pedroediaz/lexvm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

pike: improve size calculations

Browse Source
This commit is contained in:
Kyryl Melekhin
2022-02-17 18:20:06 +00:00
parent 281820d7c9
commit eb01f29134
2 changed files with 131 additions and 144 deletions
Download Patch File Download Diff File Expand all files Collapse all files

138
README
View File

@@ -86,89 +86,71 @@ length of input, m it the size of RE and t is the number of submatch groups
and subexpressions that contain them."
Research has shown that it is possible to disambiguate NFA in polynomial time
but it brings serious performance issues on non ambiguous inputs.
See the branch "disambiguate_paths" on this repo shows what is being
done to solve it and the potential performance costs. In short it
requires tracking the parent of every state added on nlist from clist.
If the state from nlist matches the consumer, the alternative clist
state related to that nlist state gets discarded and the nsub ref
can be decremented (freed). The reason why this problem does not
exist for non ambiguous regexes is because the alternative clist
state will never match due to the next state having a different
consumer. There is no need for any extra handling it gets freed normally.
I decided to not apply this solution here because I think
most use cases for regex are not ambiguious like say regex:
"a{10000}". If you try matching 10000 'a' characters in a row
like that you will have a problem where the stack usage will
jump up to 10000*(subsize) but it will never exceed the size
of regex though, but the number of NFA states will also increase
by the same amount, so at the charater 9999 you will find
9999 redundant nlist states, that will degrade performance
linearly, however it will be very slow compared to uplimited
regex like a+. The cost of this solution is somewhere around
2% general performance decrease (broadly), but a magnitude of
complexity decrease for ambiguous cases, for example
matching 64 characters went down from 30 to 9 microseconds.
Another solution to this problem can be to determine the
ambiguous paths at compile time and flag the inner
states as ambiguous ahead of time, still this can't avoid
having a loop though the alt states as their positioning
in clist can't be precomputed due to the dynamic changes.
but it brings serious performance issues on non ambiguous inputs. See the
branch "disambiguate_paths" on this repo shows what is being done to solve it
and the potential performance costs. In short it requires tracking the parent
of every state added on nlist from clist. If the state from nlist matches
the consumer, the alternative clist state related to that nlist state gets
discarded and the nsub ref can be decremented (freed). The reason why this
problem does not exist for non ambiguous regexes is because the alternative
clist state will never match due to the next state having a different consumer
. There is no need for any extra handling it gets freed normally. I decided
to not apply this solution here because I think most use cases for regex are
not ambiguious like say regex: "a{10000}". If you try matching 10000 'a'
characters in a row like that you will have a problem where the stack usage
will jump up to 10000*(subsize) but it will never exceed the size of regex
though, but the number of NFA states will also increase by the same amount,
so at the charater 9999 you will find 9999 redundant nlist states, that will
degrade performance linearly, however it will be very slow compared to
uplimited regex like a+. The cost of this solution is somewhere around 2%
general performance decrease (broadly), but a magnitude of complexity
decrease for ambiguous cases, for example matching 64 characters went down
from 30 to 9 microseconds. Another solution to this problem can be to
determine the ambiguous paths at compile time and flag the inner states as
ambiguous ahead of time, still this can't avoid having a loop though the alt
states as their positioning in clist can't be precomputed due to the dynamic
changes.
(Comment about O(mt) memory complexity)
This worst case scenario can only happen on ambiguous input, that is why nsubs
size is set to half a MB just in case, this can match 5000000
ambiguous consumers (char, class, any) assuming t is 1. In practice there
is almost never a situation where someone wants to search using regex this
large. Use of alloca() instead of VLA, could remove this limit, I just wish
it was standardized. If you ever wondered about a situation where alloca
is a must, this is the algorithm.
Most of the time memory usage is very low and the space
complexity for non ambigious regex is O(nt) where n is
the number of currently considering alternate paths
in the regex and t is the number of submatch groups.
This worst case scenario can only happen on ambiguous input. Ambiguous
consumers (char, class, any) assuming t is 1. In practice there is almost
never a situation where someone wants to search using regex this large. Most
of the time memory usage is very low and the space complexity for non
ambigious regex is O(nt) where n is the number of currently considering
alternate paths in the regex and t is the number of submatch groups.
This pikevm features an improved submatch extraction
algorithm based on Russ Cox's original design.
I - Kyryl Melekhin have found a way to optimize the tracking
properly of 1st number in the submatch pair. Based on simple
observation of how the NFA is constructed I noticed that
there is no way for addthread1() to ever reach inner SAVE
instructions in the regex, so that leaves tracking 2nd pairs by
addthread1() irrelevant to the final results (except the need to
initialize the sub after allocation). This improved the overall
performance by 25% which is massive considering that at the
time there was nothing else left to can be done to make it faster.
This pikevm implementation features an improved submatch extraction algorithm
based on Russ Cox's original design. I - Kyryl Melekhin have found a way to
optimize the tracking properly of 1st number in the submatch pair. Based on
simple observation of how the NFA is constructed I noticed that there is no
way for addthread1() to ever reach inner SAVE instructions in the regex, so
that leaves tracking 2nd pairs by addthread1() irrelevant to the final
results (except the need to initialize the sub after allocation). This
improved the overall performance by 25% which is massive considering that at
the time there was nothing else left to can be done to make it faster.
What are on##list macros?
Redundant state inside nlist can happen in couple of
ways, and has to do with the (closure) a* (star) operations and
also +. Due to the automata machine design split happens
to be above the next consumed instruction and if that
state gets added onto the list we may segfault or give
wrong submatch result. Rsplit does not have this problem
because it is generated below the consumer instruction, but
it can still add redundant states. Overall this is extremely
difficult to understand or explain, but this is just something
we have to check for. We checked for this using extra int inside
the split instructions, so this left some global state inside the
machine insts. Most of the time we just added to the next
gen number and kept incrementing it forever. This leaves a small
chance of overflowing the int and getting a run on a false state
left from previous use of the regex. Though if overflow never
happens there is no chance of getting a false state. Overflows
like this pose a high security threat, if the hacker knows
how many cycles he needs to overflow the gen variable and get
inconsistent result. It is possible to reset the marks if we
near the overflow, but as you may guess that does not come
for free.
Redundant state inside nlist can happen in couple of ways, and has to do with
the (closure) a* (star) operations and also +. Due to the automata machine
design split happens to be above the next consumed instruction and if that
state gets added onto the list we may segfault or give wrong submatch result.
Rsplit does not have this problem because it is generated below the consumer
instruction, but it can still add redundant states. Overall this is extremely
difficult to understand or explain, but this is just something we have to
check for. We checked for this using extra int inside the split instructions,
so this left some global state inside the machine insts. Most of the time we
just added to the next gen number and kept incrementing it forever. This
leaves a small chance of overflowing the int and getting a run on a false
state left from previous use of the regex. Though if overflow never happens
there is no chance of getting a false state. Overflows like this pose a high
security threat, if the hacker knows how many cycles he needs to overflow the
gen variable and get inconsistent result. It is possible to reset the marks
if we near the overflow, but as you may guess that does not come for free.
Currently I removed all dynamic global state from the instructions
fixing any overlow issue utilizing a sparse set datastructure trick
which abuses the uninitialized varibles. This allows the redundant
states to be excluded in O(1) operation. That said, don't run
valgrind on pikevm as it will go crazy, or find a way to surpress
errors from pikevm.
Currently I removed all dynamic global state from the instructions fixing any
overlow issue utilizing a sparse set datastructure trick which abuses the
uninitialized varibles. This allows the redundant states to be excluded in
O(1) operation. That said, don't run valgrind on pikevm as it will go crazy, or
find a way to surpress errors from pikevm.
Further reading
===============

77
pike.c
View File

@@ -1,5 +1,8 @@
// Copyright 2007-2009 Russ Cox. All Rights Reserved.
// Use of this source code is governed by a BSD-style
/*
Copyright 2007-2009 Russ Cox. All Rights Reserved.
Copyright 2020-2021 Kyryl Melekhin. All Rights Reserved.
Use of this source code is governed by a BSD-style
*/
#include <stdio.h>
#include <stdlib.h>
@@ -51,35 +54,36 @@ static int isword(const char *s)
typedef struct rcode rcode;
struct rcode
{
int unilen;
int len;
int sub;
int presub;
int splits;
int insts[];
int unilen; /* number of integers in insts */
int len; /* number of atoms/instructions */
int sub; /* interim val = save count; final val = nsubs size */
int presub; /* interim val = save count; final val = 1 rsub size */
int splits; /* number of split insts */
int sparsesz; /* sdense size */
int insts[]; /* re code */
};
enum
{
// Instructions which consume input bytes (and thus fail if none left)
/* Instructions which consume input bytes */
CHAR = 1,
CLASS,
MATCH,
ANY,
// Assert position
/* Assert position */
WBEG,
WEND,
BOL,
EOL,
// Other (special) instructions
/* Other (special) instructions */
SAVE,
// Instructions which take relative offset as arg
/* Instructions which take relative offset as arg */
JMP,
SPLIT,
RSPLIT,
};
// Return codes for re_sizecode() and re_comp()
/* Return codes for re_sizecode() and re_comp() */
enum {
RE_SUCCESS = 0,
RE_SYNTAX_ERROR = -2,
@@ -111,7 +115,7 @@ pc += num;
static int re_classmatch(const int *pc, int c)
{
// pc points to "classnot" byte after opcode
/* pc points to "classnot" byte after opcode */
int is_positive = *pc++;
int cnt = *pc++;
while (cnt--) {
@@ -176,7 +180,7 @@ void re_dumpcode(rcode *prog)
break;
}
}
printf("Unilen: %d, insts: %d, splits: %d, counted insts: %d\n",
printf("unilen: %d, insts: %d, splits: %d, counted insts: %d\n",
prog->unilen, prog->len, prog->splits, i);
}
@@ -196,7 +200,7 @@ static int _compilecode(const char **re_loc, rcode *prog, int sizecode)
switch (*re) {
case '\\':
re++;
if (!*re) goto syntax_error; // Trailing backslash
if (!*re) goto syntax_error; /* Trailing backslash */
if (*re == '<' || *re == '>') {
if (re - *re_loc > 2 && re[-2] == '\\')
break;
@@ -223,7 +227,7 @@ static int _compilecode(const char **re_loc, rcode *prog, int sizecode)
re++;
} else
EMIT(PC++, 1);
PC++; // Skip "# of pairs" byte
PC++; /* Skip "# of pairs" byte */
for (cnt = 0; *re != ']'; cnt++) {
if (*re == '\\') re++;
if (!*re) goto syntax_error;
@@ -372,9 +376,8 @@ int re_sizecode(const char *re, int *nsub)
dummyprog.unilen = 3;
dummyprog.sub = 0;
int res = _compilecode(&re, &dummyprog, /*sizecode*/1);
int res = _compilecode(&re, &dummyprog, 1);
if (res < 0) return res;
// If unparsed chars left
if (*re) return RE_SYNTAX_ERROR;
*nsub = dummyprog.sub;
return dummyprog.unilen;
@@ -388,9 +391,8 @@ int re_comp(rcode *prog, const char *re, int nsubs)
prog->presub = nsubs;
prog->splits = 0;
int res = _compilecode(&re, prog, /*sizecode*/0);
int res = _compilecode(&re, prog, 0);
if (res < 0) return res;
// If unparsed chars left
if (*re) return RE_SYNTAX_ERROR;
int icnt = 0, scnt = SPLIT;
for (int i = 0; i < prog->unilen; i++)
@@ -417,8 +419,11 @@ int re_comp(rcode *prog, const char *re, int nsubs)
prog->insts[prog->unilen++] = SAVE;
prog->insts[prog->unilen++] = prog->sub + 1;
prog->insts[prog->unilen++] = MATCH;
prog->splits = (scnt - SPLIT) / 2 + SPLIT;
prog->splits = (scnt - SPLIT) / 2;
prog->len = icnt + 2;
prog->presub = sizeof(rsub)+(sizeof(char*) * (nsubs + 1) * 2);
prog->sub = prog->presub * (prog->len - prog->splits + 4);
prog->sparsesz = (scnt - 2) * 2;
return RE_SUCCESS;
}
@@ -434,8 +439,14 @@ if (--csub->ref == 0) { \
freesub = csub; \
} \
#define deccheck(nn) \
{ decref(nsub) goto rec_check##nn; } \
#define rec_check(nn) \
if (si) { \
npc = pcs[--si]; \
nsub = subs[si]; \
goto rec##nn; \
} \
#define deccheck(nn) { decref(nsub) rec_check(nn) continue; } \
#define onclist(nn)
#define onnlist(nn) \
@@ -493,19 +504,13 @@ if (spc == MATCH) \
} \
#define addthread(nn, list, listidx) \
si = 0; \
rec##nn: \
spc = *npc; \
if ((unsigned int)spc < WBEG) { \
list[listidx].sub = nsub; \
list[listidx++].pc = npc; \
rec_check(nn) \
list##match() \
rec_check##nn: \
if (si) { \
npc = pcs[--si]; \
nsub = subs[si]; \
goto rec##nn; \
} \
continue; \
} \
next##nn: \
@@ -557,18 +562,18 @@ clistidx = nlistidx; \
int re_pikevm(rcode *prog, const char *s, const char **subp, int nsubp)
{
int rsubsize = sizeof(rsub)+(sizeof(char*)*nsubp);
int si, i, j, c, suboff = rsubsize, *npc, osubp = nsubp * sizeof(char*);
int clistidx = 0, nlistidx, spc, mcont = MATCH;
int rsubsize = prog->presub, suboff = rsubsize;
int spc, i, j, c, *npc, osubp = nsubp * sizeof(char*);
int si = 0, clistidx = 0, nlistidx, mcont = MATCH;
const char *sp = s, *_sp = s;
int *insts = prog->insts;
int *pcs[prog->splits];
unsigned int sdense[prog->splits * 2], sparsesz;
rsub *subs[prog->splits];
char nsubs[rsubsize * (prog->len-prog->splits+14)];
unsigned int sdense[prog->sparsesz], sparsesz;
rsub *nsub, *s1, *matched = NULL, *freesub = NULL;
rthread _clist[prog->len], _nlist[prog->len];
rthread *clist = _clist, *nlist = _nlist, *tmp;
char nsubs[prog->sub];
goto jmp_start;
for (;; sp = _sp) {
uc_len(i, sp) uc_code(c, sp)