patch-2.1.66 linux/drivers/char/ftape/zftape/zftape-read.c

Next file: linux/drivers/char/ftape/zftape/zftape-read.h
Previous file: linux/drivers/char/ftape/zftape/zftape-init.h
Back to the patch index
Back to the overall index

diff -u --recursive --new-file v2.1.65/linux/drivers/char/ftape/zftape/zftape-read.c linux/drivers/char/ftape/zftape/zftape-read.c
@@ -0,0 +1,390 @@
+/*
+ *      Copyright (C) 1996, 1997 Claus-Justus Heine
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; see the file COPYING.  If not, write to
+ the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+
+ *
+ * $Source: /homes/cvs/ftape-stacked/ftape/zftape/zftape-read.c,v $
+ * $Revision: 1.2 $
+ * $Date: 1997/10/05 19:19:06 $
+ *
+ *      This file contains the high level reading code
+ *      for the QIC-117 floppy-tape driver for Linux.
+ */
+
+#include <linux/config.h>
+#include <linux/errno.h>
+#include <linux/mm.h>
+#ifdef CONFIG_KERNELD
+#include <linux/kerneld.h>
+#endif
+
+#include <linux/zftape.h>
+
+#if LINUX_VERSION_CODE >= KERNEL_VER(2,1,6)
+#include <asm/uaccess.h>
+#else
+#include <asm/segment.h>
+#endif
+
+#include "../zftape/zftape-init.h"
+#include "../zftape/zftape-eof.h"
+#include "../zftape/zftape-ctl.h"
+#include "../zftape/zftape-write.h"
+#include "../zftape/zftape-read.h"
+#include "../zftape/zftape-rw.h"
+#include "../zftape/zftape-vtbl.h"
+
+/*      Global vars.
+ */
+int zft_just_before_eof = 0;
+
+/*      Local vars.
+ */
+static int buf_len_rd = 0;
+
+void zft_zap_read_buffers(void)
+{
+	buf_len_rd = 0;
+}
+
+int zft_read_header_segments(void)      
+{
+	TRACE_FUN(ft_t_flow);
+
+	zft_header_read = 0;
+	TRACE_CATCH(zft_vmalloc_once(&zft_hseg_buf, FT_SEGMENT_SIZE),);
+	TRACE_CATCH(ftape_read_header_segment(zft_hseg_buf),);
+	TRACE(ft_t_info, "Segments written since first format: %d",
+	      (int)GET4(zft_hseg_buf, FT_SEG_CNT));
+	zft_qic113 = (ft_format_code != fmt_normal &&
+		      ft_format_code != fmt_1100ft &&
+		      ft_format_code != fmt_425ft);
+	TRACE(ft_t_info, "ft_first_data_segment: %d, ft_last_data_segment: %d", 
+	      ft_first_data_segment, ft_last_data_segment);
+	zft_capacity = zft_get_capacity();
+	zft_old_ftape = zft_ftape_validate_label(&zft_hseg_buf[FT_LABEL]);
+	if (zft_old_ftape) {
+		TRACE(ft_t_info, 
+"Found old ftaped tape, emulating eof marks, entering read-only mode");
+		zft_ftape_extract_file_marks(zft_hseg_buf);
+		TRACE_CATCH(zft_fake_volume_headers(zft_eof_map, 
+						    zft_nr_eof_marks),);
+	} else {
+		/* the specs say that the volume table must be
+		 * initialized with zeroes during formatting, so it
+		 * MUST be readable, i.e. contain vaid ECC
+		 * information.  
+		 */
+		TRACE_CATCH(ftape_read_segment(ft_first_data_segment, 
+					       zft_deblock_buf, 
+					       FT_RD_SINGLE),);
+		TRACE_CATCH(zft_extract_volume_headers(zft_deblock_buf),);
+	}
+	zft_header_read = 1;
+	zft_set_flags(zft_unit);
+	zft_reset_position(&zft_pos);
+	TRACE_EXIT 0;
+}
+
+int zft_fetch_segment_fraction(const unsigned int segment, void *buffer,
+			       const ft_read_mode_t read_mode,
+			       const unsigned int start,
+			       const unsigned int size)
+{
+	int seg_sz;
+	TRACE_FUN(ft_t_flow);
+
+	if (segment == zft_deblock_segment) {
+		TRACE(ft_t_data_flow,
+		      "re-using segment %d already in deblock buffer",
+		      segment);
+		seg_sz = zft_get_seg_sz(segment);
+		if (start > seg_sz) {
+			TRACE_ABORT(-EINVAL, ft_t_bug,
+				    "trying to read beyond end of segment:\n"
+				    KERN_INFO "seg_sz : %d\n"
+				    KERN_INFO "start  : %d\n"
+				    KERN_INFO "segment: %d",
+				    seg_sz, start, segment);
+		}
+		if ((start + size) > seg_sz) {
+			TRACE_EXIT seg_sz - start;
+		}
+		TRACE_EXIT size;
+	}
+	seg_sz = ftape_read_segment_fraction(segment, buffer, read_mode,
+					     start, size);
+	TRACE(ft_t_data_flow, "segment %d, result %d", segment, seg_sz);
+	if ((int)seg_sz >= 0 && start == 0 && size == FT_SEGMENT_SIZE) {
+		/*  this implicitly assumes that we are always called with
+		 *  buffer == zft_deblock_buf 
+		 */
+		zft_deblock_segment = segment;
+	} else {
+		zft_deblock_segment = -1;
+	}
+	TRACE_EXIT seg_sz;
+}
+
+/*
+ * out:
+ *
+ * int *read_cnt: the number of bytes we removed from the
+ *                zft_deblock_buf (result)
+ *
+ * int *to_do   : the remaining size of the read-request. Is changed.
+ *
+ * in:
+ *
+ * char *buff      : buff is the address of the upper part of the user
+ *                   buffer, that hasn't been filled with data yet.
+ * int buf_pos_read: copy of buf_pos_rd
+ * int buf_len_read: copy of buf_len_rd
+ * char *zft_deblock_buf: ftape_zft_deblock_buf
+ *
+ * returns the amount of data actually copied to the user-buffer
+ *
+ * to_do MUST NOT SHRINK except to indicate an EOT. In this case to_do
+ * has to be set to 0. We cannot return -ENOSPC, because we return the
+ * amount of data actually * copied to the user-buffer
+ */
+static int zft_simple_read (int *read_cnt, 
+			    __u8  *dst_buf, 
+			    const int to_do, 
+			    const __u8 *src_buf, 
+			    const int seg_sz, 
+			    const zft_position *pos,
+			    const zft_volinfo *volume)
+{
+	TRACE_FUN(ft_t_flow);
+
+	if (seg_sz - pos->seg_byte_pos < to_do) {
+		*read_cnt = seg_sz - pos->seg_byte_pos;
+	} else {
+		*read_cnt = to_do;
+	}
+#if LINUX_VERSION_CODE > KERNEL_VER(2,1,3)
+	if (copy_to_user(dst_buf, 
+			 src_buf + pos->seg_byte_pos, *read_cnt) != 0) {
+		TRACE_EXIT -EFAULT;
+	}
+#else
+	TRACE_CATCH(verify_area(VERIFY_WRITE, dst_buf, *read_cnt),);
+	memcpy_tofs(dst_buf, src_buf +  pos->seg_byte_pos, *read_cnt);
+#endif
+	TRACE(ft_t_noise, "nr bytes just read: %d", *read_cnt);
+	TRACE_EXIT *read_cnt;
+}
+
+/* req_len: gets clipped due to EOT of EOF.
+ * req_clipped: is a flag indicating whether req_len was clipped or not
+ * volume: contains information on current volume (blk_sz etc.)
+ */
+static int check_read_access(int *req_len, 
+			     const zft_volinfo **volume,
+			     int *req_clipped, 
+			     const zft_position *pos)
+{
+	static __s64 remaining = 0;
+	static int eod;
+	TRACE_FUN(ft_t_flow);
+	
+	if (zft_io_state != zft_reading) {
+		if (zft_offline) { /* offline includes no_tape */
+			TRACE_ABORT(-ENXIO, ft_t_warn,
+				    "tape is offline or no cartridge");
+		}
+		if (!ft_formatted) {
+			TRACE_ABORT(-EACCES,
+				    ft_t_warn, "tape is not formatted");
+		}
+		/*  now enter defined state, read header segment if not
+		 *  already done and flush write buffers
+		 */
+		TRACE_CATCH(zft_def_idle_state(),);
+		zft_io_state = zft_reading;
+		if (zft_tape_at_eod(pos)) {
+			eod = 1;
+			TRACE_EXIT 1;
+		}
+		eod = 0;
+		*volume = zft_find_volume(pos->seg_pos);
+		/* get the space left until EOF */
+		remaining = zft_check_for_eof(*volume, pos);
+		buf_len_rd = 0;
+		TRACE(ft_t_noise, "remaining: " LL_X ", vol_no: %d",
+		      LL(remaining), (*volume)->count);
+	} else if (zft_tape_at_eod(pos)) {
+		if (++eod > 2) {
+			TRACE_EXIT -EIO; /* st.c also returns -EIO */
+		} else {
+			TRACE_EXIT 1;
+		}
+	}
+	if ((*req_len % (*volume)->blk_sz) != 0) {
+		/*  this message is informational only. The user gets the
+		 *  proper return value
+		 */
+		TRACE_ABORT(-EINVAL, ft_t_info,
+			    "req_len %d not a multiple of block size %d",
+			    *req_len, (*volume)->blk_sz);
+	}
+	/* As GNU tar doesn't accept partial read counts when the
+	 * multiple volume flag is set, we make sure to return the
+	 * requested amount of data. Except, of course, at the end of
+	 * the tape or file mark.  
+	 */
+	remaining -= *req_len;
+	if (remaining <= 0) {
+		TRACE(ft_t_noise, 
+		      "clipped request from %d to %d.", 
+		      *req_len, (int)(*req_len + remaining));
+		*req_len += remaining;
+		*req_clipped = 1;
+	} else {
+		*req_clipped = 0;
+	}
+	TRACE_EXIT 0;
+}
+
+/* this_segs_size: the current segment's size.
+ * buff: the USER-SPACE buffer provided by the calling function.
+ * req_len: how much data should be read at most.
+ * volume: contains information on current volume (blk_sz etc.)
+ */  
+static int empty_deblock_buf(__u8 *usr_buf, const int req_len,
+			     const __u8 *src_buf, const int seg_sz,
+			     zft_position *pos,
+			     const zft_volinfo *volume)
+{
+	int cnt;
+	int result = 0;
+	TRACE_FUN(ft_t_flow);
+
+	TRACE(ft_t_data_flow, "this_segs_size: %d", seg_sz);
+	if (zft_use_compression && volume->use_compression) {
+		TRACE_CATCH(zft_cmpr_lock(1 /* try to load */),);
+		TRACE_CATCH(result= (*zft_cmpr_ops->read)(&cnt,
+							  usr_buf, req_len,
+							  src_buf, seg_sz,
+							  pos, volume),);
+	} else {                                  
+		TRACE_CATCH(result= zft_simple_read (&cnt,
+						     usr_buf, req_len,
+						     src_buf, seg_sz,
+						     pos, volume),);
+	}
+	pos->volume_pos   += result;
+        pos->tape_pos     += cnt;
+	pos->seg_byte_pos += cnt;
+	buf_len_rd        -= cnt; /* remaining bytes in buffer */
+	TRACE(ft_t_data_flow, "buf_len_rd: %d, cnt: %d", buf_len_rd, cnt);
+	if(pos->seg_byte_pos >= seg_sz) {
+		pos->seg_pos++;
+		pos->seg_byte_pos = 0;
+	}
+	TRACE(ft_t_data_flow, "bytes moved out of deblock-buffer: %d", cnt);
+	TRACE_EXIT result;
+}
+
+
+/* note: we store the segment id of the segment that is inside the
+ * deblock buffer. This spares a lot of ftape_read_segment()s when we
+ * use small block-sizes. The block-size may be 1kb (SECTOR_SIZE). In
+ * this case a MTFSR 28 maybe still inside the same segment.
+ */
+int _zft_read(char* buff, int req_len)
+{
+	int req_clipped;
+	int result     = 0;
+	int bytes_read = 0;
+	static unsigned int seg_sz = 0;
+	static const zft_volinfo *volume = NULL;
+	TRACE_FUN(ft_t_flow);
+	
+	zft_resid = req_len;
+	result = check_read_access(&req_len, &volume,
+				   &req_clipped, &zft_pos);
+	switch(result) {
+	case 0: 
+		break; /* nothing special */
+	case 1: 
+		TRACE(ft_t_noise, "EOD reached");
+		TRACE_EXIT 0;   /* EOD */
+	default:
+		TRACE_ABORT(result, ft_t_noise,
+			    "check_read_access() failed with result %d",
+			    result);
+		TRACE_EXIT result;
+	}
+	while (req_len > 0) { 
+		/*  Allow escape from this loop on signal !
+		 */
+		FT_SIGNAL_EXIT(_DONT_BLOCK);
+		/* buf_len_rd == 0 means that we need to read a new
+		 * segment.
+		 */
+		if (buf_len_rd == 0) {
+			while((result = zft_fetch_segment(zft_pos.seg_pos,
+							  zft_deblock_buf,
+							  FT_RD_AHEAD)) == 0) {
+				zft_pos.seg_pos ++;
+				zft_pos.seg_byte_pos = 0;
+			}
+			if (result < 0) {
+				zft_resid -= bytes_read;
+				TRACE_ABORT(result, ft_t_noise,
+					    "zft_fetch_segment(): %d",
+					    result);
+			}
+			seg_sz = result;
+			buf_len_rd = seg_sz - zft_pos.seg_byte_pos;
+		}
+		TRACE_CATCH(result = empty_deblock_buf(buff, 
+						       req_len,
+						       zft_deblock_buf, 
+						       seg_sz, 
+						       &zft_pos,
+						       volume),
+			    zft_resid -= bytes_read);
+		TRACE(ft_t_data_flow, "bytes just read: %d", result);
+		bytes_read += result; /* what we got so far       */
+		buff       += result; /* index in user-buffer     */
+		req_len    -= result; /* what's left from req_len */
+	} /* while (req_len  > 0) */
+	if (req_clipped) {
+		TRACE(ft_t_data_flow,
+		      "maybe partial count because of eof mark");
+		if (zft_just_before_eof && bytes_read == 0) {
+			/* req_len was > 0, but user didn't get
+			 * anything the user has read in the eof-mark 
+			 */
+			zft_move_past_eof(&zft_pos);
+			ftape_abort_operation();
+		} else {
+			/* don't skip to the next file before the user
+			 * tried to read a second time past EOF Just
+			 * mark that we are at EOF and maybe decrement
+			 * zft_seg_pos to stay in the same volume;
+			 */
+			zft_just_before_eof = 1;
+			zft_position_before_eof(&zft_pos, volume);
+			TRACE(ft_t_noise, "just before eof");
+		}
+	}
+	zft_resid -= result; /* for MTSTATUS       */
+	TRACE_EXIT bytes_read;
+}

FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen, slshen@lbl.gov