Add checksum helper functions.

These functions make it easier to write code that wants to compute a
checksum for some data while allowing the user to configure the type
of checksum that gets used.

This is another piece of infrastructure for the upcoming patch to add
backup manifests.

Patch written from scratch by me, but it is similar to previous work
by Rushabh Lathia and Suraj Kharage. Suraj also reviewed this version
off-list. Advice on how not to break Windows from Davinder Singh.

Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZRTBiPyvQEwV79PU1ePTtSEo2UeVncrkJMbn1sU1gnRA@mail.gmail.com

diff --git a/src/common/Makefile b/src/common/Makefile
index a97c723fbda80d46d1b8e08fb3c7719559dc4f0b..d0be882cca4be4d96e65196a1858c0e19134d7ee 100644 (file)
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -48,6 +48,7 @@ LIBS += $(PTHREAD_LIBS)
 OBJS_COMMON = \
    archive.o \
    base64.o \
+   checksum_helper.o \
    config_info.o \
    controldata_utils.o \
    d2s.o \

diff --git a/src/common/checksum_helper.c b/src/common/checksum_helper.c
new file mode 100644 (file)
index 0000000..79a9a74
--- /dev/null
+++ b/src/common/checksum_helper.c
@@ -0,0 +1,190 @@
+/*-------------------------------------------------------------------------
+ *
+ * checksum_helper.c
+ *   Compute a checksum of any of various types using common routines
+ *
+ * Portions Copyright (c) 2016-2020, PostgreSQL Global Development Group
+ *
+ * IDENTIFICATION
+ *       src/common/checksum_helper.c
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#ifndef FRONTEND
+#include "postgres.h"
+#else
+#include "postgres_fe.h"
+#endif
+
+#include "common/checksum_helper.h"
+
+/*
+ * If 'name' is a recognized checksum type, set *type to the corresponding
+ * constant and return true. Otherwise, set *type to CHECKSUM_TYPE_NONE and
+ * return false.
+ */
+bool
+pg_checksum_parse_type(char *name, pg_checksum_type *type)
+{
+   pg_checksum_type result_type = CHECKSUM_TYPE_NONE;
+   bool        result = true;
+
+   if (pg_strcasecmp(name, "none") == 0)
+       result_type = CHECKSUM_TYPE_NONE;
+   else if (pg_strcasecmp(name, "crc32c") == 0)
+       result_type = CHECKSUM_TYPE_CRC32C;
+   else if (pg_strcasecmp(name, "sha224") == 0)
+       result_type = CHECKSUM_TYPE_SHA224;
+   else if (pg_strcasecmp(name, "sha256") == 0)
+       result_type = CHECKSUM_TYPE_SHA256;
+   else if (pg_strcasecmp(name, "sha384") == 0)
+       result_type = CHECKSUM_TYPE_SHA384;
+   else if (pg_strcasecmp(name, "sha512") == 0)
+       result_type = CHECKSUM_TYPE_SHA512;
+   else
+       result = false;
+
+   *type = result_type;
+   return result;
+}
+
+/*
+ * Get the canonical human-readable name corresponding to a checksum type.
+ */
+char *
+pg_checksum_type_name(pg_checksum_type type)
+{
+   switch (type)
+   {
+       case CHECKSUM_TYPE_NONE:
+           return "NONE";
+       case CHECKSUM_TYPE_CRC32C:
+           return "CRC32C";
+       case CHECKSUM_TYPE_SHA224:
+           return "SHA224";
+       case CHECKSUM_TYPE_SHA256:
+           return "SHA256";
+       case CHECKSUM_TYPE_SHA384:
+           return "SHA384";
+       case CHECKSUM_TYPE_SHA512:
+           return "SHA512";
+   }
+
+   Assert(false);
+   return "???";
+}
+
+/*
+ * Initialize a checksum context for checksums of the given type.
+ */
+void
+pg_checksum_init(pg_checksum_context *context, pg_checksum_type type)
+{
+   context->type = type;
+
+   switch (type)
+   {
+       case CHECKSUM_TYPE_NONE:
+           /* do nothing */
+           break;
+       case CHECKSUM_TYPE_CRC32C:
+           INIT_CRC32C(context->raw_context.c_crc32c);
+           break;
+       case CHECKSUM_TYPE_SHA224:
+           pg_sha224_init(&context->raw_context.c_sha224);
+           break;
+       case CHECKSUM_TYPE_SHA256:
+           pg_sha256_init(&context->raw_context.c_sha256);
+           break;
+       case CHECKSUM_TYPE_SHA384:
+           pg_sha384_init(&context->raw_context.c_sha384);
+           break;
+       case CHECKSUM_TYPE_SHA512:
+           pg_sha512_init(&context->raw_context.c_sha512);
+           break;
+   }
+}
+
+/*
+ * Update a checksum context with new data.
+ */
+void
+pg_checksum_update(pg_checksum_context *context, const uint8 *input,
+                  size_t len)
+{
+   switch (context->type)
+   {
+       case CHECKSUM_TYPE_NONE:
+           /* do nothing */
+           break;
+       case CHECKSUM_TYPE_CRC32C:
+           COMP_CRC32C(context->raw_context.c_crc32c, input, len);
+           break;
+       case CHECKSUM_TYPE_SHA224:
+           pg_sha224_update(&context->raw_context.c_sha224, input, len);
+           break;
+       case CHECKSUM_TYPE_SHA256:
+           pg_sha256_update(&context->raw_context.c_sha256, input, len);
+           break;
+       case CHECKSUM_TYPE_SHA384:
+           pg_sha384_update(&context->raw_context.c_sha384, input, len);
+           break;
+       case CHECKSUM_TYPE_SHA512:
+           pg_sha512_update(&context->raw_context.c_sha512, input, len);
+           break;
+   }
+}
+
+/*
+ * Finalize a checksum computation and write the result to an output buffer.
+ *
+ * The caller must ensure that the buffer is at least PG_CHECKSUM_MAX_LENGTH
+ * bytes in length. The return value is the number of bytes actually written.
+ */
+int
+pg_checksum_final(pg_checksum_context *context, uint8 *output)
+{
+   int         retval = 0;
+
+   StaticAssertStmt(sizeof(pg_crc32c) <= PG_CHECKSUM_MAX_LENGTH,
+                    "CRC-32C digest too big for PG_CHECKSUM_MAX_LENGTH");
+   StaticAssertStmt(PG_SHA224_DIGEST_LENGTH <= PG_CHECKSUM_MAX_LENGTH,
+                    "SHA224 digest too for PG_CHECKSUM_MAX_LENGTH");
+   StaticAssertStmt(PG_SHA256_DIGEST_LENGTH <= PG_CHECKSUM_MAX_LENGTH,
+                    "SHA256 digest too for PG_CHECKSUM_MAX_LENGTH");
+   StaticAssertStmt(PG_SHA384_DIGEST_LENGTH <= PG_CHECKSUM_MAX_LENGTH,
+                    "SHA384 digest too for PG_CHECKSUM_MAX_LENGTH");
+   StaticAssertStmt(PG_SHA512_DIGEST_LENGTH <= PG_CHECKSUM_MAX_LENGTH,
+                    "SHA512 digest too for PG_CHECKSUM_MAX_LENGTH");
+
+   switch (context->type)
+   {
+       case CHECKSUM_TYPE_NONE:
+           break;
+       case CHECKSUM_TYPE_CRC32C:
+           FIN_CRC32C(context->raw_context.c_crc32c);
+           retval = sizeof(pg_crc32c);
+           memcpy(output, &context->raw_context.c_crc32c, retval);
+           break;
+       case CHECKSUM_TYPE_SHA224:
+           pg_sha224_final(&context->raw_context.c_sha224, output);
+           retval = PG_SHA224_DIGEST_LENGTH;
+           break;
+       case CHECKSUM_TYPE_SHA256:
+           pg_sha256_final(&context->raw_context.c_sha256, output);
+           retval = PG_SHA256_DIGEST_LENGTH;
+           break;
+       case CHECKSUM_TYPE_SHA384:
+           pg_sha384_final(&context->raw_context.c_sha384, output);
+           retval = PG_SHA384_DIGEST_LENGTH;
+           break;
+       case CHECKSUM_TYPE_SHA512:
+           pg_sha512_final(&context->raw_context.c_sha512, output);
+           retval = PG_SHA512_DIGEST_LENGTH;
+           break;
+   }
+
+   Assert(retval <= PG_CHECKSUM_MAX_LENGTH);
+   return retval;
+}

diff --git a/src/include/common/checksum_helper.h b/src/include/common/checksum_helper.h
new file mode 100644 (file)
index 0000000..48b0745
--- /dev/null
+++ b/src/include/common/checksum_helper.h
@@ -0,0 +1,74 @@
+/*-------------------------------------------------------------------------
+ *
+ * checksum_helper.h
+ *   Compute a checksum of any of various types using common routines
+ *
+ * Portions Copyright (c) 2016-2020, PostgreSQL Global Development Group
+ *
+ * IDENTIFICATION
+ *       src/include/common/checksum_helper.h
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#ifndef CHECKSUM_HELPER_H
+#define CHECKSUM_HELPER_H
+
+#include "common/sha2.h"
+#include "port/pg_crc32c.h"
+
+/*
+ * Supported checksum types. It's not necessarily the case that code using
+ * these functions needs a cryptographically strong checksum; it may only
+ * need to detect accidental modification. That's why we include CRC-32C: it's
+ * much faster than any of the other algorithms. On the other hand, we omit
+ * MD5 here because any new that does need a cryptographically strong checksum
+ * should use something better.
+ */
+typedef enum pg_checksum_type
+{
+   CHECKSUM_TYPE_NONE,
+   CHECKSUM_TYPE_CRC32C,
+   CHECKSUM_TYPE_SHA224,
+   CHECKSUM_TYPE_SHA256,
+   CHECKSUM_TYPE_SHA384,
+   CHECKSUM_TYPE_SHA512
+} pg_checksum_type;
+
+/*
+ * This is just a union of all applicable context types.
+ */
+typedef union pg_checksum_raw_context
+{
+   pg_crc32c   c_crc32c;
+   pg_sha224_ctx c_sha224;
+   pg_sha256_ctx c_sha256;
+   pg_sha384_ctx c_sha384;
+   pg_sha512_ctx c_sha512;
+} pg_checksum_raw_context;
+
+/*
+ * This structure provides a convenient way to pass the checksum type and the
+ * checksum context around together.
+ */
+typedef struct pg_checksum_context
+{
+   pg_checksum_type type;
+   pg_checksum_raw_context raw_context;
+} pg_checksum_context;
+
+/*
+ * This is the longest possible output for any checksum algorithm supported
+ * by this file.
+ */
+#define PG_CHECKSUM_MAX_LENGTH     PG_SHA512_DIGEST_LENGTH
+
+extern bool pg_checksum_parse_type(char *name, pg_checksum_type *);
+extern char *pg_checksum_type_name(pg_checksum_type);
+
+extern void pg_checksum_init(pg_checksum_context *, pg_checksum_type);
+extern void pg_checksum_update(pg_checksum_context *, const uint8 *input,
+                              size_t len);
+extern int pg_checksum_final(pg_checksum_context *, uint8 *output);
+
+#endif

diff --git a/src/tools/msvc/Mkvcbuild.pm b/src/tools/msvc/Mkvcbuild.pm
index 5c88825f4969935927c0703be73f813092879918..72a21dbd419437362464480f8d911e874481509c 100644 (file)
--- a/src/tools/msvc/Mkvcbuild.pm
+++ b/src/tools/msvc/Mkvcbuild.pm
@@ -120,8 +120,8 @@ sub mkvcbuild
    }
 
    our @pgcommonallfiles = qw(
-     archive.c
-     base64.c config_info.c controldata_utils.c d2s.c encnames.c exec.c
+     archive.c base64.c checksum_helper.c
+      config_info.c controldata_utils.c d2s.c encnames.c exec.c
      f2s.c file_perm.c hashfn.c ip.c jsonapi.c
      keywords.c kwlookup.c link-canary.c md5.c
      pg_lzcompress.c pgfnames.c psprintf.c relpath.c rmtree.c