Use big_alloc().

[libucw.git] / lib / fastbuf.h

diff --git a/lib/fastbuf.h b/lib/fastbuf.h
index d831ed36153a877f27d0cf5f9cb85d90ac615604..922798ff183515a3d102191c6e614569373088e1 100644 (file)
--- a/lib/fastbuf.h
+++ b/lib/fastbuf.h
@@ -1,18 +1,28 @@
 /*
 /*

- *     Sherlock Library -- Fast Buffered I/O
+ *     UCW Library -- Fast Buffered I/O

  *
  *

- *     (c) 1997--2000 Martin Mares <mj@ucw.cz>
+ *     (c) 1997--2004 Martin Mares <mj@ucw.cz>
+ *     (c) 2004 Robert Spalek <robert@ucw.cz>
+ *
+ *     This software may be freely distributed and used according to the terms
+ *     of the GNU Lesser General Public License.

  */
 
  */
 

-#ifndef _SHERLOCK_FASTBUF_H
-#define _SHERLOCK_FASTBUF_H
+#ifndef _UCW_FASTBUF_H
+#define _UCW_FASTBUF_H

 
 #ifndef EOF
 #include <stdio.h>
 #endif
 
 
 #ifndef EOF
 #include <stdio.h>
 #endif
 

+#include <string.h>
+#include <alloca.h>
+
+#include "lib/unaligned.h"
+

 /*
 /*

- *  Generic buffered I/O on a top of buffer swapping functions.
+ *  Generic buffered I/O. You supply hooks to be called for low-level operations
+ *  (swapping of buffers, seeking and closing), we do the rest.

  *
  *  Buffer layout when reading:
  *
  *
  *  Buffer layout when reading:
  *


@@ -28,51 +38,123 @@
  *
  *  When writing:
  *
  *
  *  When writing:
  *

- *  +----------------+---------------------------+
- *  | written data   | free space                |
- *  +----------------+---------------------------+
- *  ^                 ^                           ^
- *  buffer=bstop      bptr                        bufend
+ *  +--------+--------------+--------------------+
+ *  | unused | written data | free space         |
+ *  +--------+--------------+--------------------+
+ *  ^         ^              ^                    ^
+ *  buffer    bstop          bptr                 bufend
+ *
+ *  Dirty tricks:
+ *
+ *    - You can mix reads and writes on the same stream, but you must
+ *     call bflush() in between and remember that the file position
+ *     points after the flushed buffer which is not necessarily the same
+ *     as after the data you've read.
+ *    - The spout/refill hooks can change not only bptr and bstop, but also
+ *     the location of the buffer; fb-mem.c takes advantage of it.
+ *    - In some cases, the user of the bdirect interface can be allowed to modify
+ *     the data in the buffer to avoid unnecessary copying. If the back-end
+ *     allows such modifications, it can set can_overwrite_buffer accordingly:
+ *             *  0 if no modification is allowed,
+ *             *  1 if the user can modify the buffer on the condition that
+ *                  the modifications will be undone before calling the next
+ *                  fastbuf operation
+ *             *  2 if the user is allowed to overwrite the data in the buffer
+ *                  if bdirect_read_commit_modified() is called afterwards.
+ *                  In this case, the back-end must be prepared for trimming
+ *                  of the buffer which is done by the commit function.

  */
 
 struct fastbuf {
  */
 
 struct fastbuf {

+  byte is_fastbuf[0];                  /* Dummy field for checking of type casts */

   byte *bptr, *bstop;                  /* Access pointers */
   byte *buffer, *bufend;               /* Start and end of the buffer */
   byte *name;                          /* File name for error messages */
   byte *bptr, *bstop;                  /* Access pointers */
   byte *buffer, *bufend;               /* Start and end of the buffer */
   byte *name;                          /* File name for error messages */

-  uns buflen;                          /* Size of the buffer */
-  sh_off_t pos;                                /* Position of buffer start in the file */
-  sh_off_t fdpos;                      /* Current position in the non-buffered file */
-  int fd;                              /* File descriptor, -1 if not a real file */
-  int is_temp_file;                    /* Is a temporary file, delete on close */
-  void *lldata;                                /* Data private to access functions below */
-  void *llpos;                         /* ... continued ... */
+  sh_off_t pos;                                /* Position of bstop in the file */

   int (*refill)(struct fastbuf *);     /* Get a buffer with new data */
   void (*spout)(struct fastbuf *);     /* Write buffer data to the file */
   void (*seek)(struct fastbuf *, sh_off_t, int);  /* Slow path for bseek(), buffer already flushed */
   void (*close)(struct fastbuf *);     /* Close the stream */
   int (*refill)(struct fastbuf *);     /* Get a buffer with new data */
   void (*spout)(struct fastbuf *);     /* Write buffer data to the file */
   void (*seek)(struct fastbuf *, sh_off_t, int);  /* Slow path for bseek(), buffer already flushed */
   void (*close)(struct fastbuf *);     /* Close the stream */

+  int (*config)(struct fastbuf *, uns, int);   /* Configure the stream */
+  int can_overwrite_buffer;            /* Can the buffer be altered? (see discussion above) 0=never, 1=temporarily, 2=permanently */

 };
 
 };
 

-/* FastIO on standard files */
+/* FastIO on standard files (specify buffer size 0 to enable mmaping) */

 
 

-struct fastbuf *bopen(byte *name, uns mode, uns buffer);
-struct fastbuf *bfdopen(int fd, uns buffer);
-void bbcopy(struct fastbuf *f, struct fastbuf *t, uns l);
+struct fastbuf *bopen(byte *name, uns mode, uns buflen);
+struct fastbuf *bopen_try(byte *name, uns mode, uns buflen);
+struct fastbuf *bopen_tmp(uns buflen);
+struct fastbuf *bfdopen(int fd, uns buflen);
+struct fastbuf *bfdopen_shared(int fd, uns buflen);
+void bfilesync(struct fastbuf *b);

 
 /* FastIO on in-memory streams */
 
 struct fastbuf *fbmem_create(unsigned blocksize);      /* Create stream and return its writing fastbuf */
 struct fastbuf *fbmem_clone_read(struct fastbuf *);    /* Create reading fastbuf */
 
 
 /* FastIO on in-memory streams */
 
 struct fastbuf *fbmem_create(unsigned blocksize);      /* Create stream and return its writing fastbuf */
 struct fastbuf *fbmem_clone_read(struct fastbuf *);    /* Create reading fastbuf */
 

+/* FastIO on memory mapped files */
+
+struct fastbuf *bopen_mm(byte *name, uns mode);
+
+/* FastI on file descriptors with limit */
+
+struct fastbuf *bopen_limited_fd(int fd, uns bufsize, uns limit);
+
+/* FastIO on static buffers */
+
+void fbbuf_init_read(struct fastbuf *f, byte *buffer, uns size, uns can_overwrite);
+void fbbuf_init_write(struct fastbuf *f, byte *buffer, uns size);
+static inline uns
+fbbuf_count_written(struct fastbuf *f)
+{
+  return f->bptr - f->bstop;
+}
+
+/* FastIO on recyclable growing buffers */
+
+struct fastbuf *fbgrow_create(unsigned basic_size);
+void fbgrow_reset(struct fastbuf *b);                  /* Reset stream and prepare for writing */
+void fbgrow_rewind(struct fastbuf *b);                 /* Prepare for reading */
+
+/* FastO with atomic writes for multi-threaded programs */
+
+struct fb_atomic {
+  struct fastbuf fb;
+  struct fb_atomic_file *af;
+  byte *expected_max_bptr;
+  uns slack_size;
+};
+#define FB_ATOMIC(f) ((struct fb_atomic *)(f)->is_fastbuf)
+
+struct fastbuf *fbatomic_open(byte *name, struct fastbuf *master, uns bufsize, int record_len);
+void fbatomic_internal_write(struct fastbuf *b);
+
+static inline void
+fbatomic_commit(struct fastbuf *b)
+{
+  if (b->bptr >= ((struct fb_atomic *)b)->expected_max_bptr)
+    fbatomic_internal_write(b);
+}
+
+/* Configuring stream parameters */
+
+int bconfig(struct fastbuf *f, uns type, int data);
+
+#define BCONFIG_IS_TEMP_FILE 0
+

 /* Universal functions working on all fastbuf's */
 
 void bclose(struct fastbuf *f);
 void bflush(struct fastbuf *f);
 void bseek(struct fastbuf *f, sh_off_t pos, int whence);
 void bsetpos(struct fastbuf *f, sh_off_t pos);
 /* Universal functions working on all fastbuf's */
 
 void bclose(struct fastbuf *f);
 void bflush(struct fastbuf *f);
 void bseek(struct fastbuf *f, sh_off_t pos, int whence);
 void bsetpos(struct fastbuf *f, sh_off_t pos);

+void brewind(struct fastbuf *f);
+sh_off_t bfilesize(struct fastbuf *f);

 
 static inline sh_off_t btell(struct fastbuf *f)
 {
 
 static inline sh_off_t btell(struct fastbuf *f)
 {

-  return f->pos + (f->bptr - f->buffer);
+  return f->pos + (f->bptr - f->bstop);

 }
 
 int bgetc_slow(struct fastbuf *f);
 }
 
 int bgetc_slow(struct fastbuf *f);


@@ -87,13 +169,13 @@ static inline int bpeekc(struct fastbuf *f)
   return (f->bptr < f->bstop) ? (int) *f->bptr : bpeekc_slow(f);
 }
 
   return (f->bptr < f->bstop) ? (int) *f->bptr : bpeekc_slow(f);
 }
 

-static inline void bungetc(struct fastbuf *f, byte c)
+static inline void bungetc(struct fastbuf *f)

 {
 {

-  *--f->bptr = c;
+  f->bptr--;

 }
 
 }
 

-void bputc_slow(struct fastbuf *f, byte c);
-static inline void bputc(struct fastbuf *f, byte c)
+void bputc_slow(struct fastbuf *f, uns c);
+static inline void bputc(struct fastbuf *f, uns c)

 {
   if (f->bptr < f->bufend)
     *f->bptr++ = c;
 {
   if (f->bptr < f->bufend)
     *f->bptr++ = c;


@@ -101,22 +183,25 @@ static inline void bputc(struct fastbuf *f, byte c)
     bputc_slow(f, c);
 }
 
     bputc_slow(f, c);
 }
 

-word bgetw_slow(struct fastbuf *f);
-static inline word bgetw(struct fastbuf *f)
+static inline uns
+bavailr(struct fastbuf *f)
+{
+  return f->bstop - f->bptr;
+}
+
+static inline uns
+bavailw(struct fastbuf *f)
+{
+  return f->bufend - f->bptr;
+}
+
+int bgetw_slow(struct fastbuf *f);
+static inline int bgetw(struct fastbuf *f)

 {
 {

-  word w;
-  if (f->bptr + 2 <= f->bstop)
+  int w;
+  if (bavailr(f) >= 2)

     {
     {

-      byte *p = f->bptr;
-#ifdef CPU_CAN_DO_UNALIGNED_WORDS
-      w = * ((word *) p);
-#else
-#ifdef CPU_BIG_ENDIAN
-      w = (p[0] << 8) | p[1];
-#else
-      w = (p[1] << 8) | p[0];
-#endif
-#endif
+      w = GET_U16(f->bptr);

       f->bptr += 2;
       return w;
     }
       f->bptr += 2;
       return w;
     }


@@ -128,18 +213,9 @@ u32 bgetl_slow(struct fastbuf *f);
 static inline u32 bgetl(struct fastbuf *f)
 {
   u32 l;
 static inline u32 bgetl(struct fastbuf *f)
 {
   u32 l;

-  if (f->bptr + 4 <= f->bstop)
+  if (bavailr(f) >= 4)

     {
     {

-      byte *p = f->bptr;
-#ifdef CPU_CAN_DO_UNALIGNED_LONGS
-      l = * ((u32 *) p);
-#else
-#ifdef CPU_BIG_ENDIAN
-      l = (p[0] << 24) | (p[1] << 16) | (p[2] << 8) | p[3];
-#else
-      l = (p[3] << 24) | (p[2] << 16) | (p[1] << 8) | p[0];
-#endif
-#endif
+      l = GET_U32(f->bptr);

       f->bptr += 4;
       return l;
     }
       f->bptr += 4;
       return l;
     }


@@ -150,10 +226,10 @@ static inline u32 bgetl(struct fastbuf *f)
 u64 bgetq_slow(struct fastbuf *f);
 static inline u64 bgetq(struct fastbuf *f)
 {
 u64 bgetq_slow(struct fastbuf *f);
 static inline u64 bgetq(struct fastbuf *f)
 {

-  if (f->bptr + 8 <= f->bstop)
+  u64 l;
+  if (bavailr(f) >= 8)

     {
     {

-      u64 l;
-      memcpy(&l, f->bptr, 8);
+      l = GET_U64(f->bptr);

       f->bptr += 8;
       return l;
     }
       f->bptr += 8;
       return l;
     }


@@ -165,14 +241,9 @@ u64 bget5_slow(struct fastbuf *f);
 static inline u64 bget5(struct fastbuf *f)
 {
   u64 l;
 static inline u64 bget5(struct fastbuf *f)
 {
   u64 l;

-  if (f->bptr + 5 <= f->bstop)
+  if (bavailr(f) >= 5)

     {
     {

-      byte *p = f->bptr;
-#ifdef CPU_BIG_ENDIAN
-      l = ((u64)p[0] << 32) | (u32)((p[1] << 24) | (p[2] << 16) | (p[3] << 8) | p[4]);
-#else
-      l = ((u64)p[4] << 32) | (u32)((p[3] << 24) | (p[2] << 16) | (p[1] << 8) | p[0]);
-#endif
+      l = GET_U40(f->bptr);

       f->bptr += 5;
       return l;
     }
       f->bptr += 5;
       return l;
     }


@@ -180,23 +251,12 @@ static inline u64 bget5(struct fastbuf *f)
     return bget5_slow(f);
 }
 
     return bget5_slow(f);
 }
 

-void bputw_slow(struct fastbuf *f, word w);
-static inline void bputw(struct fastbuf *f, word w)
+void bputw_slow(struct fastbuf *f, uns w);
+static inline void bputw(struct fastbuf *f, uns w)

 {
 {

-  if (f->bptr + 2 <= f->bufend)
+  if (bavailw(f) >= 2)

     {
     {

-      byte *p = f->bptr;
-#ifdef CPU_CAN_DO_UNALIGNED_WORDS
-      * ((word *) p) = w;
-#else
-#ifdef CPU_BIG_ENDIAN
-      p[0] = w >> 8U;
-      p[1] = w;
-#else
-      p[1] = w >> 8U;
-      p[0] = w;
-#endif
-#endif
+      PUT_U16(f->bptr, w);

       f->bptr += 2;
     }
   else
       f->bptr += 2;
     }
   else


@@ -206,24 +266,9 @@ static inline void bputw(struct fastbuf *f, word w)
 void bputl_slow(struct fastbuf *f, u32 l);
 static inline void bputl(struct fastbuf *f, u32 l)
 {
 void bputl_slow(struct fastbuf *f, u32 l);
 static inline void bputl(struct fastbuf *f, u32 l)
 {

-  if (f->bptr + 4 <= f->bufend)
+  if (bavailw(f) >= 4)

     {
     {

-      byte *p = f->bptr;
-#ifdef CPU_CAN_DO_UNALIGNED_LONGS
-      * ((u32 *) p) = l;
-#else
-#ifdef CPU_BIG_ENDIAN
-      p[0] = l >> 24U;
-      p[1] = l >> 16U;
-      p[2] = l >> 8U;
-      p[3] = l;
-#else
-      p[3] = l >> 24U;
-      p[2] = l >> 16U;
-      p[1] = l >> 8U;
-      p[0] = l;
-#endif
-#endif
+      PUT_U32(f->bptr, l);

       f->bptr += 4;
     }
   else
       f->bptr += 4;
     }
   else


@@ -233,9 +278,9 @@ static inline void bputl(struct fastbuf *f, u32 l)
 void bputq_slow(struct fastbuf *f, u64 l);
 static inline void bputq(struct fastbuf *f, u64 l)
 {
 void bputq_slow(struct fastbuf *f, u64 l);
 static inline void bputq(struct fastbuf *f, u64 l)
 {

-  if (f->bptr + 8 <= f->bufend)
+  if (bavailw(f) >= 8)

     {
     {

-      memcpy(f->bptr, &l, 8);
+      PUT_U64(f->bptr, l);

       f->bptr += 8;
     }
   else
       f->bptr += 8;
     }
   else


@@ -245,23 +290,9 @@ static inline void bputq(struct fastbuf *f, u64 l)
 void bput5_slow(struct fastbuf *f, u64 l);
 static inline void bput5(struct fastbuf *f, u64 l)
 {
 void bput5_slow(struct fastbuf *f, u64 l);
 static inline void bput5(struct fastbuf *f, u64 l)
 {

-  if (f->bptr + 5 <= f->bufend)
+  if (bavailw(f) >= 5)

     {
     {

-      byte *p = f->bptr;
-      u32 low = l;
-#ifdef CPU_BIG_ENDIAN
-      p[0] = l >> 32U;
-      p[1] = low >> 24U;
-      p[2] = low >> 16U;
-      p[3] = low >> 8U;
-      p[4] = low;
-#else
-      p[4] = l >> 32U;
-      p[3] = low >> 24U;
-      p[2] = low >> 16U;
-      p[1] = low >> 8U;
-      p[0] = low;
-#endif
+      PUT_U40(f->bptr, l);

       f->bptr += 5;
     }
   else
       f->bptr += 5;
     }
   else


@@ -271,7 +302,7 @@ static inline void bput5(struct fastbuf *f, u64 l)
 uns bread_slow(struct fastbuf *f, void *b, uns l, uns check);
 static inline uns bread(struct fastbuf *f, void *b, uns l)
 {
 uns bread_slow(struct fastbuf *f, void *b, uns l, uns check);
 static inline uns bread(struct fastbuf *f, void *b, uns l)
 {

-  if (f->bptr + l <= f->bstop)
+  if (bavailr(f) >= l)

     {
       memcpy(b, f->bptr, l);
       f->bptr += l;
     {
       memcpy(b, f->bptr, l);
       f->bptr += l;


@@ -283,7 +314,7 @@ static inline uns bread(struct fastbuf *f, void *b, uns l)
 
 static inline uns breadb(struct fastbuf *f, void *b, uns l)
 {
 
 static inline uns breadb(struct fastbuf *f, void *b, uns l)
 {

-  if (f->bptr + l <= f->bstop)
+  if (bavailr(f) >= l)

     {
       memcpy(b, f->bptr, l);
       f->bptr += l;
     {
       memcpy(b, f->bptr, l);
       f->bptr += l;


@@ -296,7 +327,7 @@ static inline uns breadb(struct fastbuf *f, void *b, uns l)
 void bwrite_slow(struct fastbuf *f, void *b, uns l);
 static inline void bwrite(struct fastbuf *f, void *b, uns l)
 {
 void bwrite_slow(struct fastbuf *f, void *b, uns l);
 static inline void bwrite(struct fastbuf *f, void *b, uns l)
 {

-  if (f->bptr + l <= f->bufend)
+  if (bavailw(f) >= l)

     {
       memcpy(f->bptr, b, l);
       f->bptr += l;
     {
       memcpy(f->bptr, b, l);
       f->bptr += l;


@@ -306,8 +337,23 @@ static inline void bwrite(struct fastbuf *f, void *b, uns l)
 }
 
 byte *bgets(struct fastbuf *f, byte *b, uns l);        /* Non-std */
 }
 
 byte *bgets(struct fastbuf *f, byte *b, uns l);        /* Non-std */

+int bgets_nodie(struct fastbuf *f, byte *b, uns l);

 byte *bgets0(struct fastbuf *f, byte *b, uns l);
 
 byte *bgets0(struct fastbuf *f, byte *b, uns l);
 

+struct mempool;
+struct bb_t;
+uns bgets_bb(struct fastbuf *f, struct bb_t *b, uns limit);
+byte *bgets_mp(struct fastbuf *f, struct mempool *mp);
+
+struct bgets_stk_struct {
+  struct fastbuf *f;
+  byte *old_buf, *cur_buf, *src;
+  uns old_len, cur_len, src_len;
+};
+void bgets_stk_init(struct bgets_stk_struct *s);
+void bgets_stk_step(struct bgets_stk_struct *s);
+#define bgets_stk(fb) ({ struct bgets_stk_struct _s; _s.f = (fb); for (bgets_stk_init(&_s); _s.cur_len; _s.cur_buf = alloca(_s.cur_len), bgets_stk_step(&_s)); _s.cur_buf; })
+

 static inline void
 bputs(struct fastbuf *f, byte *b)
 {
 static inline void
 bputs(struct fastbuf *f, byte *b)
 {


@@ -327,24 +373,87 @@ bputsn(struct fastbuf *f, byte *b)
   bputc(f, '\n');
 }
 
   bputc(f, '\n');
 }
 

-/* Direct I/O on buffers */
+void bbcopy_slow(struct fastbuf *f, struct fastbuf *t, uns l);
+static inline void
+bbcopy(struct fastbuf *f, struct fastbuf *t, uns l)
+{
+  if (bavailr(f) >= l && bavailw(t) >= l)
+    {
+      memcpy(t->bptr, f->bptr, l);
+      t->bptr += l;
+      f->bptr += l;
+    }
+  else
+    bbcopy_slow(f, t, l);
+}

 
 

-int bdirect_read(struct fastbuf *f, byte **buf);
-int bdirect_write_prepare(struct fastbuf *f, byte **buf);
-void bdirect_write_commit(struct fastbuf *f, byte *pos);
+int bskip_slow(struct fastbuf *f, uns len);
+static inline int bskip(struct fastbuf *f, uns len)
+{
+  if (bavailr(f) >= len)
+    {
+      f->bptr += len;
+      return 1;
+    }
+  else
+    return bskip_slow(f, len);
+}

 
 

-/* Depending on compile-time configuration, we select the right function for reading/writing of file offsets */
+/* I/O on addr_int_t */

 
 

-#ifdef SHERLOCK_CONFIG_LARGE_DB
-#define bgeto(f) bget5(f)
-#define bputo(f,l) bput5(f,l)
-#define bgetp(f) bgetq(f)
-#define bputp(f,l) bputq(f,l)
+#ifdef CPU_64BIT_POINTERS
+#define bputa(x,p) bputq(x,p)
+#define bgeta(x) bgetq(x)

 #else
 #else

-#define bgeto(f) bgetl(f)
-#define bputo(f,l) bputl(f,l)
-#define bgetp(f) bgetl(f)
-#define bputp(f,l) bputl(f,l)
+#define bputa(x,p) bputl(x,p)
+#define bgeta(x) bgetl(x)

 #endif
 
 #endif
 

+/* Direct I/O on buffers */
+
+static inline uns
+bdirect_read_prepare(struct fastbuf *f, byte **buf)
+{
+  if (f->bptr == f->bstop && !f->refill(f))
+    {
+      *buf = NULL;  // This is not needed, but it helps to get rid of spurious warnings
+      return 0;
+    }
+  *buf = f->bptr;
+  return bavailr(f);
+}
+
+static inline void
+bdirect_read_commit(struct fastbuf *f, byte *pos)
+{
+  f->bptr = pos;
+}
+
+static inline void
+bdirect_read_commit_modified(struct fastbuf *f, byte *pos)
+{
+  f->bptr = pos;
+  f->buffer = pos;     /* Avoid seeking backwards in the buffer */
+}
+
+static inline uns
+bdirect_write_prepare(struct fastbuf *f, byte **buf)
+{
+  if (f->bptr == f->bufend)
+    f->spout(f);
+  *buf = f->bptr;
+  return bavailw(f);
+}
+
+static inline void
+bdirect_write_commit(struct fastbuf *f, byte *pos)
+{
+  f->bptr = pos;
+}
+
+/* Formatted output */
+
+int bprintf(struct fastbuf *b, char *msg, ...) FORMAT_CHECK(printf,2,3);
+int vbprintf(struct fastbuf *b, char *msg, va_list args);
+

 #endif
 #endif