v7-0006-pg_dump-compression-API-read_func-gets_func.patch

From d127c9afdff0710babd8db28d6b9870c537cbff7 Mon Sep 17 00:00:00 2001
From: Daniel Gustafsson <dgustafsson@postgresql.org>
Date: Wed, 25 Jun 2025 15:53:00 +0200
Subject: [PATCH v7 6/6] pg_dump compression API: read_func / gets_func

read_func is defined to return the number of bytes read, zero for EOF
and throw an error on all error conditions

 * lz4: Make sure to call throw an error and not return -1
 * gzip: gzread returning zero cannot be assumed to indicate EOF as
   it is documented to return zero for some types of errors.
 * none: Don't call ferror on every fread unless it returne zero and
   we can expect there to be an error or EOF.

gets_func is defined go return *s on success and NULL on error or
if EOF is found before any char has been read.

 * lz4, zstd: Convert _read_internal functions to not automatically
   call pg_fatal on errors to be able to handle gets returning NULL
   on error.
---
 src/bin/pg_dump/compress_gzip.c | 15 +++++++++++++--
 src/bin/pg_dump/compress_io.h   |  2 +-
 src/bin/pg_dump/compress_lz4.c  | 19 +++++++++++++++----
 src/bin/pg_dump/compress_none.c |  2 +-
 src/bin/pg_dump/compress_zstd.c | 28 +++++++++++++++++++++++-----
 5 files changed, 53 insertions(+), 13 deletions(-)

diff --git a/src/bin/pg_dump/compress_gzip.c b/src/bin/pg_dump/compress_gzip.c
index 2edcbffdd8d..4a067e1402c 100644
--- a/src/bin/pg_dump/compress_gzip.c
+++ b/src/bin/pg_dump/compress_gzip.c
@@ -258,10 +258,21 @@ Gzip_read(void *ptr, size_t size, CompressFileHandle *CFH)
 	int			gzret;
 
 	gzret = gzread(gzfp, ptr, size);
-	if (gzret < 0)
+
+	/*
+	 * gzread returns zero on EOF as well as some error conditions, and less
+	 * than zero on other error conditions, so we need to inspect for EOF on
+	 * zero.
+	 */
+	if (gzret <= 0)
 	{
 		int			errnum;
-		const char *errmsg = gzerror(gzfp, &errnum);
+		const char *errmsg;
+
+		if (gzret == 0 && gzeof(gzfp))
+			return 0;
+
+		errmsg = gzerror(gzfp, &errnum);
 
 		pg_fatal("could not read from input file: %s",
 				 errnum == Z_ERRNO ? strerror(errno) : errmsg);
diff --git a/src/bin/pg_dump/compress_io.h b/src/bin/pg_dump/compress_io.h
index e4d98ca08d7..82f5c71f764 100644
--- a/src/bin/pg_dump/compress_io.h
+++ b/src/bin/pg_dump/compress_io.h
@@ -127,7 +127,7 @@ struct CompressFileHandle
 	 * 'ptr'.
 	 *
 	 * Returns number of bytes read (this might be less than 'size' if EOF was
-	 * reached).  Throws error for error conditions other than EOF.
+	 * reached).  Exits via pg_fatal all for error conditions.
 	 */
 	size_t		(*read_func) (void *ptr, size_t size,
 							  CompressFileHandle *CFH);
diff --git a/src/bin/pg_dump/compress_lz4.c b/src/bin/pg_dump/compress_lz4.c
index 700915fa90d..e2f7c468293 100644
--- a/src/bin/pg_dump/compress_lz4.c
+++ b/src/bin/pg_dump/compress_lz4.c
@@ -459,7 +459,11 @@ LZ4Stream_read_internal(LZ4State *state, void *ptr, int ptrsize, bool eol_flag)
 
 	/* Lazy init */
 	if (!LZ4Stream_init(state, size, false /* decompressing */ ))
+	{
+		pg_log_error("unable to initialize LZ4 library: %s",
+					 LZ4F_getErrorName(state->errcode));
 		return -1;
+	}
 
 	/* No work needs to be done for a zero-sized output buffer */
 	if (size <= 0)
@@ -486,7 +490,10 @@ LZ4Stream_read_internal(LZ4State *state, void *ptr, int ptrsize, bool eol_flag)
 
 		rsize = fread(readbuf, 1, size, state->fp);
 		if (rsize < size && !feof(state->fp))
+		{
+			pg_log_error("could not read from input file: %m");
 			return -1;
+		}
 
 		rp = (char *) readbuf;
 		rend = (char *) readbuf + rsize;
@@ -503,6 +510,8 @@ LZ4Stream_read_internal(LZ4State *state, void *ptr, int ptrsize, bool eol_flag)
 			if (LZ4F_isError(status))
 			{
 				state->errcode = status;
+				pg_log_error("could not read from input file: %s",
+							 LZ4F_getErrorName(state->errcode));
 				return -1;
 			}
 
@@ -639,11 +648,13 @@ LZ4Stream_gets(char *ptr, int size, CompressFileHandle *CFH)
 	int			ret;
 
 	ret = LZ4Stream_read_internal(state, ptr, size - 1, true);
-	if (ret < 0 || (ret == 0 && !LZ4Stream_eof(CFH)))
-		pg_fatal("could not read from input file: %s", LZ4Stream_get_error(CFH));
 
-	/* Done reading */
-	if (ret == 0)
+	/*
+	 * LZ4Stream_read_internal returning 0 or -1 means that it was either an
+	 * EOF or an error, but gets_func is defined to return NULL in either case
+	 * so we can treat both the same here.
+	 */
+	if (ret <= 0)
 		return NULL;
 
 	/*
diff --git a/src/bin/pg_dump/compress_none.c b/src/bin/pg_dump/compress_none.c
index 9a24ef96cd0..c6234eec25c 100644
--- a/src/bin/pg_dump/compress_none.c
+++ b/src/bin/pg_dump/compress_none.c
@@ -90,7 +90,7 @@ read_none(void *ptr, size_t size, CompressFileHandle *CFH)
 	size_t		ret;
 
 	ret = fread(ptr, 1, size, fp);
-	if (ferror(fp))
+	if (ret == 0 && ferror(fp))
 		pg_fatal("could not read from input file: %m");
 
 	return ret;
diff --git a/src/bin/pg_dump/compress_zstd.c b/src/bin/pg_dump/compress_zstd.c
index 706c2427212..e24d45e1bbe 100644
--- a/src/bin/pg_dump/compress_zstd.c
+++ b/src/bin/pg_dump/compress_zstd.c
@@ -260,7 +260,7 @@ InitCompressorZstd(CompressorState *cs,
  */
 
 static size_t
-Zstd_read(void *ptr, size_t size, CompressFileHandle *CFH)
+Zstd_read_internal(void *ptr, size_t size, CompressFileHandle *CFH, bool exit_on_error)
 {
 	ZstdCompressorState *zstdcs = (ZstdCompressorState *) CFH->private_data;
 	ZSTD_inBuffer *input = &zstdcs->input;
@@ -278,7 +278,11 @@ Zstd_read(void *ptr, size_t size, CompressFileHandle *CFH)
 		zstdcs->input.src = pg_malloc0(input_allocated_size);
 		zstdcs->dstream = ZSTD_createDStream();
 		if (zstdcs->dstream == NULL)
-			pg_fatal("could not initialize compression library");
+		{
+			if (exit_on_error)
+				pg_fatal("could not initialize compression library");
+			return -1;
+		}
 	}
 
 	output->size = size;
@@ -306,7 +310,11 @@ Zstd_read(void *ptr, size_t size, CompressFileHandle *CFH)
 		{
 			cnt = fread(unconstify(void *, input->src), 1, input_allocated_size, zstdcs->fp);
 			if (ferror(zstdcs->fp))
-				pg_fatal("could not read from input file: %m");
+			{
+				if (exit_on_error)
+					pg_fatal("could not read from input file: %m");
+				return -1;
+			}
 
 			input->size = cnt;
 
@@ -323,7 +331,11 @@ Zstd_read(void *ptr, size_t size, CompressFileHandle *CFH)
 			res = ZSTD_decompressStream(zstdcs->dstream, output, input);
 
 			if (ZSTD_isError(res))
-				pg_fatal("could not decompress data: %s", ZSTD_getErrorName(res));
+			{
+				if (exit_on_error)
+					pg_fatal("could not decompress data: %s", ZSTD_getErrorName(res));
+				return -1;
+			}
 
 			if (output->pos == output->size)
 				break;			/* No more room for output */
@@ -402,7 +414,7 @@ Zstd_gets(char *buf, int len, CompressFileHandle *CFH)
 	 */
 	for (i = 0; i < len - 1; ++i)
 	{
-		if (CFH->read_func(&buf[i], 1, CFH) != 1)
+		if (Zstd_read_internal(&buf[i], 1, CFH, false) != 1)
 			break;
 		if (buf[i] == '\n')
 		{
@@ -414,6 +426,12 @@ Zstd_gets(char *buf, int len, CompressFileHandle *CFH)
 	return i > 0 ? buf : NULL;
 }
 
+static size_t
+Zstd_read(void *ptr, size_t size, CompressFileHandle *CFH)
+{
+	return Zstd_read_internal(ptr, size, CFH, true);
+}
+
 static bool
 Zstd_close(CompressFileHandle *CFH)
 {
-- 
2.39.3 (Apple Git-146)