Skip to contentSkip to conversation

hw/nvram/fw_cfg: fix FWCfgDataGeneratorClass::get_data() consumption

PATCH Applied Laszlo Ersek 3 participants 4 messages Sep 16, 2020 – Sep 18, 2020

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 hw/nvram/fw_cfg: fix FWCfgDataGeneratorClass::get_data() consumption

The documentation on g_byte_array_free()
<https://developer.gnome.org/glib/stable/glib-Byte-Arrays.html#g-byte-array-free>
says:

Quoted 4 lines

Returns

the element data if free_segment is FALSE, otherwise NULL. The element data should be freed using g_free().

Because we currently call g_byte_array_free() with free_segment=TRUE, we end up passing data=NULL to fw_cfg_add_file().

On the plus side, fw_cfg_data_read() and fw_cfg_dma_transfer() both deal with NULL data gracefully: QEMU does not crash when the guest reads such an item, the guest just gets a properly sized, but zero-filled blob.

However, the bug breaks UEFI HTTPS boot, as the IANA_TLS_CIPHER array, generated otherwise correctly by the "tls-cipher-suites" object, is in effect replaced with a zero blob.

Fix the issue by passing free_segment=FALSE to g_byte_array_free():

- the caller (fw_cfg_add_from_generator()) temporarily assumes ownership of the generated byte array,

- then ownership of the byte array is transfered to fw_cfg, as fw_cfg_add_file() links (not copies) "data" into fw_cfg.

Cc "Daniel P. Berrangé" <[email protected]>
Cc "Philippe Mathieu-Daudé" <[email protected]>
Cc Gerd Hoffmann <[email protected]>
Fixes 3203148917d035b09f71986ac2eaa19a352d6d9d
---
 hw/nvram/fw_cfg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/nvram/fw_cfg.c +1 -1
index f3a4728288eb..0e95d057fd51 100644
@@ -1056,7 +1056,7 @@ bool fw_cfg_add_from_generator(FWCfgState *s, const char *filename,
         return false;
     }
     size = array->len;
-    fw_cfg_add_file(s, filename, g_byte_array_free(array, TRUE), size);
+    fw_cfg_add_file(s, filename, g_byte_array_free(array, FALSE), size);
 
     return true;
 }
Signature · 1 line
2.19.1.3.g30247aa5d201
Reviewed-by on v1 1/1
#1 lore raw
Quoted 11 lines · Laszlo Ersek

The documentation on g_byte_array_free()
<https://developer.gnome.org/glib/stable/glib-Byte-Arrays.html#g-byte-array-free>
says:

Quoted 4 lines

Returns

the element data if free_segment is FALSE, otherwise NULL. The element data should be freed using g_free().

Because we currently call g_byte_array_free() with free_segment=TRUE, we end up passing data=NULL to fw_cfg_add_file().

Mea culpa...

Quoted 9 lines

On the plus side, fw_cfg_data_read() and fw_cfg_dma_transfer() both deal with NULL data gracefully: QEMU does not crash when the guest reads such an item, the guest just gets a properly sized, but zero-filled blob.

However, the bug breaks UEFI HTTPS boot, as the IANA_TLS_CIPHER array, generated otherwise correctly by the "tls-cipher-suites" object, is in effect replaced with a zero blob.

Quoted 7 lines

Fix the issue by passing free_segment=FALSE to g_byte_array_free():

- the caller (fw_cfg_add_from_generator()) temporarily assumes ownership of the generated byte array,

- then ownership of the byte array is transfered to fw_cfg, as fw_cfg_add_file() links (not copies) "data" into fw_cfg.

Thanks!

Quoted 24 lines
Cc "Daniel P. Berrangé" <[email protected]>
Cc "Philippe Mathieu-Daudé" <[email protected]>
Cc Gerd Hoffmann <[email protected]>
Fixes 3203148917d035b09f71986ac2eaa19a352d6d9d
---
 hw/nvram/fw_cfg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/nvram/fw_cfg.c +1 -1
index f3a4728288eb..0e95d057fd51 100644
@@ -1056,7 +1056,7 @@ bool fw_cfg_add_from_generator(FWCfgState *s, const char *filename,
         return false;
     }
     size = array->len;
-    fw_cfg_add_file(s, filename, g_byte_array_free(array, TRUE), size);
+    fw_cfg_add_file(s, filename, g_byte_array_free(array, FALSE), size);
 
     return true;
 }
Reviewed-by on v1 1/1
#2 lore raw
In reply to Laszlo Ersek
Quoted 28 lines · Laszlo Ersek

The documentation on g_byte_array_free()
<https://developer.gnome.org/glib/stable/glib-Byte-Arrays.html#g-byte-array-free>
says:

Quoted 4 lines

Returns

the element data if free_segment is FALSE, otherwise NULL. The element data should be freed using g_free().

Because we currently call g_byte_array_free() with free_segment=TRUE, we end up passing data=NULL to fw_cfg_add_file().

On the plus side, fw_cfg_data_read() and fw_cfg_dma_transfer() both deal with NULL data gracefully: QEMU does not crash when the guest reads such an item, the guest just gets a properly sized, but zero-filled blob.

However, the bug breaks UEFI HTTPS boot, as the IANA_TLS_CIPHER array, generated otherwise correctly by the "tls-cipher-suites" object, is in effect replaced with a zero blob.

Fix the issue by passing free_segment=FALSE to g_byte_array_free():

- the caller (fw_cfg_add_from_generator()) temporarily assumes ownership of the generated byte array,

- then ownership of the byte array is transfered to fw_cfg, as fw_cfg_add_file() links (not copies) "data" into fw_cfg.

Quoted 8 lines
Cc "Daniel P. Berrangé" <[email protected]>
Cc "Philippe Mathieu-Daudé" <[email protected]>
Cc Gerd Hoffmann <[email protected]>
Fixes 3203148917d035b09f71986ac2eaa19a352d6d9d
---
 hw/nvram/fw_cfg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Regards,
Daniel

Signature · 3 lines
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|
#3 lore raw
In reply to Laszlo Ersek
Quoted 43 lines · Laszlo Ersek

The documentation on g_byte_array_free()
<https://developer.gnome.org/glib/stable/glib-Byte-Arrays.html#g-byte-array-free>
says:

Quoted 4 lines

Returns

the element data if free_segment is FALSE, otherwise NULL. The element data should be freed using g_free().

Because we currently call g_byte_array_free() with free_segment=TRUE, we end up passing data=NULL to fw_cfg_add_file().

On the plus side, fw_cfg_data_read() and fw_cfg_dma_transfer() both deal with NULL data gracefully: QEMU does not crash when the guest reads such an item, the guest just gets a properly sized, but zero-filled blob.

However, the bug breaks UEFI HTTPS boot, as the IANA_TLS_CIPHER array, generated otherwise correctly by the "tls-cipher-suites" object, is in effect replaced with a zero blob.

Fix the issue by passing free_segment=FALSE to g_byte_array_free():

- the caller (fw_cfg_add_from_generator()) temporarily assumes ownership of the generated byte array,

- then ownership of the byte array is transfered to fw_cfg, as fw_cfg_add_file() links (not copies) "data" into fw_cfg.

Cc "Daniel P. Berrangé" <[email protected]>
Cc "Philippe Mathieu-Daudé" <[email protected]>
Cc Gerd Hoffmann <[email protected]>
Fixes 3203148917d035b09f71986ac2eaa19a352d6d9d
---
 hw/nvram/fw_cfg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/nvram/fw_cfg.c +0 -0
index f3a4728288eb..0e95d057fd51 100644

hw/nvram/fw_cfg.c bool fw_cfg_add_from_generator(FWCfgState *s, const char *filename, · line 1062

Quoted 8 lines
hw/nvram/fw_cfg.c +1 -1
@@ -1056,7 +1056,7 @@ bool fw_cfg_add_from_generator(FWCfgState *s, const char *filename,
         return false;
     }
     size = array->len;
-    fw_cfg_add_file(s, filename, g_byte_array_free(array, TRUE), size);
+    fw_cfg_add_file(s, filename, g_byte_array_free(array, FALSE), size);
 
     return true;
 }

Thanks, applied to my fw_cfg tree.