len: u32 = len.try_into().unwrap(); | ||
len.write_to_bytes(meta_bytes); | ||
fn write_to_bytes_impl(self, b: &mut [u8; 16]) { | ||
let position = (self.position.get() as u64).to_le_bytes(); | ||
let len = (self.num_elems as u64).to_le_bytes(); | ||
|
||
// Element width is selected at runtime on a per-table basis by omitting trailing | ||
// zero bytes in table elements. This works very naturally when table elements are | ||
// simple numbers but `LazyArray` is a pair of integers. If naively encoded, the second | ||
// element would shield the trailing zeroes in the first. Interleaving the bytes | ||
// of the position and length exposes trailing zeroes in both to the optimization. | ||
// We encode length second because we generally expect it to be smaller. | ||
for i in 0..8 { | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Can you please write a comment here stating the motivation for this interleaving? This code would be very difficult to understand without context. |
||
b[2 * i] = position[i]; | ||
b[2 * i + 1] = len[i]; | ||
} | ||
} | ||
|
||
fn from_bytes_impl(position_bytes: &[u8; 4], meta_bytes: &[u8; 4]) -> Option<LazyArray<T>> { | ||
let position = NonZeroUsize::new(u32::from_bytes(position_bytes) as usize)?; | ||
let len = u32::from_bytes(meta_bytes) as usize; | ||
fn from_bytes_impl(position: &[u8; 8], meta: &[u8; 8]) -> Option<LazyArray<T>> { | ||
let position = NonZeroUsize::new(u64::from_bytes(&position) as usize)?; | ||
let len = u64::from_bytes(&meta) as usize; | ||
Some(LazyArray::from_position_and_num_elems(position, len)) | ||
} | ||
} | ||
|
||
// Decoding helper for the encoding scheme used by `LazyArray`. | ||
// Interleaving the bytes of the two integers exposes trailing bytes in the first integer | ||
// to the varint scheme that we use for tables. | ||
#[inline] | ||
fn decode_interleaved(encoded: &[u8; 16]) -> ([u8; 8], [u8; 8]) { | ||
let mut first = [0u8; 8]; | ||
let mut second = [0u8; 8]; | ||
for i in 0..8 { | ||
first[i] = encoded[2 * i]; | ||
second[i] = encoded[2 * i + 1]; | ||
} | ||
(first, second) | ||
} | ||
|
||
impl<T> FixedSizeEncoding for LazyArray<T> { | ||
type ByteArray = [u8; 8]; | ||
type ByteArray = [u8; 16]; | ||
|
||
#[inline] | ||
fn from_bytes(b: &[u8; 8]) -> Self { | ||
let ([position_bytes, meta_bytes], []) = b.as_chunks::<4>() else { panic!() }; | ||
if *meta_bytes == [0; 4] { | ||
fn from_bytes(b: &[u8; 16]) -> Self { | ||
let (position, meta) = decode_interleaved(b); | ||
|
||
if meta == [0; 8] { | ||
return Default::default(); | ||
} | ||
LazyArray::from_bytes_impl(position_bytes, meta_bytes).unwrap() | ||
LazyArray::from_bytes_impl(&position, &meta).unwrap() | ||
} | ||
|
||
#[inline] | ||
fn write_to_bytes(self, b: &mut [u8; 8]) { | ||
fn write_to_bytes(self, b: &mut [u8; 16]) { | ||
assert!(!self.is_default()); | ||
self.write_to_bytes_impl(b) | ||
} | ||
} | ||
|
||
impl<T> FixedSizeEncoding for Option<LazyArray<T>> { | ||
type ByteArray = [u8; 8]; | ||
type ByteArray = [u8; 16]; | ||
|
||
#[inline] | ||
fn from_bytes(b: &[u8; 8]) -> Self { | ||
let ([position_bytes, meta_bytes], []) = b.as_chunks::<4>() else { panic!() }; | ||
LazyArray::from_bytes_impl(position_bytes, meta_bytes) | ||
fn from_bytes(b: &[u8; 16]) -> Self { | ||
let (position, meta) = decode_interleaved(b); | ||
|
||
LazyArray::from_bytes_impl(&position, &meta) | ||
} | ||
|
||
#[inline] | ||
fn write_to_bytes(self, b: &mut [u8; 8]) { | ||
fn write_to_bytes(self, b: &mut [u8; 16]) { | ||
match self { | ||
None => unreachable!(), | ||
Some(lazy) => lazy.write_to_bytes_impl(b), | ||
|
@@ -380,13 +420,14 @@ impl<T> FixedSizeEncoding for Option<LazyArray<T>> { | |
|
||
/// Helper for constructing a table's serialization (also see `Table`). | ||
pub(super) struct TableBuilder<I: Idx, T: FixedSizeEncoding> { | ||
width: usize, | ||
blocks: IndexVec<I, T::ByteArray>, | ||
_marker: PhantomData<T>, | ||
} | ||
|
||
impl<I: Idx, T: FixedSizeEncoding> Default for TableBuilder<I, T> { | ||
fn default() -> Self { | ||
TableBuilder { blocks: Default::default(), _marker: PhantomData } | ||
TableBuilder { width: 0, blocks: Default::default(), _marker: PhantomData } | ||
} | ||
} | ||
|
||
|
@@ -414,40 +455,63 @@ impl<I: Idx, const N: usize, T: FixedSizeEncoding<ByteArray = [u8; N]>> TableBui | |
// > store bit-masks of which item in each bucket is actually serialized). | ||
let block = self.blocks.ensure_contains_elem(i, || [0; N]); | ||
value.write_to_bytes(block); | ||
if self.width != N { | ||
let width = N - trailing_zeros(block); | ||
self.width = self.width.max(width); | ||
} | ||
} | ||
} | ||
|
||
pub(crate) fn encode(&self, buf: &mut FileEncoder) -> LazyTable<I, T> { | ||
let pos = buf.position(); | ||
|
||
let width = self.width; | ||
for block in &self.blocks { | ||
buf.emit_raw_bytes(block); | ||
buf.emit_raw_bytes(&block[..width]); | ||
} | ||
let num_bytes = self.blocks.len() * N; | ||
|
||
LazyTable::from_position_and_encoded_size( | ||
NonZeroUsize::new(pos as usize).unwrap(), | ||
num_bytes, | ||
width, | ||
self.blocks.len(), | ||
) | ||
} | ||
} | ||
|
||
fn trailing_zeros(x: &[u8]) -> usize { | ||
x.iter().rev().take_while(|b| **b == 0).count() | ||
} | ||
|
||
impl<I: Idx, const N: usize, T: FixedSizeEncoding<ByteArray = [u8; N]> + ParameterizedOverTcx> | ||
LazyTable<I, T> | ||
where | ||
for<'tcx> T::Value<'tcx>: FixedSizeEncoding<ByteArray = [u8; N]>, | ||
{ | ||
/// Given the metadata, extract out the value at a particular index (if any). | ||
#[inline(never)] | ||
pub(super) fn get<'a, 'tcx, M: Metadata<'a, 'tcx>>(&self, metadata: M, i: I) -> T::Value<'tcx> { | ||
trace!("LazyTable::lookup: index={:?} len={:?}", i, self.encoded_size); | ||
trace!("LazyTable::lookup: index={:?} len={:?}", i, self.len); | ||
|
||
// Access past the end of the table returns a Default | ||
if i.index() >= self.len { | ||
return Default::default(); | ||
} | ||
|
||
let start = self.position.get(); | ||
let bytes = &metadata.blob()[start..start + self.encoded_size]; | ||
let (bytes, []) = bytes.as_chunks::<N>() else { panic!() }; | ||
bytes.get(i.index()).map_or_else(Default::default, FixedSizeEncoding::from_bytes) | ||
let width = self.width; | ||
let start = self.position.get() + (width * i.index()); | ||
let end = start + width; | ||
let bytes = &metadata.blob()[start..end]; | ||
|
||
if let Ok(fixed) = bytes.try_into() { | ||
FixedSizeEncoding::from_bytes(fixed) | ||
} else { | ||
let mut fixed = [0u8; N]; | ||
fixed[..width].copy_from_slice(bytes); | ||
FixedSizeEncoding::from_bytes(&fixed) | ||
} | ||
} | ||
|
||
/// Size of the table in entries, including possible gaps. | ||
pub(super) fn size(&self) -> usize { | ||
self.encoded_size / N | ||
self.len | ||
} | ||
} |
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
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.
Maybe add some comments on those?