[Qemu-devel] [PATCH v2] block: handle "rechs" and "large" translation options

[Qemu-devel] [PATCH v2] block: handle "rechs" and "large" translation options

[Paolo Bonzini]

Sure, CHS translation is an obscure topic, and legacy options for
hard-disk geometries are obscure as well.  But since QEMU does nothing
with it except telling the BIOS, and since there "large" and "rechs"
are listed in the enums, parsing them seems to be the bare minimum.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 blockdev.c |  4 ++++
 vl.c       | 22 ++++++++++++++++------
 2 files changed, 20 insertions(+), 6 deletions(-)

diff --git a/blockdev.c b/blockdev.c
index 36ceece..d71f815 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -776,6 +776,10 @@ DriveInfo *drive_init(QemuOpts *all_opts, BlockInterfaceType block_default_type)
             translation = BIOS_ATA_TRANSLATION_NONE;
         } else if (!strcmp(value, "lba")) {
             translation = BIOS_ATA_TRANSLATION_LBA;
+        } else if (!strcmp(value, "large")) {
+            translation = BIOS_ATA_TRANSLATION_LARGE;
+        } else if (!strcmp(value, "rechs")) {
+            translation = BIOS_ATA_TRANSLATION_RECHS;
         } else if (!strcmp(value, "auto")) {
             translation = BIOS_ATA_TRANSLATION_AUTO;
         } else {
diff --git a/vl.c b/vl.c
index 7f4fe0d..46112af 100644
--- a/vl.c
+++ b/vl.c
@@ -3053,14 +3053,19 @@ int main(int argc, char **argv, char **envp)
                         goto chs_fail;
                     if (*p == ',') {
                         p++;
-                        if (!strcmp(p, "none"))
+                        if (!strcmp(p, "large")) {
+                            translation = BIOS_ATA_TRANSLATION_LARGE;
+                        } else if (!strcmp(p, "rechs")) {
+                            translation = BIOS_ATA_TRANSLATION_RECHS;
+                        } else if (!strcmp(p, "none")) {
                             translation = BIOS_ATA_TRANSLATION_NONE;
-                        else if (!strcmp(p, "lba"))
+                        } else if (!strcmp(p, "lba")) {
                             translation = BIOS_ATA_TRANSLATION_LBA;
-                        else if (!strcmp(p, "auto"))
+                        } else if (!strcmp(p, "auto")) {
                             translation = BIOS_ATA_TRANSLATION_AUTO;
-                        else
+                        } else {
                             goto chs_fail;
+                        }
                     } else if (*p != '\0') {
                     chs_fail:
                         fprintf(stderr, "qemu: invalid physical CHS format\n");
@@ -3074,10 +3079,15 @@ int main(int argc, char **argv, char **envp)
                         qemu_opt_set(hda_opts, "heads", num);
                         snprintf(num, sizeof(num), "%d", secs);
                         qemu_opt_set(hda_opts, "secs", num);
-                        if (translation == BIOS_ATA_TRANSLATION_LBA)
+                        if (translation == BIOS_ATA_TRANSLATION_LARGE) {
+                            qemu_opt_set(hda_opts, "trans", "large");
+                        } else if (translation == BIOS_ATA_TRANSLATION_RECHS) {
+                            qemu_opt_set(hda_opts, "trans", "rechs");
+                        } else if (translation == BIOS_ATA_TRANSLATION_LBA) {
                             qemu_opt_set(hda_opts, "trans", "lba");
-                        if (translation == BIOS_ATA_TRANSLATION_NONE)
+                        } else if (translation == BIOS_ATA_TRANSLATION_NONE) {
                             qemu_opt_set(hda_opts, "trans", "none");
+                        }
                     }
                 }
                 break;
-- 
1.8.4.2

Re: [Qemu-devel] [PATCH v2] block: handle "rechs" and "large" translation options

[Markus Armbruster]

Paolo Bonzini <pbonzini@redhat.com> writes:

> Sure, CHS translation is an obscure topic, and legacy options for
> hard-disk geometries are obscure as well.  But since QEMU does nothing
> with it except telling the BIOS, and since there "large" and "rechs"
> are listed in the enums, parsing them seems to be the bare minimum.
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

This is semantically on top of "[PATCH 00/12] qdev: cleanup legacy
properties", which adds these two values to isa-hd property
bios-chs-trans.  Here, you add them to the two obsolete ways to specify
the property.  While I don't think that's necessary, I don't object to
it either.

Despite my doubts on the need for the patch:

Reviewed-by: Markus Armbruster <armbru@redhat.com>

Re: [Qemu-devel] [PATCH v2] block: handle "rechs" and "large" translation options

[Stefan Hajnoczi]

On Thu, Jan 30, 2014 at 04:57:25PM +0100, Paolo Bonzini wrote:
> Sure, CHS translation is an obscure topic, and legacy options for
> hard-disk geometries are obscure as well.  But since QEMU does nothing
> with it except telling the BIOS, and since there "large" and "rechs"
> are listed in the enums, parsing them seems to be the bare minimum.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  blockdev.c |  4 ++++
>  vl.c       | 22 ++++++++++++++++------
>  2 files changed, 20 insertions(+), 6 deletions(-)

Acked-by: Stefan Hajnoczi <stefanha@gmail.com>

Feel free to squash into your other series.

(Please consider doing a separate patch with the documentation updates I
mentioned in response to v1.)