| Age | Commit message (Collapse) | Author |
|---|
|
Prior to this commit we only ever protected MemoryChunk's requested_size
field with Valgrind NOACCESS. This means that if the hdrmask field is
ever accessed accidentally then we're not going to get any warnings from
Valgrind about it. Valgrind would have warned us about the problem fixed
in 92957ed98 had we already been doing this.
Per suggestion from Tom Lane
Reviewed-by: Richard Guo
Discussion: https://postgr.es/m/[email protected]
Discussion: https://postgr.es/m/CAApHDvr=FZNGbj252Z6M9BSFKoq6BMxgkQ2yEAGUYoo7RquqZg@mail.gmail.com
|
|
The chunk header is unsigned, and the output format takes unsigned, so
casting it to signed in between is incorrect.
|
|
If asked to decrease the size of a large (>8K) palloc chunk,
AllocSetRealloc could improperly change the Valgrind state of memory
beyond the new end of the chunk: it would mark data UNDEFINED as far
as the old end of the chunk after having done the realloc(3) call,
thus tromping on the state of memory that no longer belongs to it.
One would normally expect that memory to now be marked NOACCESS,
so that this mislabeling might prevent detection of later errors.
If realloc() had chosen to move the chunk someplace else (unlikely,
but well within its rights) we could also mismark perfectly-valid
DEFINED data as UNDEFINED, causing false-positive valgrind reports
later. Also, any malloc bookkeeping data placed within this area
might now be wrongly marked, causing additional problems.
Fix by replacing relevant uses of "oldsize" with "Min(size, oldsize)".
It's sufficient to mark as far as "size" when that's smaller, because
whatever remains in the new chunk size will be marked NOACCESS below,
and we expect realloc() to have taken care of marking the memory
beyond the new official end of the chunk.
While we're here, also rename the function's "oldsize" variable
to "oldchksize" to more clearly explain what it actually holds,
namely the distance to the end of the chunk (that is, requested size
plus trailing padding). This is more consistent with the use of
"size" and "chksize" to hold the new requested size and chunk size.
Add a new variable "oldsize" in the one stanza where we're actually
talking about the old requested size.
Oversight in commit c477f3e44. Back-patch to all supported branches,
as that was, just in case anybody wants to do valgrind testing on back
branches.
Karina Litskevich
Discussion: https://postgr.es/m/CACiT8iaAET-fmzjjZLjaJC4zwSJmrFyL7LAdHwaYyjjQOQ4hcg@mail.gmail.com
|
|
To allow testing for general support for fast bitscan intrinsics,
add symbols HAVE_BITSCAN_REVERSE and HAVE_BITSCAN_FORWARD.
Also do related cleanup in AllocSetFreeIndex(): Previously, we
tested for HAVE__BUILTIN_CLZ and copied the relevant internals of
pg_leftmost_one_pos32(), with a special fallback that does less
work than the general fallback for that function. Now that we have
a more general test, we just call pg_leftmost_one_pos32() directly
for platforms with intrinsic support. On gcc at least, there is no
difference in the binary for non-assert builds.
Discussion: https://www.postgresql.org/message-id/CAFBsxsEPc%2BFnX_0vmmQ5DHv60sk4rL_RZJ%2BMD6ei%3D76L0kFMvA%40mail.gmail.com
|
|
Invent separate macros for "invalid" values of these types, so that
we needn't embed knowledge of their representations into calling code.
These are all zeroes anyway ATM, so this is not fixing any live bug,
but it makes the code cleaner and more future-proof.
I (tgl) also chose to move DSM_HANDLE_INVALID into dsm_impl.h,
since it seems like it should live beside the typedef for dsm_handle.
Hou Zhijie, Nathan Bossart, Kyotaro Horiguchi, Tom Lane
Discussion: https://postgr.es/m/OS0PR01MB5716860B1454C34E5B179B6694C99@OS0PR01MB5716.jpnprd01.prod.outlook.com
|
|
Author: Justin Pryzby
Discussion: https://postgr.es/m/[email protected]
|
|
This fixes a couple of unused variable warnings that could be seen when
compiling with MEMORY_CONTEXT_CHECKING but not USE_ASSERT_CHECKING.
Defining MEMORY_CONTEXT_CHECKING without asserts is a little unusual,
however, we shouldn't be producing any warnings from such a build.
Author: Richard Guo
Discussion: https://postgr.es/m/CAMbWs4_D-vgLEh7eO47p=73u1jWO78NWf6Qfv1FndY1kG-Q-jA@mail.gmail.com
|
|
Backpatch-through: 11
|
|
This introduces palloc_aligned() and MemoryContextAllocAligned() which
allow callers to obtain memory which is allocated to the given size and
also aligned to the specified alignment boundary. The alignment
boundaries may be any power-of-2 value. Currently, the alignment is
capped at 2^26, however, we don't expect values anything like that large.
The primary expected use case is to align allocations to perhaps CPU
cache line size or to maybe I/O page size. Certain use cases can benefit
from having aligned memory by either having better performance or more
predictable performance.
The alignment is achieved by requesting 'alignto' additional bytes from
the underlying allocator function and then aligning the address that is
returned to the requested alignment. This obviously does waste some
memory, so alignments should be kept as small as what is required.
It's also important to note that these alignment bytes eat into the
maximum allocation size. So something like:
palloc_aligned(MaxAllocSize, 64, 0);
will not work as we cannot request MaxAllocSize + 64 bytes.
Additionally, because we're just requesting the requested size plus the
alignment requirements from the given MemoryContext, if that context is
the Slab allocator, then since slab can only provide chunks of the size
that's specified when the slab context is created, then this is not going
to work. Slab will generate an error to indicate that the requested size
is not supported.
The alignment that is requested in palloc_aligned() is stored along with
the allocated memory. This allows the alignment to remain intact through
repalloc() calls.
Author: Andres Freund, David Rowley
Reviewed-by: Maxim Orlov, Andres Freund, John Naylor
Discussion: https://postgr.es/m/CAApHDvpxLPUMV1mhxs6g7GNwCP6Cs6hfnYQL5ffJQTuFAuxt8A%40mail.gmail.com
|
|
d21ded75f changed the way slab.c works but introduced a bug that meant we
could end up with the slab's curBlocklistIndex pointing to the wrong list.
The condition which was checking for this was failing to account for two
things:
1. The curBlocklistIndex could be 0 as we've currently got no non-full
blocks to put chunks on. In this case, the dlist_is_empty() check cannot
be performed as there can be any number of completely full blocks at that
index.
2. The curBlocklistIndex may be greater than the index we just moved the
block onto. Since we need to ensure we fill up fuller blocks first, we
must reset curBlocklistIndex when changing any blocklist element that's
less than the curBlocklistIndex too.
Reported-by: Takamichi Osumi
Discussion: https://postgr.es/m/TYCPR01MB8373329C6329768D7E093D68EDEB9@TYCPR01MB8373.jpnprd01.prod.outlook.com
|
|
Discussion: https://postgr.es/m/[email protected]
|
|
Slab has traditionally been fairly slow when compared with the AllocSet or
Generation memory allocators. Part of this slowness came from having to
write out an entire block when we allocate a new block in order to
populate the free list indexes within the block's memory. Additional
slowness came from having to move a block onto another dlist each time we
palloc or pfree a chunk from it.
Here we optimize both of those cases and do a little bit extra to improve
the performance of the slab allocator.
Here, instead of writing out the free list indexes when allocating a new
block, we introduce the concept of "unused" chunks. When a block is first
allocated all chunks are unused. These chunks only make it onto the
free list when they are pfree'd. When allocating new chunks on an
existing block, we have the choice of consuming a chunk from the free list
or an unused chunk. When both exist, we opt to use one from the free
list, as these have been used already and the memory of them is more
likely to be cached by the CPU.
Here we also reduce the number of block lists from there being one for
every possible value of free chunks on a block to just having a small
fixed number of block lists. We keep the 0th block list for completely
full blocks and anything else stores blocks for some range of free chunks
with fuller blocks appearing on lower block list array elements. This
reduces how often we must move a block to another list when we allocate or
free chunks, but still allows us to prefer to put new chunks on fuller
blocks and perhaps allow blocks with fewer chunks to be free'd later
once all their remaining chunks have been pfree'd.
Additionally, we now store a list of "emptyblocks", which are blocks that
no longer contain any allocated chunks. We now keep up to 10 of these
around to avoid having to thrash malloc/free when allocation patterns
continually cause blocks to become free of any allocated chunks only to
allocate more chunks again. Now only once we have 10 of these, we free
the block. This does raise the high water mark for the total memory that
a slab context can consume. It does not seem entirely unreasonable that
we might one day want to make this a property of SlabContext rather than a
compile-time constant. Let's wait and see if there is any evidence to
support that this is required before doing it.
Author: Andres Freund, David Rowley
Tested-by: Tomas Vondra, John Naylor
Discussion: https://postgr.es/m/[email protected]
|
|
Because we added StaticAssertStmt() first before StaticAssertDecl(),
some uses as well as the instructions in c.h are now a bit backwards
from the "native" way static assertions are meant to be used in C.
This updates the guidance and moves some static assertions to better
places.
Specifically, since the addition of StaticAssertDecl(), we can put
static assertions at the file level. This moves a number of static
assertions out of function bodies, where they might have been stuck
out of necessity, to perhaps better places at the file level or in
header files.
Also, when the static assertion appears in a position where a
declaration is allowed, then using StaticAssertDecl() is more native
than StaticAssertStmt().
Reviewed-by: John Naylor <[email protected]>
Discussion: https://www.postgresql.org/message-id/flat/941a04e7-dd6f-c0e4-8cdf-a33b3338cbda%40enterprisedb.com
|
|
These zero out the space added by repalloc. This is a common pattern
that is quite hairy to code by hand.
Reviewed-by: Tom Lane <[email protected]>
Discussion: https://www.postgresql.org/message-id/[email protected]
|
|
These don't offer anything over plain Assert, and their usage had
already been declared obsolescent.
Author: Nathan Bossart <[email protected]>
Reviewed-by: Michael Paquier <[email protected]>
Discussion: https://www.postgresql.org/message-id/20221009210148.GA900071@nathanxps13
|
|
Per discussion, the existing routine name able to initialize a SRF
function with materialize mode is unpopular, so rename it. Equally, the
flags of this function are renamed, as of:
- SRF_SINGLE_USE_EXPECTED -> MAT_SRF_USE_EXPECTED_DESC
- SRF_SINGLE_BLESS -> MAT_SRF_BLESS
The previous function and flags introduced in 9e98583 are kept around
for compatibility purposes, so as any extension code already compiled
with v15 continues to work as-is. The declarations introduced here for
compatibility will be removed from HEAD in a follow-up commit.
The new names have been suggested by Andres Freund and Melanie
Plageman.
Discussion: https://postgr.es/m/[email protected]
Backpatch-through: 15
|
|
We lack a version of repalloc() that supports MCXT_ALLOC_NO_OOM
semantics, so invent repalloc_extended() with the usual set of
flags. repalloc_huge() becomes a legacy wrapper for that.
Also, fix dynahash.c so that it can support HASH_ENTER_NULL
requests when using the default palloc-based allocator.
The only reason it didn't do that already was the lack of the
MCXT_ALLOC_NO_OOM option when that code was written, ages ago.
While here, simplify a few overcomplicated tests in mcxt.c.
Discussion: https://postgr.es/m/[email protected]
|
|
Before commit c6e0fe1f2, functions such as AllocSetFree could pretty
safely presume that they were given a valid chunk pointer for their
own type of context, because the indirect call through a memory
context object and method struct would be very unlikely to work
otherwise. But now, if pfree() is mistakenly invoked on a pointer
to garbage, we have three chances in eight of ending up at one of
these functions. That means we need to take extra measures to
verify that we are looking at what we're supposed to be looking at,
especially in debug builds.
Hence, add code to verify that the chunk's back-link to a block header
leads to a memory context object that satisfies the right sort of
IsA() check. This is still a bit weaker than what we did before,
but for the moment assume that an IsA() check is sufficient.
As a compromise between speed and safety, implement these checks
as Asserts when dealing with small chunks but plain test-and-elogs
when dealing with large (external) chunks. The latter case should
not be too performance-critical, but the former case probably is.
In slab.c, all chunks are small; but nonetheless use a plain test
in SlabRealloc, because that is certainly not performance-critical,
indeed we should be suspicious that it's being called in error.
In aset.c, additionally add some assertions that the "value" field
of the chunk header is within the small range allowed for freelist
indexes. Without that, we might find ourselves trying to wipe
most of memory when CLOBBER_FREED_MEMORY is enabled, or scribbling
on a "freelist header" that's far away from the context object.
Eventually, field experience might show us that it's smarter for
these tests to be active always, but for now we'll try to get
away with just having them as assertions.
While at it, also be more uniform about asserting that context
objects passed as parameters are of the type we expect. Some
places missed that altogether, and slab.c was for no very good
reason doing it differently from the other allocators.
Discussion: https://postgr.es/m/[email protected]
|
|
Commit c6e0fe1f2 was a shade too trusting that any pointer passed
to pfree, repalloc, etc will point at a valid chunk. Notably,
passing a pointer that was actually obtained from malloc tended
to result in obscure assertion failures, if not worse. (On FreeBSD
I've seen such mistakes take down the entire cluster, seemingly as
a result of clobbering shared memory.)
To improve matters, extend the mcxt_methods[] array so that it
has entries for every possible MemoryContextMethodID bit-pattern,
with the currently unassigned ID codes pointing to error-reporting
functions. Then, fiddle with the ID assignments so that patterns
likely to be associated with bad pointers aren't valid ID codes.
In particular, we should avoid assigning bit patterns 000 (zeroed
memory) and 111 (wipe_mem'd memory).
It turns out that on glibc (Linux), malloc uses chunk headers that
have flag bits in the same place we keep MemoryContextMethodID,
and that the bit patterns 000, 001, 010 are the only ones we'll
see as long as the backend isn't threaded. So we can have very
robust detection of pfree'ing a malloc-assigned block on that
platform, at least so long as we can refrain from using up those
ID codes. On other platforms, we don't have such a good guarantee,
but keeping 000 reserved will be enough to catch many such cases.
While here, make GetMemoryChunkMethodID() local to mcxt.c, as there
seems no need for it to be exposed even in memutils_internal.h.
Patch by me, with suggestions from Andres Freund and David Rowley.
Discussion: https://postgr.es/m/[email protected]
|
|
MemoryContextContains is no longer reliable in the wake of c6e0fe1f2,
because there's no longer very much redundancy in chunk headers.
(It wasn't *completely* reliable even before that, as there was a
chance of a false positive if you passed it something that didn't
point to an mcxt chunk at all. But it was generally good enough.)
Hence, remove it. There is no remaining core code that requires it.
Extensions that have been using it might be able to substitute a
test like "GetMemoryChunkContext(ptr) == context", recognizing that
this explicitly requires that the pointer point to some chunk.
Tom Lane and David Rowley
Discussion: https://postgr.es/m/[email protected]
|
|
Autoconf is showing its age, fewer and fewer contributors know how to wrangle
it. Recursive make has a lot of hard to resolve dependency issues and slow
incremental rebuilds. Our home-grown MSVC build system is hard to maintain for
developers not using Windows and runs tests serially. While these and other
issues could individually be addressed with incremental improvements, together
they seem best addressed by moving to a more modern build system.
After evaluating different build system choices, we chose to use meson, to a
good degree based on the adoption by other open source projects.
We decided that it's more realistic to commit a relatively early version of
the new build system and mature it in tree.
This commit adds an initial version of a meson based build system. It supports
building postgres on at least AIX, FreeBSD, Linux, macOS, NetBSD, OpenBSD,
Solaris and Windows (however only gcc is supported on aix, solaris). For
Windows/MSVC postgres can now be built with ninja (faster, particularly for
incremental builds) and msbuild (supporting the visual studio GUI, but
building slower).
Several aspects (e.g. Windows rc file generation, PGXS compatibility, LLVM
bitcode generation, documentation adjustments) are done in subsequent commits
requiring further review. Other aspects (e.g. not installing test-only
extensions) are not yet addressed.
When building on Windows with msbuild, builds are slower when using a visual
studio version older than 2019, because those versions do not support
MultiToolTask, required by meson for intra-target parallelism.
The plan is to remove the MSVC specific build system in src/tools/msvc soon
after reaching feature parity. However, we're not planning to remove the
autoconf/make build system in the near future. Likely we're going to keep at
least the parts required for PGXS to keep working around until all supported
versions build with meson.
Some initial help for postgres developers is at
https://wiki.postgresql.org/wiki/Meson
With contributions from Thomas Munro, John Naylor, Stone Tickle and others.
Author: Andres Freund <[email protected]>
Author: Nazir Bilal Yavuz <[email protected]>
Author: Peter Eisentraut <[email protected]>
Reviewed-By: Peter Eisentraut <[email protected]>
Discussion: https://postgr.es/m/[email protected]
|
|
5265e91fd changed MemoryContextContains to update it so that it works
correctly with the new MemoryChunk code added in c6e0fe1f2. However,
5265e91fd was done with the assumption that MemoryContextContains would
only ever be given pointers to memory that had been returned by one of our
MemoryContext allocators. It seems that's not true and many of our 32-bit
buildfarm animals are clearly showing that.
There are some code paths that call MemoryContextContains with a pointer
pointing part way into an allocated chunk. The example of this found by
the 32-bit buildfarm animals is the int2int4_sum() function. This
function returns transdata->sum, which is not a pointer to memory that was
allocated directly. This return value is then subsequently passed to
MemoryContextContains which causes it to crash due to it thinking the
memory directly prior to that pointer is a MemoryChunk. What's actually
in that memory is the field in the struct that comes prior to the "sum"
field. This problem didn't occur in 64-bit world because BIGINT is a
byval type and the code which was calling MemoryContextContains with the
bad pointer only does so with non-byval types.
Here, instead of reverting 5265e91fd and making MemoryContextContains
completely broken again, let's just make it always return false for now.
Effectively prior to 5265e91fd it was doing that anyway, this at least
makes that more explicit. The only repercussions of this with the current
MemoryContextContains calls are that we perform a datumCopy() when we
might not need to. This should make the 32-bit buildfarm animals happy
again and give us more time to consider a long-term fix.
Discussion: https://postgr.es/m/20220907130552.sfjri7jublfxyyi4%40jrouhaud
|
|
c6e0fe1f2 recently changed the way we store headers for allocated chunks
of memory. Prior to that commit, we stored a pointer to the owning
MemoryContext directly prior to the pointer to the allocated memory.
That's no longer true and c6e0fe1f2 neglected to update
MemoryContextContains() so that it correctly obtains the owning context
with the new method.
A side effect of this change and c6e0fe1f2, in general, is that it's even
less safe than it was previously to pass MemoryContextContains() an
arbitrary pointer which was not allocated by one of our MemoryContexts.
Previously some comments in MemoryContextContains() seemed to indicate
that the worst that could happen by passing an arbitrary pointer would be
a false positive return value. It seems to me that this was a rather
wishful outlook as we subsequently proceeded to subtract sizeof(void *)
from the given pointer and then dereferenced that memory. So it seems
quite likely that we could have segfaulted instead of returning a false
positive. However, it's not impossible that the memory sizeof(void *)
bytes before the pointer could have been owned by the process, but it's
far less likely to work now as obtaining a pointer to the owning
MemoryContext is less direct than before c6e0fe1f2 and will access memory
that's possibly much further away to obtain the owning MemoryContext.
Because of this, I took the liberty of updating the comment to warn
against any future usages of the function and checked the existing core
usages to ensure that we only ever pass in a pointer to memory allocated
by a MemoryContext.
Extension authors updating their code for PG16 who are using
MemoryContextContains should check to ensure that only NULL pointers and
pointers to chunks allocated with a MemoryContext will ever be passed to
MemoryContextContains.
Reported-by: Andres Freund
Discussion: https://postgr.es/m/[email protected]
|
|
Traditionally, in MEMORY_CONTEXT_CHECKING builds, we only ever marked a
sentinel byte just beyond the requested size if there happened to be
enough space on the chunk to do so. For Slab and Generation context
types, we only rounded the size of the chunk up to the next maxalign
boundary, so it was often not that likely that those would ever have space
for the sentinel given that the majority of allocation requests are going
to be for sizes which are maxaligned. For AllocSet, it was a little
different as smaller allocations are rounded up to the next power-of-2
value rather than the next maxalign boundary, so we're a bit more likely
to have space for the sentinel byte, especially when we get away from tiny
sized allocations such as 8 or 16 bytes.
Here we make more of an effort to allow space so that there is enough room
for the sentinel byte in more cases. This makes it more likely that we'll
detect when buggy code accidentally writes beyond the end of any of its
memory allocations.
Each of the 3 MemoryContext types has been changed as follows:
The Slab allocator will now always set a sentinel byte. Both the current
usages of this MemoryContext type happen to use chunk sizes which were on
the maxalign boundary, so these never used sentinel bytes previously.
For the Generation allocator, we now always ensure there's enough space in
the allocation for a sentinel byte.
For AllocSet, this commit makes an adjustment for allocation sizes which
are greater than allocChunkLimit. We now ensure there is always space for
a sentinel byte. We don't alter the sentinel behavior for request sizes
<= allocChunkLimit. Making way for the sentinel byte for power-of-2
request sizes would require doubling up to the next power of 2. Some
analysis done on the request sizes made during installcheck shows that a
fairly large portion of allocation requests are for power-of-2 sizes. The
amount of additional memory for the sentinel there seems prohibitive, so
we do nothing for those here.
Author: David Rowley
Discussion: https://postgr.es/m/[email protected]
|
|
c6e0fe1f2 added a new pointer field to SlabBlock to make it 4 bytes larger
on 32-bit machines. Prior to that commit, the size of that struct was a
multiple of 8, which meant that MAXALIGN(sizeof(SlabBlock)) was the same
as sizeof(SlabBlock), however, after c6e0fe1f2, due to the addition of the
new pointer field to store a pointer to the owning context, that was no
longer true on builds with sizeof(void *) == 4.
This problem was highlighted by an Assert failure which was checking that
the pointer given to pfree() was MAXALIGNED. Various 32-bit ARM buildfarm
animals were failing. These have MAXIMUM_ALIGNOF of 8. The only 32-bit
testing I'd managed to do on c6e0fe1f2 had been on x86, which has a
MAXIMUM_ALIGNOF of 4, therefore did not exhibit this issue.
Here we define Slab_BLOCKHDRSZ and copy what is being done in aset.c and
generation.c for doing calculations based on the size of the context's
block type. This means that SlabAlloc() will now always return a
MAXALIGNed pointer.
This also fixes an incorrect sentinel_ok() check in SlabCheck() which was
incorrectly checking the wrong sentinel byte. This must have previously
not caused any issues due to the fullChunkSize never being large enough to
store the sentinel byte.
Diagnosed-by: Tomas Vondra, Tom Lane
Author: Tomas Vondra, David Rowley
Discussion: https://postgr.es/m/CAA4eK1%2B1JyW5TiL%3DyV-3Uq1CrfnTyn0Xrk5uArt31Z%3D8rgPhXQ%40mail.gmail.com
|
|
Whenever we palloc a chunk of memory, traditionally, we prefix the
returned pointer with a pointer to the memory context to which the chunk
belongs. This is required so that we're able to easily determine the
owning context when performing operations such as pfree() and repalloc().
For the AllocSet context, prior to this commit we additionally prefixed
the pointer to the owning context with the size of the chunk. This made
the header 16 bytes in size. This 16-byte overhead was required for all
AllocSet allocations regardless of the allocation size.
For the generation context, the problem was worse; in addition to the
pointer to the owning context and chunk size, we also stored a pointer to
the owning block so that we could track the number of freed chunks on a
block.
The slab allocator had a 16-byte chunk header.
The changes being made here reduce the chunk header size down to just 8
bytes for all 3 of our memory context types. For small to medium sized
allocations, this significantly increases the number of chunks that we can
fit on a given block which results in much more efficient use of memory.
Additionally, this commit completely changes the rule that pointers to
palloc'd memory must be directly prefixed by a pointer to the owning
memory context and instead, we now insist that they're directly prefixed
by an 8-byte value where the least significant 3-bits are set to a value
to indicate which type of memory context the pointer belongs to. Using
those 3 bits as an index (known as MemoryContextMethodID) to a new array
which stores the methods for each memory context type, we're now able to
pass the pointer given to functions such as pfree() and repalloc() to the
function specific to that context implementation to allow them to devise
their own methods of finding the memory context which owns the given
allocated chunk of memory.
The reason we're able to reduce the chunk header down to just 8 bytes is
because of the way we make use of the remaining 61 bits of the required
8-byte chunk header. Here we also implement a general-purpose MemoryChunk
struct which makes use of those 61 remaining bits to allow the storage of
a 30-bit value which the MemoryContext is free to use as it pleases, and
also the number of bytes which must be subtracted from the chunk to get a
reference to the block that the chunk is stored on (also 30 bits). The 1
additional remaining bit is to denote if the chunk is an "external" chunk
or not. External here means that the chunk header does not store the
30-bit value or the block offset. The MemoryContext can use these
external chunks at any time, but must use them if any of the two 30-bit
fields are not large enough for the value(s) that need to be stored in
them. When the chunk is marked as external, it is up to the MemoryContext
to devise its own means to determine the block offset.
Using 3-bits for the MemoryContextMethodID does mean we're limiting
ourselves to only having a maximum of 8 different memory context types.
We could reduce the bit space for the 30-bit value a little to make way
for more than 3 bits, but it seems like it might be better to do that only
if we ever need more than 8 context types. This would only be a problem
if some future memory context type which does not use MemoryChunk really
couldn't give up any of the 61 remaining bits in the chunk header.
With this MemoryChunk, each of our 3 memory context types can quickly
obtain a reference to the block any given chunk is located on. AllocSet
is able to find the context to which the chunk is owned, by first
obtaining a reference to the block by subtracting the block offset as is
stored in the 'hdrmask' field and then referencing the block's 'aset'
field. The Generation context uses the same method, but GenerationBlock
did not have a field pointing back to the owning context, so one is added
by this commit.
In aset.c and generation.c, all allocations larger than allocChunkLimit
are stored on dedicated blocks. When there's just a single chunk on a
block like this, it's easy to find the block from the chunk, we just
subtract the size of the block header from the chunk pointer. The size of
these chunks is also known as we store the endptr on the block, so we can
just subtract the pointer to the allocated memory from that. Because we
can easily find the owning block and the size of the chunk for these
dedicated blocks, we just always use external chunks for allocation sizes
larger than allocChunkLimit. For generation.c, this sidesteps the problem
of non-external MemoryChunks being unable to represent chunk sizes >= 1GB.
This is less of a problem for aset.c as we store the free list index in
the MemoryChunk's spare 30-bit field (the value of which will never be
close to using all 30-bits). We can easily reverse engineer the chunk size
from this when needed. Storing this saves AllocSetFree() from having to
make a call to AllocSetFreeIndex() to determine which free list to put the
newly freed chunk on.
For the slab allocator, this commit adds a new restriction that slab
chunks cannot be >= 1GB in size. If there happened to be any users of
slab.c which used chunk sizes this large, they really should be using
AllocSet instead.
Here we also add a restriction that normal non-dedicated blocks cannot be
1GB or larger. It's now not possible to pass a 'maxBlockSize' >= 1GB
during the creation of an AllocSet or Generation context. Allocations can
still be larger than 1GB, it's just these will always be on dedicated
blocks (which do not have the 1GB restriction).
Author: Andres Freund, David Rowley
Discussion: https://postgr.es/m/CAApHDvpjauCRXcgcaL6+e3eqecEHoeRm9D-kcbuvBitgPnW=vw@mail.gmail.com
|
|
Per discussion, we choose not to change this. This just gives a
little bit more information.
Discussion: https://www.postgresql.org/message-id/flat/cf26e970-8e92-59f1-247a-aa265235075b%40enterprisedb.com
|
|
Commit 4f658dc8 provided the traditional BSD fls() function in
src/port/fls.c so it could be used in several places. Later we added a
bunch of similar facilities in pg_bitutils.h, based on compiler
builtins that map to hardware instructions. It's a bit confusing to
have both 1-based and 0-based variants of this operation in use in
different parts of the tree, and neither is blessed by a standard.
Let's drop fls.c and the configure probe, and reuse the newer code.
Reviewed-by: David Rowley <[email protected]>
Reviewed-by: Tom Lane <[email protected]>
Discussion: https://postgr.es/m/CA%2BhUKG%2B7dSX1XF8yFGmYk-%3D48dbjH2kmzZj16XvhbrWP-9BzRg%40mail.gmail.com
|
|
This replaces all MemSet() calls with struct initialization where that
is easily and obviously possible. (For example, some cases have to
worry about padding bits, so I left those.)
(The same could be done with appropriate memset() calls, but this
patch is part of an effort to phase out MemSet(), so it doesn't touch
memset() calls.)
Reviewed-by: Ranier Vilela <[email protected]>
Reviewed-by: Alvaro Herrera <[email protected]>
Discussion: https://www.postgresql.org/message-id/[email protected]
|
|
Previously, we encoded both NULL and the first byte at the base address
as 0. That confusion led to the assertion in commit e07d4ddc, which
failed when min_dynamic_shared_memory was used. Give them distinct
encodings, by switching to 1-based offsets for non-NULL pointers. Also
improve macro hygiene in passing (missing/misplaced parentheses), and
remove open-coded access to the raw offset value from freepage.c/h.
Although e07d4ddc was back-patched to 10, the only code that actually
makes use of relptr at the base address arrived in 84b1c63a, so no need
to back-patch further than 14 for now.
Reported-by: Justin Pryzby <[email protected]>
Reviewed-by: Robert Haas <[email protected]>
Discussion: https://postgr.es/m/20220519193839.GT19626%40telsasoft.com
|
|
These are useless and distracting. We wouldn't have written the code
with them to begin with, so there's no reason to keep them.
Author: Justin Pryzby <[email protected]>
Discussion: https://postgr.es/m/[email protected]
Discussion: https://postgr.es/m/attachment/133167/0016-Extraneous-blank-lines.patch
|
|
Author: Justin Pryzby
Discussion: https://postgr.es/m/[email protected]
|
|
Here we make a series of improvements to the generation memory
allocator, namely:
1. Allow generation contexts to have a minimum, initial and maximum block
sizes. The standard allocator allows this already but when the generation
context was added, it only allowed fixed-sized blocks. The problem with
fixed-sized blocks is that it's difficult to choose how large to make the
blocks. If the chosen size is too small then we'd end up with a large
number of blocks and a large number of malloc calls. If the block size is
made too large, then memory is wasted.
2. Add support for "keeper" blocks. This is a special block that is
allocated along with the context itself but is never freed. Instead,
when the last chunk in the keeper block is freed, we simply mark the block
as empty to allow new allocations to make use of it.
3. Add facility to "recycle" newly empty blocks instead of freeing them
and having to later malloc an entire new block again. We do this by
recording a single GenerationBlock which has become empty of any chunks.
When we run out of space in the current block, we check to see if there is
a "freeblock" and use that if it contains enough space for the allocation.
Author: David Rowley, Tomas Vondra
Reviewed-by: Andy Fan
Discussion: https://postgr.es/m/[email protected]
|
|
Set-returning functions that use the Materialize mode, creating a
tuplestore to include all the tuples returned in a set rather than doing
so in multiple calls, use roughly the same set of steps to prepare
ReturnSetInfo for this job:
- Check if ReturnSetInfo supports returning a tuplestore and if the
materialize mode is enabled.
- Create a tuplestore for all the tuples part of the returned set in the
per-query memory context, stored in ReturnSetInfo->setResult.
- Build a tuple descriptor mostly from get_call_result_type(), then
stored in ReturnSetInfo->setDesc. Note that there are some cases where
the SRF's tuple descriptor has to be the one specified by the function
caller.
This refactoring is done so as there are (well, should be) no behavior
changes in any of the in-core functions refactored, and the centralized
function that checks and sets up the function's ReturnSetInfo can be
controlled with a set of bits32 options. Two of them prove to be
necessary now:
- SRF_SINGLE_USE_EXPECTED to use expectedDesc as tuple descriptor, as
expected by the function's caller.
- SRF_SINGLE_BLESS to validate the tuple descriptor for the SRF.
The same initialization pattern is simplified in 28 places per my
count as of src/backend/, shaving up to ~900 lines of code. These
mostly come from the removal of the per-query initializations and the
sanity checks now grouped in a single location. There are more
locations that could be simplified in contrib/, that are left for a
follow-up cleanup.
fcc2817, 07daca5 and d61a361 have prepared the areas of the code related
to this change, to ease this refactoring.
Author: Melanie Plageman, Michael Paquier
Reviewed-by: Álvaro Herrera, Justin Pryzby
Discussion: https://postgr.es/m/CAAKRu_azyd1Z3W_r7Ou4sorTjRCs+PxeHw1CWJeXKofkE6TuZg@mail.gmail.com
|
|
SPI_commit previously left it up to the caller to recover from any error
occurring during commit. Since that's complicated and requires use of
low-level xact.c facilities, it's not too surprising that no caller got
it right. Let's move the responsibility for cleanup into spi.c. Doing
that requires redefining SPI_commit as starting a new transaction, so
that it becomes equivalent to SPI_commit_and_chain except that you get
default transaction characteristics instead of preserving the prior
transaction's characteristics. We can make this pretty transparent
API-wise by redefining SPI_start_transaction() as a no-op. Callers
that expect to do something in between might be surprised, but
available evidence is that no callers do so.
Having made that API redefinition, we can fix this mess by having
SPI_commit[_and_chain] trap errors and start a new, clean transaction
before re-throwing the error. Likewise for SPI_rollback[_and_chain].
Some cleanup is also needed in AtEOXact_SPI, which was nowhere near
smart enough to deal with SPI contexts nested inside a committing
context.
While plperl and pltcl need no changes beyond removing their now-useless
SPI_start_transaction() calls, plpython needs some more work because it
hadn't gotten the memo about catching commit/rollback errors in the
first place. Such an error resulted in longjmp'ing out of the Python
interpreter, which leaks Python stack entries at present and is reported
to crash Python 3.11 altogether. Add the missing logic to catch such
errors and convert them into Python exceptions.
We are probably going to have to back-patch this once Python 3.11 ships,
but it's a sufficiently basic change that I'm a bit nervous about doing
so immediately. Let's let it bake awhile in HEAD first.
Peter Eisentraut and Tom Lane
Discussion: https://postgr.es/m/[email protected]
Discussion: https://postgr.es/m/[email protected]
|
|
The following set-returning functions have their logic simplified, to be
more consistent with other in-core areas:
- pg_prepared_statement()'s tuple descriptor is now created with
get_call_result_type() instead of being created from scratch, saving
from some duplication with pg_proc.dat.
- show_all_file_settings(), similarly, now uses get_call_result_type()
to build its tuple descriptor instead of creating it from scratch.
- pg_options_to_table() made use of a static routine called only once.
This commit removes this internal routine to make the function easier to
follow.
- pg_config() was using a unique logic style, doing checks on the tuple
descriptor passed down in expectedDesc, but it has no need to do so.
This switches the function to use a tuplestore with a tuple descriptor
retrieved from get_call_result_type(), instead.
This simplifies an upcoming patch aimed at refactoring the way
tuplestores are created and checked in set-returning functions, this
change making sense as its own independent cleanup by shaving some
code.
Author: Melanie Plageman, Michael Paquier
Reviewed-by: Justin Pryzby
Discussion: https://postgr.es/m/CAAKRu_azyd1Z3W_r7Ou4sorTjRCs+PxeHw1CWJeXKofkE6TuZg@mail.gmail.com
|
|
This routine is a no-op since dd04e95 from 2003, with a macro kept
around for compatibility purposes. This has led to the same code
patterns being copy-pasted around for no effect, sometimes in confusing
ways like in pg_logical_slot_get_changes_guts() from logical.c where the
code was actually incorrect.
This issue has been discussed on two different threads recently, so
rather than living with this legacy, remove any uses of this routine in
the C code to simplify things. The compatibility macro is kept to avoid
breaking any out-of-core modules that depend on it.
Reported-by: Tatsuhito Kasahara, Justin Pryzby
Author: Tatsuhito Kasahara
Discussion: https://postgr.es/m/[email protected]
Discussion: https://postgr.es/m/CAP0=ZVJeeYfAeRfmzqAF2Lumdiv4S4FewyBnZd4DPTrsSQKJKw@mail.gmail.com
|
|
When pg_log_backend_memory_contexts() is executed, the target backend
should use LOG_SERVER_ONLY to log its memory contexts, to prevent them
from being sent to its connected client regardless of client_min_messages.
But previously the backend unexpectedly used LOG to log the message
"logging memory contexts of PID %d" and it could be sent to the client.
This is a bug in memory context logging.
To fix the bug, this commit changes that message so that it's logged with
LOG_SERVER_ONLY.
Back-patch to v14 where pg_log_backend_memory_contexts() was added.
Author: Fujii Masao
Reviewed-by: Bharath Rupireddy, Atsushi Torikoshi
Discussion: https://postgr.es/m/[email protected]
|
|
Backpatch-through: 10
|
|
|
|
Commit 84f5c2908 forgot to consider the possibility that
EnsurePortalSnapshotExists could run inside a subtransaction with
lifespan shorter than the Portal's. In that case, the new active
snapshot would be popped at the end of the subtransaction, leaving
a dangling pointer in the Portal, with mayhem ensuing.
To fix, make sure the ActiveSnapshot stack entry is marked with
the same subtransaction nesting level as the associated Portal.
It's certainly safe to do so since we won't be here at all unless
the stack is empty; hence we can't create an out-of-order stack.
Let's also apply this logic in the case where PortalRunUtility
sets portalSnapshot, just to be sure that path can't cause similar
problems. It's slightly less clear that that path can't create
an out-of-order stack, so add an assertion guarding it.
Report and patch by Bertrand Drouvot (with kibitzing by me).
Back-patch to v11, like the previous commit.
Discussion: https://postgr.es/m/[email protected]
|
|
COMMIT/ROLLBACK necessarily destroys all snapshots within the session.
The original implementation of intra-procedure transactions just
cavalierly did that, ignoring the fact that this left us executing in
a rather different environment than normal. In particular, it turns
out that handling of toasted datums depends rather critically on there
being an outer ActiveSnapshot: otherwise, when SPI or the core
executor pop whatever snapshot they used and return, it's unsafe to
dereference any toasted datums that may appear in the query result.
It's possible to demonstrate "no known snapshots" and "missing chunk
number N for toast value" errors as a result of this oversight.
Historically this outer snapshot has been held by the Portal code,
and that seems like a good plan to preserve. So add infrastructure
to pquery.c to allow re-establishing the Portal-owned snapshot if it's
not there anymore, and add enough bookkeeping support that we can tell
whether it is or not.
We can't, however, just re-establish the Portal snapshot as part of
COMMIT/ROLLBACK. As in normal transaction start, acquiring the first
snapshot should wait until after SET and LOCK commands. Hence, teach
spi.c about doing this at the right time. (Note that this patch
doesn't fix the problem for any PLs that try to run intra-procedure
transactions without using SPI to execute SQL commands.)
This makes SPI's no_snapshots parameter rather a misnomer, so in HEAD,
rename that to allow_nonatomic.
replication/logical/worker.c also needs some fixes, because it wasn't
careful to hold a snapshot open around AFTER trigger execution.
That code doesn't use a Portal, which I suspect someday we're gonna
have to fix. But for now, just rearrange the order of operations.
This includes back-patching the recent addition of finish_estate()
to centralize the cleanup logic there.
This also back-patches commit 2ecfeda3e into v13, to improve the
test coverage for worker.c (it was that test that exposed that
worker.c's snapshot management is wrong).
Per bug #15990 from Andreas Wicht. Back-patch to v11 where
intra-procedure COMMIT was added.
Discussion: https://postgr.es/m/[email protected]
|
|
Commit 3e98c0bafb added pg_backend_memory_contexts view to display
the memory contexts of the backend process. However its target process
is limited to the backend that is accessing to the view. So this is
not so convenient when investigating the local memory bloat of other
backend process. To improve this situation, this commit adds
pg_log_backend_memory_contexts() function that requests to log
the memory contexts of the specified backend process.
This information can be also collected by calling
MemoryContextStats(TopMemoryContext) via a debugger. But
this technique cannot be used in some environments because no debugger
is available there. So, pg_log_backend_memory_contexts() allows us to
see the memory contexts of specified backend more easily.
Only superusers are allowed to request to log the memory contexts
because allowing any users to issue this request at an unbounded rate
would cause lots of log messages and which can lead to denial of service.
On receipt of the request, at the next CHECK_FOR_INTERRUPTS(),
the target backend logs its memory contexts at LOG_SERVER_ONLY level,
so that these memory contexts will appear in the server log but not
be sent to the client. It logs one message per memory context.
Because if it buffers all memory contexts into StringInfo to log them
as one message, which may require the buffer to be enlarged very much
and lead to OOM error since there can be a large number of memory
contexts in a backend.
When a backend process is consuming huge memory, logging all its
memory contexts might overrun available disk space. To prevent this,
now this patch limits the number of child contexts to log per parent
to 100. As with MemoryContextStats(), it supposes that practical cases
where the log gets long will typically be huge numbers of siblings
under the same parent context; while the additional debugging value
from seeing details about individual siblings beyond 100 will not be large.
There was another proposed patch to add the function to return
the memory contexts of specified backend as the result sets,
instead of logging them, in the discussion. However that patch is
not included in this commit because it had several issues to address.
Thanks to Tatsuhito Kasahara, Andres Freund, Tom Lane, Tomas Vondra,
Michael Paquier, Kyotaro Horiguchi and Zhihong Yu for the discussion.
Bump catalog version.
Author: Atsushi Torikoshi
Reviewed-by: Kyotaro Horiguchi, Zhihong Yu, Fujii Masao
Discussion: https://postgr.es/m/[email protected]
|
|
This patch essentially is cleaning up technical debt left behind
by the original implementation of plpgsql procedures, particularly
commit d92bc83c4. That patch (or more precisely, follow-on patches
fixing its worst bugs) forced us to re-plan CALL and DO statements
each time through, if we're in a non-atomic context. That wasn't
for any fundamental reason, but just because use of a saved plan
requires having a ResourceOwner to hold a reference count for the
plan, and we had no suitable resowner at hand, nor would the
available APIs support using one if we did. While it's not that
expensive to create a "plan" for CALL/DO, the cycles do add up
in repeated executions.
This patch therefore makes the following API changes:
* GetCachedPlan/ReleaseCachedPlan are modified to let the caller
specify which resowner to use to pin the plan, rather than forcing
use of CurrentResourceOwner.
* spi.c gains a "SPI_execute_plan_extended" entry point that lets
callers say which resowner to use to pin the plan. This borrows the
idea of an options struct from the recently added SPI_prepare_extended,
hopefully allowing future options to be added without more API breaks.
This supersedes SPI_execute_plan_with_paramlist (which I've marked
deprecated) as well as SPI_execute_plan_with_receiver (which is new
in v14, so I just took it out altogether).
* I also took the opportunity to remove the crude hack of letting
plpgsql reach into SPI private data structures to mark SPI plans as
"no_snapshot". It's better to treat that as an option of
SPI_prepare_extended.
Now, when running a non-atomic procedure or DO block that contains
any CALL or DO commands, plpgsql creates a ResourceOwner that
will be used to pin the plans of the CALL/DO commands. (In an
atomic context, we just use CurrentResourceOwner, as before.)
Having done this, we can just save CALL/DO plans normally,
whether or not they are used across transaction boundaries.
This seems to be good for something like 2X speedup of a CALL
of a trivial procedure with a few simple argument expressions.
By restricting the creation of an extra ResourceOwner like this,
there's essentially zero penalty in cases that can't benefit.
Pavel Stehule, with some further hacking by me
Discussion: https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com
|
|
Backpatch-through: 9.5
|
|
Invent a new flag bit HASH_STRINGS to specify C-string hashing, which
was formerly the default; and add assertions insisting that exactly
one of the bits HASH_STRINGS, HASH_BLOBS, and HASH_FUNCTION be set.
This is in hopes of preventing recurrences of the type of oversight
fixed in commit a1b8aa1e4 (i.e., mistakenly omitting HASH_BLOBS).
Also, when HASH_STRINGS is specified, insist that the keysize be
more than 8 bytes. This is a heuristic, but it should catch
accidental use of HASH_STRINGS for integer or pointer keys.
(Nearly all existing use-cases set the keysize to NAMEDATALEN or
more, so there's little reason to think this restriction should
be problematic.)
Tweak hash_create() to insist that the HASH_ELEM flag be set, and
remove the defaults it had for keysize and entrysize. Since those
defaults were undocumented and basically useless, no callers
omitted HASH_ELEM anyway.
Also, remove memset's zeroing the HASHCTL parameter struct from
those callers that had one. This has never been really necessary,
and while it wasn't a bad coding convention it was confusing that
some callers did it and some did not. We might as well save a few
cycles by standardizing on "not".
Also improve the documentation for hash_create().
In passing, improve reinit.c's usage of a hash table by storing
the key as a binary Oid rather than a string; and, since that's
a temporary hash table, allocate it in CurrentMemoryContext for
neatness.
Discussion: https://postgr.es/m/[email protected]
|
|
Otherwise just printing an empty string makes the memory context debug
output slightly confusing.
Discussion: https://www.postgresql.org/message-id/flat/ccb353ef-89ff-09b3-8046-1d2514624b9c%402ndquadrant.com
|
|
Reported-by: Ranier Vilela
Discussion: https://postgr.es/m/CAEudQAo1+AcGppxDSg8k+zF4+Kv+eJyqzEDdbpDg58-=MQcerQ@mail.gmail.com
Author: Ranier Vilela
Backpatch-through: master
|
|
Previously the codes for pg_backend_memory_contexts were in
src/backend/utils/mmgr/mcxt.c. This commit moves them to
src/backend/utils/adt/mcxtfuncs.c so that mcxt.c basically includes
only the low-level interface for memory contexts.
Author: Atsushi Torikoshi
Reviewed-by: Michael Paquier, Fujii Masao
Discussion: https://postgr.es/m/[email protected]
|
|
This view displays the usages of all the memory contexts of the server
process attached to the current session. This information is useful to
investigate the cause of backend-local memory bloat.
This information can be also collected by calling
MemoryContextStats(TopMemoryContext) via a debugger. But this technique
cannot be uesd in some environments because no debugger is available there.
And it outputs lots of text messages and it's not easy to analyze them.
So, pg_backend_memory_contexts view allows us to access to backend-local
memory contexts information more easily.
Bump catalog version.
Author: Atsushi Torikoshi, Fujii Masao
Reviewed-by: Tatsuhito Kasahara, Andres Freund, Daniel Gustafsson, Robert Haas, Michael Paquier
Discussion: https://postgr.es/m/[email protected]
|
