branches/zip: Write compressed pages to disk.

os_aio_simulated_handle(): Temporarily disable os_file_check_page_trailers(),
which cannot be invoked on compressed pages.

dict_table_add_system_columns(): New function, split from
dict_table_add_to_cache().

mlog_parse_index(): Add system columns to the dummy table and identify
DB_TRX_ID and DB_ROLL_PTR in the dummy index.

buf_LRU_get_free_block(): Note that page_zip->data should be allocated from
an aligned memory pool.

buf_flush_buffered_writes(): Write compressed pages to disk.

buf_flush_post_to_doublewrite_buf(): Copy compressed pages to the
doublewrite buffer.  Zero fill any excess space.

buf_flush_init_for_writing(): Treat all compressed pages the same.

buf_read_page_low(): Read compressed pages from disk.

buf_page_io_complete(): Process compressed pages.

trx_sys_doublewrite_init_or_restore_page(): Process compressed pages.

mlog_write_initial_log_record_fast(): Enable a debug printout
#ifdef UNIV_LOG_DEBUG.

fsp_header_init(), fsp_fill_free_list(): Pass the compressed page size
to buf_page_create().

page_zip_compress_write_log(): Flatten the if-else if-else logic.

page_zip_parse_write_blob_ptr(): Do not test page_zip if page==NULL.

page_zip_parse_write_node_ptr(): Do not test page_zip if page==NULL.
Invoke mlog_close() correctly.

row_sel_store_row_id_to_prebuilt(): Add UNIV_UNLIKELY hint to an
assertion-like test.

btr/btr0cur.c

@@ -3626,6 +3626,7 @@ btr_store_big_rec_extern_fields(
 
 			if (UNIV_LIKELY_NULL(page_zip)) {
 				int		err;
+				page_zip_des_t*	blob_page_zip;
 
 				mach_write_to_2(page + FIL_PAGE_TYPE,
 						FIL_PAGE_TYPE_ZBLOB);
@@ -3649,6 +3650,16 @@ btr_store_big_rec_extern_fields(
 					0, c_stream.avail_out);
 				mlog_log_string(page + FIL_PAGE_TYPE,
 					page_zip->size - FIL_PAGE_TYPE, &mtr);
+				/* Copy the page to compressed storage,
+				because it will be flushed to disk
+				from there. */
+				blob_page_zip = buf_block_get_page_zip(
+						buf_block_align(page));
+				ut_ad(blob_page_zip);
+				ut_ad(blob_page_zip->size == page_zip->size);
+				memcpy(blob_page_zip->data, page,
+						page_zip->size);
+				/* TODO: retain blob_page_zip, release page */
 
 				if (err == Z_OK && prev_page_no != FIL_NULL) {
 

buf/buf0buf.c

@@ -1961,20 +1961,44 @@ buf_page_io_complete(
 	if (io_type == BUF_IO_READ) {
 		ulint	read_page_no;
 		ulint	read_space_id;
+		byte*	frame;
 
 		if (block->page_zip.size) {
 			ut_a(block->space);
 
-			switch (fil_page_get_type(block->page_zip.data)) {
+			frame = block->page_zip.data;
+
+			switch (fil_page_get_type(frame)) {
 			case FIL_PAGE_INDEX:
+				if (block->frame) {
+					if (!page_zip_decompress(
+							&block->page_zip,
+							block->frame)) {
+						goto corrupt;
+					}
+				}
+				break;
+			case FIL_PAGE_INODE:
+			case FIL_PAGE_IBUF_BITMAP:
+			case FIL_PAGE_TYPE_FSP_HDR:
+			case FIL_PAGE_TYPE_XDES:
 			case FIL_PAGE_TYPE_ZBLOB:
-				/* TODO: checksum, but do not decompress */
+				/* Copy to uncompressed storage. */
+				memcpy(block->frame, frame,
+						block->page_zip.size);
 				break;
-			default:
-				/* TODO: how to distinguish uncompressed
-				and compressed pages? */
+			case 0:
+				/* uninitialized page */
 				break;
+			default:
+				ut_print_timestamp(stderr);
+				fprintf(stderr,
+			"InnoDB: unknown compressed page type %lu\n",
+					fil_page_get_type(frame));
+				goto corrupt;
 			}
+		} else {
+			frame = block->frame;
 		}
 
 		/* If this page is not uninitialized and not in the
@@ -2013,22 +2037,19 @@ buf_page_io_complete(
 		/* From version 3.23.38 up we store the page checksum
 		to the 4 first bytes of the page end lsn field */
 
-		if (buf_page_is_corrupted(block->frame/* TODO */,
-					block->page_zip.size)) {
+		if (buf_page_is_corrupted(frame, block->page_zip.size)) {
+corrupt:
 			fprintf(stderr,
 		"InnoDB: Database page corruption on disk or a failed\n"
-		"InnoDB: file read of page %lu.\n", (ulong) block->offset);
-
-			fputs(
-		"InnoDB: You may have to recover from a backup.\n", stderr);
-
-			buf_page_print(block->frame, block->page_zip.size);
-
+		"InnoDB: file read of page %lu.\n"
+		"InnoDB: You may have to recover from a backup.\n",
+				(ulong) block->offset);
+			buf_page_print(frame, block->page_zip.size);
 			fprintf(stderr,
 		"InnoDB: Database page corruption on disk or a failed\n"
-		"InnoDB: file read of page %lu.\n", (ulong) block->offset);
-			fputs(
-		"InnoDB: You may have to recover from a backup.\n", stderr);
+		"InnoDB: file read of page %lu.\n"
+		"InnoDB: You may have to recover from a backup.\n",
+				(ulong) block->offset);
 			fputs(
 		"InnoDB: It is also possible that your operating\n"
 		"InnoDB: system has corrupted its own file cache\n"

buf/buf0flu.c

@@ -365,8 +365,22 @@ flush:
 
 	for (i = 0; i < trx_doublewrite->first_free; i++) {
 		block = trx_doublewrite->buf_block_arr[i];
-		if (UNIV_LIKELY(!block->page_zip.data) && UNIV_UNLIKELY(
-				memcmp(block->frame + (FIL_PAGE_LSN + 4),
+		ut_a(block->state == BUF_BLOCK_FILE_PAGE);
+		if (UNIV_UNLIKELY(block->page_zip.size)) {
+			ulint	blk_size
+				= UNIV_PAGE_SIZE / block->page_zip.size;
+
+			fil_io(OS_FILE_WRITE | OS_AIO_SIMULATED_WAKE_LATER,
+					FALSE, block->space,
+					block->offset / blk_size,
+					(block->offset % blk_size)
+					* block->page_zip.size,
+					block->page_zip.size,
+					(void*)block->page_zip.data,
+					(void*)block);
+			continue;
+		} else if (UNIV_UNLIKELY(memcmp(
+				block->frame + (FIL_PAGE_LSN + 4),
 				block->frame + (UNIV_PAGE_SIZE
 				- FIL_PAGE_END_LSN_OLD_CHKSUM + 4), 4))) {
 			ut_print_timestamp(stderr);
@@ -379,7 +393,6 @@ flush:
 				(ulong)block->io_fix,
 				(ulong)block->state);
 		}
-		ut_a(block->state == BUF_BLOCK_FILE_PAGE);
 
 		fil_io(OS_FILE_WRITE | OS_AIO_SIMULATED_WAKE_LATER,
 			FALSE, block->space, block->offset, 0, UNIV_PAGE_SIZE,
@@ -417,6 +430,7 @@ buf_flush_post_to_doublewrite_buf(
 /*==============================*/
 	buf_block_t*	block)	/* in: buffer block to write */
 {
+	ulint	zip_size;
 try_again:
 	mutex_enter(&(trx_doublewrite->mutex));
 
@@ -431,10 +445,21 @@ try_again:
 		goto try_again;
 	}
 
-	/* TODO: page_zip */
+	zip_size = block->page_zip.size;
+
+	if (UNIV_UNLIKELY(zip_size)) {
+		/* Copy the compressed page and clear the rest. */
+		memcpy(trx_doublewrite->write_buf
+				+ UNIV_PAGE_SIZE * trx_doublewrite->first_free,
+				block->page_zip.data, zip_size);
+		memset(trx_doublewrite->write_buf
+				+ UNIV_PAGE_SIZE * trx_doublewrite->first_free
+				+ zip_size, 0, UNIV_PAGE_SIZE - zip_size);
+	} else {
 		memcpy(trx_doublewrite->write_buf
 				+ UNIV_PAGE_SIZE * trx_doublewrite->first_free,
 				block->frame, UNIV_PAGE_SIZE);
+	}
 
 	trx_doublewrite->buf_block_arr[trx_doublewrite->first_free] = block;
 
@@ -468,9 +493,10 @@ buf_flush_init_for_writing(
 	ulint		zip_size = fil_space_get_zip_size(space);
 
 	if (zip_size && zip_size != ULINT_UNDEFINED) {
+		ut_a(page_zip);
+		ut_a(page_zip->size == zip_size);
 		switch (UNIV_EXPECT(fil_page_get_type(page), FIL_PAGE_INDEX)) {
 		case FIL_PAGE_TYPE_ZBLOB:
-			ut_ad(!page_zip);
 			mach_write_to_4(page + FIL_PAGE_OFFSET, page_no);
 			mach_write_to_4(page + FIL_PAGE_ZBLOB_SPACE_ID, space);
 			mach_write_to_8(page + FIL_PAGE_LSN, newest_lsn);
@@ -484,8 +510,8 @@ buf_flush_init_for_writing(
 		case FIL_PAGE_IBUF_BITMAP:
 		case FIL_PAGE_TYPE_FSP_HDR:
 		case FIL_PAGE_TYPE_XDES:
-			/* This is essentially an uncompressed page. */
-			break;
+			/* These are essentially uncompressed pages. */
+			memcpy(page_zip->data, page, zip_size);
 		case FIL_PAGE_INDEX:
 			ut_a(zip_size == page_zip->size);
 			mach_write_to_4(page

buf/buf0lru.c

@@ -424,7 +424,7 @@ loop:
 			}
 
 			if (zip_size) {
-				/* TODO: allocate this from a separate pool */
+				/* TODO: allocate zip from an aligned pool */
 				block->page_zip.data = ut_malloc(zip_size);
 			} else {
 				block->page_zip.data = NULL;

buf/buf0rea.c

@@ -140,10 +140,20 @@ buf_read_page_low(
 
 	ut_a(block->state == BUF_BLOCK_FILE_PAGE);
 
+	if (zip_size) {
+		ulint	zip_blk = UNIV_PAGE_SIZE / zip_size;
+
+		*err = fil_io(OS_FILE_READ | wake_later,
+				sync, space,
+				offset / zip_blk, (offset % zip_blk)
+				* zip_size, zip_size,
+				(void*)block->page_zip.data, (void*)block);
+	} else {
 		*err = fil_io(OS_FILE_READ | wake_later,
 				sync, space,
 				offset, 0, UNIV_PAGE_SIZE,
 				(void*)block->frame, (void*)block);
+	}
 	ut_a(*err == DB_SUCCESS);
 
 	if (sync) {

dict/dict0dict.c

@@ -793,29 +793,20 @@ dict_table_get_and_increment_handle_count(
 }
 
 /**************************************************************************
-Adds a table object to the dictionary cache. */
+Adds system columns to a table object. */
 
 void
-dict_table_add_to_cache(
-/*====================*/
-	dict_table_t*	table)	/* in: table */
+dict_table_add_system_columns(
+/*==========================*/
+	dict_table_t*	table)	/* in/out: table */
 {
-	ulint	fold;
-	ulint	id_fold;
-	ulint	i;
-
 	ut_ad(table);
 #ifdef UNIV_SYNC_DEBUG
 	ut_ad(mutex_own(&(dict_sys->mutex)));
 #endif /* UNIV_SYNC_DEBUG */
 	ut_ad(table->n_def == table->n_cols - DATA_N_SYS_COLS);
 	ut_ad(table->magic_n == DICT_TABLE_MAGIC_N);
-	ut_ad(table->cached == FALSE);
-
-	fold = ut_fold_string(table->name);
-	id_fold = ut_fold_dulint(table->id);
-
-	table->cached = TRUE;
+	ut_ad(!table->cached);
 
 	/* NOTE: the system columns MUST be added in the following order
 	(so that they can be indexed by the numerical value of DATA_ROW_ID,
@@ -849,6 +840,26 @@ dict_table_add_to_cache(
 #if DATA_N_SYS_COLS != 4
 #error "DATA_N_SYS_COLS != 4"
 #endif
+}
+
+/**************************************************************************
+Adds a table object to the dictionary cache. */
+
+void
+dict_table_add_to_cache(
+/*====================*/
+	dict_table_t*	table)	/* in: table */
+{
+	ulint	fold;
+	ulint	id_fold;
+	ulint	i;
+
+	dict_table_add_system_columns(table);
+
+	table->cached = TRUE;
+
+	fold = ut_fold_string(table->name);
+	id_fold = ut_fold_dulint(table->id);
 
 	/* Look for a table with the same name: error if such exists */
 	{

fsp/fsp0fsp.c

@@ -890,7 +890,7 @@ fsp_header_init(
 
 	mtr_x_lock(fil_space_get_latch(space), mtr);
 
-	page = buf_page_create(space, 0, 0/* TODO: zip_size!=16k? */, mtr);
+	page = buf_page_create(space, 0, zip_size, mtr);
 	buf_page_get(space, 0, RW_X_LATCH, mtr);
 #ifdef UNIV_SYNC_DEBUG
 	buf_page_dbg_add_level(page, SYNC_FSP_PAGE);
@@ -1218,6 +1218,7 @@ fsp_fill_free_list(
 {
 	ulint	limit;
 	ulint	size;
+	ulint	zip_size;
 	xdes_t*	descr;
 	ulint	count		= 0;
 	ulint	frag_n_used;
@@ -1233,6 +1234,8 @@ fsp_fill_free_list(
 	size = mtr_read_ulint(header + FSP_SIZE, MLOG_4BYTES, mtr);
 	limit = mtr_read_ulint(header + FSP_FREE_LIMIT, MLOG_4BYTES, mtr);
 
+	zip_size = mach_read_from_4(FSP_PAGE_ZIP_SIZE + header);
+
 	if (space == 0 && srv_auto_extend_last_data_file
 			&& size < limit + FSP_EXTENT_SIZE * FSP_FREE_ADD) {
 
@@ -1272,8 +1275,8 @@ fsp_fill_free_list(
 			pages should be ignored. */
 
 			if (i > 0) {
-				/* TODO: zip_size != 16384 */
-				descr_page = buf_page_create(space, i, 0, mtr);
+				descr_page = buf_page_create(
+						space, i, zip_size, mtr);
 				buf_page_get(space, i, RW_X_LATCH, mtr);
 #ifdef UNIV_SYNC_DEBUG
 				buf_page_dbg_add_level(descr_page,
@@ -1291,10 +1294,9 @@ fsp_fill_free_list(
 
 			mtr_start(&ibuf_mtr);
 
-			/* TODO: no ibuf on compressed tablespaces */
 			ibuf_page = buf_page_create(space,
 					i + FSP_IBUF_BITMAP_OFFSET,
-					0, &ibuf_mtr);
+					zip_size, &ibuf_mtr);
 			buf_page_get(space, i + FSP_IBUF_BITMAP_OFFSET,
 							RW_X_LATCH, &ibuf_mtr);
 #ifdef UNIV_SYNC_DEBUG

include/dict0dict.h

@@ -163,6 +163,13 @@ dict_table_autoinc_update(
 	dict_table_t*	table,	/* in: table */
 	ib_longlong	value);	/* in: value which was assigned to a row */
 /**************************************************************************
+Adds system columns to a table object. */
+
+void
+dict_table_add_system_columns(
+/*==========================*/
+	dict_table_t*	table);	/* in/out: table */
+/**************************************************************************
 Adds a table object to the dictionary cache. */
 
 void

include/mtr0log.ic

@@ -182,9 +182,9 @@ mlog_write_initial_log_record_fast(
 	mtr->n_log_recs++;
 
 #ifdef UNIV_LOG_DEBUG
-/*	fprintf(stderr,
+	fprintf(stderr,
 		"Adding to mtr log record type %lu space %lu page no %lu\n",
-						type, space, offset); */
+						(ulong) type, space, offset);
 #endif
 
 #ifdef UNIV_DEBUG

mtr/mtr0log.c

@@ -534,6 +534,7 @@ mlog_parse_index(
 		n = mach_read_from_2(ptr);
 		ptr += 2;
 		n_uniq = mach_read_from_2(ptr);
+		ptr += 2;
 		ut_ad(n_uniq <= n);
 		if (end_ptr < ptr + (n + 1) * 2) {
 			return(NULL);
@@ -548,13 +549,13 @@ mlog_parse_index(
 	ind->table = table;
 	ind->n_uniq = n_uniq;
 	if (n_uniq != n) {
+		ut_a(n_uniq + DATA_ROLL_PTR <= n);
 		ind->type = DICT_CLUSTERED;
 	}
-	/* avoid ut_ad(index->cached) in dict_index_get_n_unique_in_tree */
-	ind->cached = TRUE;
 	if (comp) {
 		for (i = 0; i < n; i++) {
-			ulint	len = mach_read_from_2(ptr += 2);
+			ulint	len = mach_read_from_2(ptr);
+			ptr += 2;
 			/* The high-order bit of len is the NOT NULL flag;
 			the rest is 0 or 0x7fff for variable-length fields,
 			and 1..0x7ffe for fixed-length fields. */
@@ -567,8 +568,25 @@ mlog_parse_index(
 			dict_index_add_col(ind,
 				dict_table_get_nth_col(table, i), 0);
 		}
-		ptr += 2;
+		dict_table_add_system_columns(table);
+		if (n_uniq != n) {
+			/* Identify DB_TRX_ID and DB_ROLL_PTR in the index. */
+			ut_a(dtype_get_len(dict_col_get_type(
+				dict_field_get_col(dict_index_get_nth_field(
+				ind, n_uniq + (DATA_TRX_ID - 1)))))
+				== DATA_TRX_ID_LEN);
+			ut_a(dtype_get_len(dict_col_get_type(
+				dict_field_get_col(dict_index_get_nth_field(
+				ind, n_uniq + (DATA_ROLL_PTR - 1)))))
+				== DATA_ROLL_PTR_LEN);
+			dict_table_get_nth_col(table, i + DATA_TRX_ID)
+				->clust_pos = n_uniq + (DATA_TRX_ID - 1);
+			dict_table_get_nth_col(table, i + DATA_ROLL_PTR)
+				->clust_pos = n_uniq + (DATA_ROLL_PTR - 1);
+		}
 	}
+	/* avoid ut_ad(index->cached) in dict_index_get_n_unique_in_tree */
+	ind->cached = TRUE;
 	*index = ind;
 	return(ptr);
 }

os/os0file.c

@@ -3943,7 +3943,8 @@ consecutive_loop:
 
 	/* Do the i/o with ordinary, synchronous i/o functions: */
 	if (slot->type == OS_FILE_WRITE) {
-		if (array == os_aio_write_array /* TODO: && !page_zip */) {
+#if 0 /* TODO: && !page_zip */
+		if (array == os_aio_write_array) {
 			if ((total_len % UNIV_PAGE_SIZE != 0)
 				|| (slot->offset % UNIV_PAGE_SIZE != 0)) {
 				fprintf(stderr,
@@ -3955,13 +3956,14 @@ consecutive_loop:
 			}
 			os_file_check_page_trailers(combined_buf, total_len);
 		}
-
+#endif
 		ret = os_file_write(slot->name, slot->file, combined_buf,
 				slot->offset, slot->offset_high, total_len);
-
-		if (array == os_aio_write_array /* TODO: && !page_zip */) {
+#if 0 /* TODO: && !page_zip */
+		if (array == os_aio_write_array) {
 			os_file_check_page_trailers(combined_buf, total_len);
 		}
+#endif
 	} else {
 		ret = os_file_read(slot->file, combined_buf,
 				slot->offset, slot->offset_high, total_len);

page/page0zip.c

@@ -211,16 +211,14 @@ page_zip_compress_write_log(
 	Subtract 2 for the infimum and supremum records. */
 	trailer_size = page_dir_get_n_heap(page_zip->data) - 2;
 	/* Multiply by uncompressed of size stored per record */
-	if (page_is_leaf(page)) {
-		if (dict_index_is_clust(index)) {
+	if (!page_is_leaf(page)) {
+		trailer_size *= PAGE_ZIP_DIR_SLOT_SIZE + REC_NODE_PTR_SIZE;
+	} else if (dict_index_is_clust(index)) {
 		trailer_size *= PAGE_ZIP_DIR_SLOT_SIZE
 				+ DATA_TRX_ID_LEN + DATA_ROLL_PTR_LEN;
 	} else {
 		trailer_size *= PAGE_ZIP_DIR_SLOT_SIZE;
 	}
-	} else {
-		trailer_size *= PAGE_ZIP_DIR_SLOT_SIZE + REC_NODE_PTR_SIZE;
-	}
 	/* Add the space occupied by BLOB pointers. */
 	trailer_size += page_zip->n_blobs * BTR_EXTERN_FIELD_REF_SIZE;
 	ut_a(page_zip->m_end > PAGE_DATA);
@@ -2311,8 +2309,7 @@ page_zip_parse_write_blob_ptr(
 
 	if (UNIV_UNLIKELY(offset < PAGE_ZIP_START)
 			|| UNIV_UNLIKELY(offset >= UNIV_PAGE_SIZE)
-			|| UNIV_UNLIKELY(z_offset >= UNIV_PAGE_SIZE)
-			|| UNIV_UNLIKELY(!page_zip)) {
+			|| UNIV_UNLIKELY(z_offset >= UNIV_PAGE_SIZE)) {
 corrupt:
 		recv_sys->found_corrupt_log = TRUE;
 
@@ -2320,15 +2317,16 @@ corrupt:
 	}
 
 	if (page) {
-#if defined UNIV_DEBUG || defined UNIV_ZIP_DEBUG
-		ut_a(page_zip_validate(page_zip, page));
-#endif /* UNIV_DEBUG || UNIV_ZIP_DEBUG */
-
-		if (UNIV_UNLIKELY(!page_is_leaf(page))) {
+		if (UNIV_UNLIKELY(!page_zip)
+				|| UNIV_UNLIKELY(!page_is_leaf(page))) {
 
 			goto corrupt;
 		}
 
+#if defined UNIV_DEBUG || defined UNIV_ZIP_DEBUG
+		ut_a(page_zip_validate(page_zip, page));
+#endif /* UNIV_DEBUG || UNIV_ZIP_DEBUG */
+
 		memcpy(page + offset,
 				ptr + 4, BTR_EXTERN_FIELD_REF_SIZE);
 		memcpy(page_zip->data + z_offset,
@@ -2451,8 +2449,7 @@ page_zip_parse_write_node_ptr(
 
 	if (UNIV_UNLIKELY(offset < PAGE_ZIP_START)
 			|| UNIV_UNLIKELY(offset >= UNIV_PAGE_SIZE)
-			|| UNIV_UNLIKELY(z_offset >= UNIV_PAGE_SIZE)
-			|| UNIV_UNLIKELY(!page_zip)) {
+			|| UNIV_UNLIKELY(z_offset >= UNIV_PAGE_SIZE)) {
 corrupt:
 		recv_sys->found_corrupt_log = TRUE;
 
@@ -2465,15 +2462,16 @@ corrupt:
 		byte*	storage;
 		ulint	heap_no;
 
-#if defined UNIV_DEBUG || defined UNIV_ZIP_DEBUG
-		ut_a(page_zip_validate(page_zip, page));
-#endif /* UNIV_DEBUG || UNIV_ZIP_DEBUG */
-
-		if (UNIV_UNLIKELY(page_is_leaf(page))) {
+		if (UNIV_UNLIKELY(!page_zip)
+				|| UNIV_UNLIKELY(page_is_leaf(page))) {
 
 			goto corrupt;
 		}
 
+#if defined UNIV_DEBUG || defined UNIV_ZIP_DEBUG
+		ut_a(page_zip_validate(page_zip, page));
+#endif /* UNIV_DEBUG || UNIV_ZIP_DEBUG */
+
 		field = page + offset;
 		storage = page_zip->data + z_offset;
 
@@ -2558,7 +2556,8 @@ page_zip_write_node_ptr(
 		mach_write_to_2(log_ptr, storage - page_zip->data);
 		log_ptr += 2;
 		memcpy(log_ptr, field, REC_NODE_PTR_SIZE);
-		mlog_close(mtr, log_ptr + 6);
+		log_ptr += REC_NODE_PTR_SIZE;
+		mlog_close(mtr, log_ptr);
 	}
 }
 

row/row0sel.c

@@ -2369,7 +2369,7 @@ row_sel_store_row_id_to_prebuilt(
 	data = rec_get_nth_field(index_rec, offsets,
 			dict_index_get_sys_col_pos(index, DATA_ROW_ID), &len);
 
-	if (len != DATA_ROW_ID_LEN) {
+	if (UNIV_UNLIKELY(len != DATA_ROW_ID_LEN)) {
 		fprintf(stderr,
 "InnoDB: Error: Row id field is wrong length %lu in ", (ulong) len);
 		dict_index_name_print(stderr, prebuilt->trx, index);

trx/trx0sys.c

@@ -465,19 +465,32 @@ trx_sys_doublewrite_init_or_restore_pages(
 			do nothing */
 		} else {
 			ulint	zip_size;
-			/* Read in the actual page from the data files */
-
-			fil_io(OS_FILE_READ, TRUE, space_id, page_no, 0,
-					UNIV_PAGE_SIZE, read_buf, NULL);
-			/* Check if the page is corrupt */
+			ulint	zip_blk;
 
 			if (space_id) {
 				zip_size = fil_space_get_zip_size(space_id);
+				if (UNIV_LIKELY(!zip_size)) {
+					goto read_uncompressed;
+				}
+				zip_blk = UNIV_PAGE_SIZE / zip_size;
+				/* Read in the actual page from the file */
+				fil_io(OS_FILE_READ, TRUE, space_id,
+					page_no / zip_blk,
+					(page_no % zip_blk)
+					* zip_size, zip_size, read_buf, NULL);
 			} else {
+read_uncompressed:
 				zip_size = 0;
+				zip_blk = 1;
+				/* Read in the actual page from the file */
+				fil_io(OS_FILE_READ, TRUE, space_id, page_no,
+					0, UNIV_PAGE_SIZE, read_buf, NULL);
 			}
 
-			if (buf_page_is_corrupted(read_buf, zip_size)) {
+			/* Check if the page is corrupt */
+
+			if (UNIV_UNLIKELY(buf_page_is_corrupted(
+					read_buf, zip_size))) {
 
 				fprintf(stderr,
 		"InnoDB: Warning: database page corruption or a failed\n"
@@ -506,9 +519,17 @@ trx_sys_doublewrite_init_or_restore_pages(
 				doublewrite buffer to the intended
 				position */
 
+				if (zip_size) {
+					fil_io(OS_FILE_WRITE, TRUE, space_id,
+						page_no / zip_blk,
+						(page_no % zip_blk)
+						* zip_size, zip_size,
+						page, NULL);
+				} else {
 					fil_io(OS_FILE_WRITE, TRUE, space_id,
 						page_no, 0,
 						UNIV_PAGE_SIZE, page, NULL);
+				}
 				fprintf(stderr,
 		"InnoDB: Recovered the page from the doublewrite buffer.\n");
 			}