All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jason Wessel <jason.wessel@windriver.com>
To: qemu-devel@nongnu.org
Subject: [Qemu-devel] [PATCH 4/5] Block Memory boot devices
Date: Sat, 20 May 2006 16:02:03 -0500	[thread overview]
Message-ID: <446F83CB.5090704@windriver.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 292 bytes --]


This patch adds memory block devices to QEMU.  It was mainly for the 
purpose of booting a pentium qemu without a disk, but it is also very 
useful for testing and debugging different types of file systems from 
the OS using only memory. 

signed-off-by: jason.wessel@windriver.com

Jason.


[-- Attachment #2: block-mem_device.patch --]
[-- Type: text/plain, Size: 9921 bytes --]

Index: qemu/Makefile
===================================================================
--- qemu.orig/Makefile
+++ qemu/Makefile
@@ -29,7 +29,7 @@ subdir-%: dyngen$(EXESUF)
 
 recurse-all: $(patsubst %,subdir-%, $(TARGET_DIRS))
         
-qemu-img$(EXESUF): qemu-img.c block.c block-cow.c block-qcow.c aes.c block-vmdk.c block-cloop.c block-dmg.c block-bochs.c block-vpc.c block-vvfat.c
+qemu-img$(EXESUF): qemu-img.c block.c block-cow.c block-qcow.c aes.c block-vmdk.c block-cloop.c block-dmg.c block-bochs.c block-vpc.c block-vvfat.c block-mem.c
 	$(CC) -DQEMU_TOOL $(CFLAGS) $(LDFLAGS) $(DEFINES) -o $@ $^ -lz $(LIBS)
 
 dyngen$(EXESUF): dyngen.c
Index: qemu/Makefile.target
===================================================================
--- qemu.orig/Makefile.target
+++ qemu/Makefile.target
@@ -273,7 +273,7 @@ endif
 
 # must use static linking to avoid leaving stuff in virtual address space
 VL_OBJS=vl.o osdep.o block.o readline.o monitor.o pci.o console.o loader.o
-VL_OBJS+=block-cow.o block-qcow.o aes.o block-vmdk.o block-cloop.o block-dmg.o block-bochs.o block-vpc.o block-vvfat.o
+VL_OBJS+=block-cow.o block-qcow.o aes.o block-vmdk.o block-cloop.o block-dmg.o block-bochs.o block-vpc.o block-vvfat.o block-mem.o
 ifdef CONFIG_WIN32
 VL_OBJS+=tap-win32.o
 endif
Index: qemu/block-mem.c
===================================================================
--- /dev/null
+++ qemu/block-mem.c
@@ -0,0 +1,216 @@
+/*
+ * Block driver for volatile memory-based devices
+ *
+ * Copyright (c) 2006 Wind River Systems, Inc.
+ * Written by Alex deVries and Jason Wessel
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+#include "vl.h"
+#include "block_int.h"
+
+/**************************************************************/
+/* Memory scratch disk driver using host memory */
+
+/* mem_segment is a linked list of segments */
+
+struct mem_segment {
+    char name[256];
+    char * data;
+    unsigned int size;
+    struct mem_segment * next;
+} ;
+
+static struct mem_segment * segment_head = NULL;
+
+
+struct BDRVMemState {
+    struct mem_segment * segment;
+};
+
+
+/* Given a segment name, find the mem_segment (if it exists) */
+
+static struct mem_segment * find_segment(char * segment_name) {
+
+    struct mem_segment * p;
+
+    if (!segment_name) return NULL;
+
+    for (p = segment_head; p ; p=p->next) {
+        if (strcmp(p->name,segment_name)==0) return p;
+    }
+    return NULL;
+}
+
+/* Read the block device */
+
+static int mem_read(BlockDriverState *bs, int64_t sector_num,
+                    uint8_t *buf, int nb_sectors)
+{
+    struct BDRVMemState *s;
+    struct mem_segment * p ;
+
+    uint64_t len = nb_sectors * 512;
+    uint64_t offset = sector_num * 512;
+
+    if (!bs) return -1;
+    s = bs->opaque;
+
+    if (s && s->segment) return -1;
+    p = s->segment;
+
+    /* Check if it is out of bounds */
+    if (offset > p->size) return -1;
+
+    /* Make sure we don't read off the end */
+    if (len + offset > p->size) len = p->size - offset;
+
+    memcpy(buf,p->data,len);
+    return len;
+}
+
+/* Write to the block device */
+
+static int mem_write(BlockDriverState *bs, int64_t sector_num,
+                     const uint8_t *buf, int nb_sectors)
+{
+    struct BDRVMemState *s;
+    struct mem_segment * p ;
+
+    uint64_t len = nb_sectors * 512;
+    uint64_t offset = sector_num * 512;
+
+    if (!bs) return -1;
+    s = bs->opaque;
+
+    if (s && s->segment) return -1;
+    p = s->segment;
+
+    /* Check if it is out of bounds */
+    if (offset > p->size) return -1;
+
+    /* Make sure we don't read off the end */
+    if (len + offset > p->size) len = p->size - offset;
+
+    memcpy(p->data,buf,len);
+    return len;
+}
+
+/* Close the device */
+
+static void mem_close(BlockDriverState *bs)
+{
+    return;
+}
+
+static int mem_probe(const uint8_t *buf, int buf_size, const char *filename)
+{
+    if (!strncmp(filename, "mem:",4))
+        return 100;
+
+    return 0;
+}
+
+/* Create the segment.  Create a segment with the segment name of
+ * filename, and size of image_sectors * 512
+ */
+static int mem_create(const char *filename, int64_t image_sectors,
+                      const char *image_filename, int flags)
+{
+    struct mem_segment* p;
+
+    /* When image_sectors is zero, parse the sector size from
+     * the file name.
+     */
+    if (!filename || filename[0] == 0)
+    {
+        return -1;
+    }
+
+    if (image_sectors == 0) {
+        const char *s = filename;
+        if (strncmp(s,"mem:",4) != 0)
+            return -1;
+        s += 4;
+        s = strchr(s,':');
+        if (!s)
+            return -1;
+        s++;
+        image_sectors = atoi(s);
+        if (image_sectors == 0)
+            return -1;
+    }
+
+    /* Make sure the segment isn't already being used */
+
+    if (find_segment((char *) filename)) return -1;
+
+    /* Create memory for the segment */
+    p = qemu_malloc(sizeof(struct mem_segment));
+    if (!p) return -1;
+
+    /* Allocate the data area */
+    p->data = qemu_malloc(image_sectors * 512);
+    if (!p->data) return -1;
+
+    /* Setup the name */
+    snprintf(p->name,256,filename);
+    p->size=image_sectors * 512;
+
+    /* Tack it to the head */
+    p->next = segment_head;
+    segment_head=p;
+
+    return 0;
+
+}
+
+/* Open the block device */
+
+static int mem_open(BlockDriverState *bs, const char *filename)
+{
+
+    struct mem_segment * p = find_segment((char *) filename);
+
+
+    if (!p  || !p->data) {
+        /* If the segment did not exist, try to create it on the fly */
+        mem_create(filename, 0, 0, 0);
+        p = find_segment((char *) filename);
+        if (!p  || !p->data)
+            return -1;
+    }
+
+    bs->opaque = (void * ) p;
+
+    return 0;
+}
+
+BlockDriver bdrv_mem = {
+    "mem",
+    sizeof(struct BDRVMemState),
+    mem_probe,
+    mem_open,
+    mem_read,
+    mem_write,
+    mem_close,
+    mem_create,
+    0,
+};
Index: qemu/block.c
===================================================================
--- qemu.orig/block.c
+++ qemu/block.c
@@ -784,6 +784,7 @@ BlockDriver bdrv_raw = {
 void bdrv_init(void)
 {
     bdrv_register(&bdrv_raw);
+    bdrv_register(&bdrv_mem);
 #ifndef _WIN32
     bdrv_register(&bdrv_cow);
 #endif
Index: qemu/qemu-doc.texi
===================================================================
--- qemu.orig/qemu-doc.texi
+++ qemu/qemu-doc.texi
@@ -814,6 +814,7 @@ written), compressed and encrypted disk 
 * disk_images_snapshot_mode:: Snapshot mode
 * qemu_img_invocation::       qemu-img Invocation
 * disk_images_fat_images::    Virtual FAT disk images
+* disk_images_mem_images::    Memory only scratch disk
 @end menu
 
 @node disk_images_quickstart
@@ -878,6 +879,21 @@ What you should @emph{never} do:
 @item write to the FAT directory on the host system while accessing it with the guest system.
 @end itemize
 
+@node disk_images_mem_images
+@subsection Memory only scratch disk
+
+QEMU can create an in memory scratch disk device instead of using a disk image.  This is the default on a pentium system to satisfy the BIOS if you do not provide a disk image.  In the default case you only have a single 512byte sector which is the MBR of the scratch disk, but you can create a large scratch disk.  You can also choose to use memory scratch devices with any other target.  The syntax is as follows:
+
+@example
+qemu -kernel bzImage -hda mem:0:1
+@end example
+
+The scratch disks use the keyword "mem" followed by ":" then the memory device number "0", and finally the number of 512 byte sectors to allocate.  The above invocation is equivalent to when the -hda and its parameter are omitted for the pentium qemu.  Below is an example of creating a 1 megabyte scratch disk as the second IDE drive.
+
+@example
+qemu -kernel bzImage -hda hdisk.img -hdb mem:1:524288
+@end example
+
 @node pcsys_network
 @section Network emulation
 
Index: qemu/vl.c
===================================================================
--- qemu.orig/vl.c
+++ qemu/vl.c
@@ -5556,6 +5556,14 @@ int main(int argc, char **argv)
         bdrv_set_type_hint(bs_table[cdrom_index], BDRV_TYPE_CDROM);
     }
 
+#ifdef TARGET_I386
+    /* Allow a i386 target to boot with no "-hda" argument */
+    if (!hd_filename[0]) {
+        hd_filename[0] = strdup("mem:0:1");
+        boot_device = 'c';
+    }
+#endif /* TARGET_I386 */
+
     /* open the virtual block devices */
     for(i = 0; i < MAX_DISKS; i++) {
         if (hd_filename[i]) {
Index: qemu/vl.h
===================================================================
--- qemu.orig/vl.h
+++ qemu/vl.h
@@ -468,6 +468,7 @@ int cpu_load(QEMUFile *f, void *opaque, 
 typedef struct BlockDriverState BlockDriverState;
 typedef struct BlockDriver BlockDriver;
 
+extern BlockDriver bdrv_mem;
 extern BlockDriver bdrv_raw;
 extern BlockDriver bdrv_cow;
 extern BlockDriver bdrv_qcow;

                 reply	other threads:[~2006-05-20 21:02 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=446F83CB.5090704@windriver.com \
    --to=jason.wessel@windriver.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.