/[pcre]/code/trunk/doc/pcrejit.3

Diff of /code/trunk/doc/pcrejit.3

Parent Directory | Revision Log | View Patch Patch

-revision 678 by ph10,
Sun Aug 28 15:23:03 2011 UTC
+revision 737 by ph10,
Wed Oct 19 17:37:29 2011 UTC
 Line 5 
 PCRE - Perl-compatible regular expressio
  .rs
  .sp
  Just-in-time compiling is a heavyweight optimization that can greatly speed up
  pattern matching. However, it comes at the cost of extra processing before the
  match is performed. Therefore, it is of most benefit when the same pattern is
  going to be matched many times. This does not necessarily mean many calls of
  \fPpcre_exec()\fP; if the pattern is not anchored, matching attempts may take
  place many times at various positions in the subject, even for a single call to
  \fBpcre_exec()\fP. If the subject string is very long, it may still pay to use
  JIT for one-off matches.
  .P
  JIT support applies only to the traditional matching function,
  \fBpcre_exec()\fP. It does not apply when \fBpcre_dfa_exec()\fP is being used.
  The code for this support was written by Zoltan Herczeg.
  .
 Line 26 
 JIT support is an optional feature of PC
  JIT. The support is limited to the following hardware platforms:
  .sp
    ARM v5, v7, and Thumb2
-   MIPS 32-bit
-   Power PC 32-bit and 64-bit
    Intel x86 32-bit and 64-bit
- .sp
+   MIPS 32-bit
- If --enable-jit is set on an unsupported platform, compilation fails.
+   Power PC 32-bit and 64-bit (experimental)
+ .sp
+ The Power PC support is designated as experimental because it has not been
+ fully tested. If --enable-jit is set on an unsupported platform, compilation
+ fails.
  .P
  A program can tell if JIT support is available by calling \fBpcre_config()\fP
  with the PCRE_CONFIG_JIT option. The result is 1 when JIT is available, and 0
  otherwise. However, a simple program does not need to check this in order to
  use JIT. The API is implemented in a way that falls back to the ordinary PCRE
  code if JIT is not available.
-Line 47 
 You have to do two things to make use of
+Line 49 
 You have to do two things to make use of
    (1) Call \fBpcre_study()\fP with the PCRE_STUDY_JIT_COMPILE option for
        each compiled pattern, and pass the resulting \fBpcre_extra\fP block to
        \fBpcre_exec()\fP.
+ .sp
    (2) Use \fBpcre_free_study()\fP to free the \fBpcre_extra\fP block when it is
-       no longer needed instead of just freeing it yourself. This ensures that
+       no longer needed instead of just freeing it yourself. This
-       any JIT data is also freed.
+       ensures that any JIT data is also freed.
  .sp
  In some circumstances you may need to call additional functions. These are
  described in the section entitled
  .\" HTML <a href="#stackcontrol">
  .\" </a>
-Line 60 
 described in the section entitled
+Line 62 
 described in the section entitled
  .\"
  below.
  .P
  If JIT support is not available, PCRE_STUDY_JIT_COMPILE is ignored, and no JIT
  data is set up. Otherwise, the compiled pattern is passed to the JIT compiler,
  which turns it into machine code that executes much faster than the normal
  interpretive code. When \fBpcre_exec()\fP is passed a \fBpcre_extra\fP block
  containing a pointer to JIT code, it obeys that instead of the normal code. The
  result is identical, but the code runs much faster.
  .P
  There are some \fBpcre_exec()\fP options that are not supported for JIT
  execution. There are also some pattern items that JIT cannot handle. Details
  are given below. In both cases, execution automatically falls back to the
  interpretive code.
  .P
  If the JIT compiler finds an unsupported item, no JIT data is generated. You
  can find out if JIT execution is available after studying a pattern by calling
  \fBpcre_fullinfo()\fP with the PCRE_INFO_JIT option. A result of 1 means that
- JIT compilationw was successful. A result of 0 means that JIT support is not
+ JIT compilation was successful. A result of 0 means that JIT support is not
  available, or the pattern was not studied with PCRE_STUDY_JIT_COMPILE, or the
  JIT compiler was not able to handle the pattern.
+ .P
+ Once a pattern has been studied, with or without JIT, it can be used as many
+ times as you like for matching different subject strings.
  .
  .
  .SH "UNSUPPORTED OPTIONS AND PATTERN ITEMS"
  .rs
  .sp
  The only \fBpcre_exec()\fP options that are supported for JIT execution are
  PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, and
  PCRE_NOTEMPTY_ATSTART. Note in particular that partial matching is not
  supported.
  .P
  The unsupported pattern items are:
  .sp
-   \eC            match a single byte, even in UTF-8 mode
+   \eC            match a single byte; not supported in UTF-8 mode
    (?Cn)          callouts
    (?(<name>)...  conditional test on setting of a named subpattern
    (?(R)...       conditional test on whole pattern recursion
    (?(Rn)...      conditional test on recursion, by number
    (?(R&name)...  conditional test on recursion, by name
    (*COMMIT)      )
-Line 101 
 The unsupported pattern items are:
+Line 106 
 The unsupported pattern items are:
    (*PRUNE)       ) the backtracking control verbs
    (*SKIP)        )
    (*THEN)        )
  .sp
  Support for some of these may be added in future.
  .
  .
  .SH "RETURN VALUES FROM JIT EXECUTION"
  .rs
  .sp
  When a pattern is matched using JIT execution, the return values are the same
  as those given by the interpretive \fBpcre_exec()\fP code, with the addition of
  one new error code: PCRE_ERROR_JIT_STACKLIMIT. This means that the memory used
  for the JIT stack was insufficient. See
  .\" HTML <a href="#stackcontrol">
  .\" </a>
  "Controlling the JIT stack"
  .\"
- below for a discussion of JIT stack usage.
+ below for a discussion of JIT stack usage. For compatibility with the
+ interpretive \fBpcre_exec()\fP code, no more than two-thirds of the
+ \fIovector\fP argument is used for passing back captured substrings.
  .P
  The error code PCRE_ERROR_MATCHLIMIT is returned by the JIT code if searching a
  very large pattern tree goes on for too long, as it is in the same circumstance
-Line 128 
 execution.
+Line 135 
 execution.
  .SH "SAVING AND RESTORING COMPILED PATTERNS"
  .rs
  .sp
  The code that is generated by the JIT compiler is architecture-specific, and is
- also position dependent. For those reasons it cannot be saved and restored like
+ also position dependent. For those reasons it cannot be saved (in a file or
- the bytecode and other data of a compiled pattern. You should be able run
+ database) and restored later like the bytecode and other data of a compiled
- \fBpcre_study()\fP on a saved and restored pattern, and thereby recreate the
+ pattern. Saving and restoring compiled patterns is not something many people
- JIT data, but because JIT compilation uses significant resources, it is
+ do. More detail about this facility is given in the
- probably not worth doing.
+ .\" HREF
+ \fBpcreprecompile\fP
+ .\"
+ documentation. It should be possible to run \fBpcre_study()\fP on a saved and
+ restored pattern, and thereby recreate the JIT data, but because JIT
+ compilation uses significant resources, it is probably not worth doing this;
+ you might as well recompile the original pattern.
  .
  .
  .\" HTML <a name="stackcontrol"></a>
  .SH "CONTROLLING THE JIT STACK"
  .rs
  .sp
  When the compiled JIT code runs, it needs a block of memory to use as a stack.
  By default, it uses 32K on the machine stack. However, some large or
  complicated patterns need more than this. The error PCRE_ERROR_JIT_STACKLIMIT
  is given when there is not enough stack. Three functions are provided for
- setting up alternative blocks of memory for use as JIT stacks.
+ managing blocks of memory for use as JIT stacks.
  .P
  The \fBpcre_jit_stack_alloc()\fP function creates a JIT stack. Its arguments
- are a starting size and a maximum size, and it returns an opaque value
+ are a starting size and a maximum size, and it returns a pointer to an opaque
- of type \fBpcre_jit_stack\fP that represents a JIT stack, or NULL if there is
+ structure of type \fBpcre_jit_stack\fP, or NULL if there is an error. The
- an error. The \fBpcre_jit_stack_free()\fP function can be used to free a stack
+ \fBpcre_jit_stack_free()\fP function can be used to free a stack that is no
- that is no longer needed.
+ longer needed. (For the technically minded: the address space is allocated by
+ mmap or VirtualAlloc.)
  .P
- The \fBpcre_assign_jit_stack()\fP function specifies which stack JIT code
+ JIT uses far less memory for recursion than the interpretive code,
+ and a maximum stack size of 512K to 1M should be more than enough for any
+ pattern.
+ .P
+ The \fBpcre_assign_jit_stack()\fP function specifies which stack JIT code
  should use. Its arguments are as follows:
  .sp
    pcre_extra         *extra
    pcre_jit_callback  callback
    void               *data
  .sp
  The \fIextra\fP argument must be the result of studying a pattern with
  PCRE_STUDY_JIT_COMPILE. There are three cases for the values of the other two
  options:
  .sp
    (1) If \fIcallback\fP is NULL and \fIdata\fP is NULL, an internal 32K block
-Line 170 
 options:
+Line 188 
 options:
        a valid JIT stack, the result of calling \fBpcre_jit_stack_alloc()\fP.
  .sp
    (3) If \fIcallback\fP not NULL, it must point to a function that is called
        with \fIdata\fP as an argument at the start of matching, in order to
        set up a JIT stack. If the result is NULL, the internal 32K stack
        is used; otherwise the return value must be a valid JIT stack,
        the result of calling \fBpcre_jit_stack_alloc()\fP.
  .sp
  You may safely assign the same JIT stack to more than one pattern, as long as
  they are all matched sequentially in the same thread. In a multithread
  application, each thread must use its own JIT stack.
  .P
+ Strictly speaking, even more is allowed. You can assign the same stack to any
+ number of patterns as long as they are not used for matching by multiple
+ threads at the same time. For example, you can assign the same stack to all
+ compiled patterns, and use a global mutex in the callback to wait until the
+ stack is available for use. However, this is an inefficient solution, and
+ not recommended.
+ .P
+ This is a suggestion for how a typical multithreaded program might operate:
+ .sp
+   During thread initalization
+     thread_local_var = pcre_jit_stack_alloc(...)
+ .sp
+   During thread exit
+     pcre_jit_stack_free(thread_local_var)
+ .sp
+   Use a one-line callback function
+     return thread_local_var
+ .sp
  All the functions described in this section do nothing if JIT is not available,
  and \fBpcre_assign_jit_stack()\fP does nothing unless the \fBextra\fP argument
  is non-NULL and points to a \fBpcre_extra\fP block that is the result of a
  successful study with PCRE_STUDY_JIT_COMPILE.
  .
  .
  .SH "EXAMPLE CODE"
  .rs
  .sp
  This is a single-threaded example that specifies a JIT stack without using a
  callback.
  .sp
    int rc;
+   int ovector[30];
    pcre *re;
    pcre_extra *extra;
    pcre_jit_stack *jit_stack;
  .sp
    re = pcre_compile(pattern, 0, &error, &erroffset, NULL);
    /* Check for errors */
    extra = pcre_study(re, PCRE_STUDY_JIT_COMPILE, &error);
-   jit_stack = pcre_jit_stack_alloc(1, 512 * 1024);
+   jit_stack = pcre_jit_stack_alloc(32*1024, 512*1024);
    /* Check for error (NULL) */
    pcre_assign_jit_stack(extra, NULL, jit_stack);
-   rc = pcre_exec(re, extra, subject, length, 0, 0, ovector, ovecsize);
+   rc = pcre_exec(re, extra, subject, length, 0, 0, ovector, 30);
    /* Check results */
    pcre_free(re);
    pcre_free_study(extra);
+   pcre_jit_stack_free(jit_stack);
  .sp
  .
  .
-Line 229 
 Cambridge CB2 3QH, England.
+Line 267 
 Cambridge CB2 3QH, England.
  .rs
  .sp
  .nf
- Last updated: 28 August 2011
+ Last updated: 19 October 2011
  Copyright (c) 1997-2011 University of Cambridge.
  .fi

 Legend:



Removed from v.678
 


changed lines


 
Added in v.737
 Legend:



Removed from v.678
 


changed lines


 
Added in v.737
-Removed from v.678
+Added in v.737

	ViewVC Help
Powered by ViewVC 1.1.5