[Qemu-devel] [PULL 01/20] fdc: fix detection under Linux

From: John Snow <jsnow@redhat.com>
To: qemu-devel@nongnu.org
Cc: peter.maydell@linaro.org, jsnow@redhat.com
Subject: [Qemu-devel] [PULL 01/20] fdc: fix detection under Linux
Date: Wed,  3 Feb 2016 15:32:16 -0500	[thread overview]
Message-ID: <1454531555-32022-2-git-send-email-jsnow@redhat.com> (raw)
In-Reply-To: <1454531555-32022-1-git-send-email-jsnow@redhat.com>

Accidentally, I removed a "feature" where empty drives had geometry
values applied to them, which allows seek on empty drives to work
"by accident," as QEMU actually tries to disallow that.

Seeks on empty drives should work, though, but the easiest thing is to
restore the misfeature where empty drives have non-zero geometries
applied.

Document the hack accordingly.

[Maintainer edit]

This fix corrects a regression introduced in d5d47efc, where
pick_geometry was modified such that it would not operate on empty
drives, and as a result if there is no diskette inserted, QEMU
no longer populates it with geometry bounds. As a result, seek fails
when QEMU denies to move the current track, but reports success anyway.
This can confuse the guest, leading to kernel panics in the guest.


Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1454106932-17236-1-git-send-email-jsnow@redhat.com
---
 hw/block/fdc.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/hw/block/fdc.c b/hw/block/fdc.c
index 818e8a4..9ef899d 100644
--- a/hw/block/fdc.c
+++ b/hw/block/fdc.c
@@ -179,6 +179,21 @@ typedef struct FDrive {
 
 static FloppyDriveType get_fallback_drive_type(FDrive *drv);
 
+/* Hack: FD_SEEK is expected to work on empty drives. However, QEMU
+ * currently goes through some pains to keep seeks within the bounds
+ * established by last_sect and max_track. Correcting this is difficult,
+ * as refactoring FDC code tends to expose nasty bugs in the Linux kernel.
+ *
+ * For now: allow empty drives to have large bounds so we can seek around,
+ * with the understanding that when a diskette is inserted, the bounds will
+ * properly tighten to match the geometry of that inserted medium.
+ */
+static void fd_empty_seek_hack(FDrive *drv)
+{
+    drv->last_sect = 0xFF;
+    drv->max_track = 0xFF;
+}
+
 static void fd_init(FDrive *drv)
 {
     /* Drive */
@@ -394,6 +409,7 @@ static void fd_revalidate(FDrive *drv)
         if (!blk_is_inserted(drv->blk)) {
             FLOPPY_DPRINTF("No disk in drive\n");
             drv->disk = FLOPPY_DRIVE_TYPE_NONE;
+            fd_empty_seek_hack(drv);
         } else if (!drv->media_validated) {
             rc = pick_geometry(drv);
             if (rc) {
-- 
2.4.3
