Sometimes you know only know which debug logging conditionals you want to
turn off. Now debug log messages can be excluded by specifying the
conditional in the $debug string with a prepended dash. So say you want
all debug logging on except for btrfs and scripting, then set
debug=all,-btrfs,-scri
This makes grub_strword more versatile. All code as of this change using
grub_strword uses it in a conditional or boolean operator context where its
return value would be silently converted to a 1 or 0, so this does not
affect current usage. A few lines in loader/xnu.c have been changed to
enhance
When debugging using the "all" condition frequently gives way too much info.
So selecting a subset of conditions is desirable. This change makes it
easier to identify which conditions are desired. Before this change log
messages needed to be looked up in the source code to find the condition
used.
These are pretty self-explanatory. They've been useful in debugging grub
issues. Note that patch #2 and #3 are particularly synergistic.
Also a note about patch #4. It was used in a version of patch #3, though not
any more. I decided to leave it in this patch series because it enhances
grub_strwor
Signed-off-by: Glenn Washburn
---
grub-core/kern/err.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/grub-core/kern/err.c b/grub-core/kern/err.c
index 53c734de7..a12fda48b 100644
--- a/grub-core/kern/err.c
+++ b/grub-core/kern/err.c
@@ -107,7 +107,7 @@ grub_print_error (void
grub_dl_unload_unneeded has been unused by default since de04eecfa9 put
its only call within an "#if 0", so there's no point in it sitting
around taking up space in the core image. Turn the "#if 0" into "#ifdef
GRUB_UNLOAD_UNNEEDED_MODULES" just in case somebody is turning this code
back on locall
This function is only used by gfxmenu. Moving it there is unfortunately
a bit messy, but it frees up a fairly substantial amount (in relative
terms) of precious core image space on i386-pc.
DOWN: obj/i386-pc/grub-core/kernel.img (31740 > 31528) - change: -212
UP: obj/i386-pc/grub-core/gfxme
This is not strictly necessary because grub_free does nothing if it is
passed a NULL pointer. However, it simplifies the code and makes it a tad
bit easier to read.
Signed-off-by: Glenn Washburn
---
grub-core/disk/luks2.c | 34 +-
1 file changed, 9 insertions(+),
Signed-off-by: Glenn Washburn
---
grub-core/disk/luks2.c | 18 --
1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/grub-core/disk/luks2.c b/grub-core/disk/luks2.c
index 199e11473..4b259cbdb 100644
--- a/grub-core/disk/luks2.c
+++ b/grub-core/disk/luks2.c
@@ -411,7 +4
Signed-off-by: Glenn Washburn
---
grub-core/disk/luks2.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/grub-core/disk/luks2.c b/grub-core/disk/luks2.c
index 4b259cbdb..2ed9e5a80 100644
--- a/grub-core/disk/luks2.c
+++ b/grub-core/disk/luks2.c
@@ -328,7 +328,7 @@ luks2_re
luks2_get_keyslot can fail for a variety of reasons that do not neccesarily
mean the next keyslot should not be tried (eg. a new kdf type). So always
try the next slot. This will make GRUB more resilient to non-spec json data
that 3rd party systems may add. We do not care if some of the keyslots ar
Patch #1: Allows GRUB to be more resilient in the fact or unexpected json
data, thus allowing access to LUKS2 volumes in cases where currently it
would be inaccessible
Patch #2-3: Add some text to go along with the error in case it gets
bubbled up to the user
Patch #4: Simplifies some error h
The messages associated with other similar GRUB_ERR_OUT_OF_RANGE errors
were lacking the trailing full stop. Syncing up the strings saves a
small amount of precious core image space on i386-pc.
DOWN: obj/i386-pc/grub-core/kernel.img (31740 > 31708) - change: -32
DOWN: i386-pc core image (bios
On Fri, 2021-03-19 at 12:16 -0500, Glenn Washburn wrote:
> diff --git a/grub-core/term/serial.c b/grub-core/term/serial.c
>
> > index f9271b092..7d4dbb2de 100644
> > --- a/grub-core/term/serial.c
> > +++ b/grub-core/term/serial.c
> > @@ -160,6 +160,18 @@ grub_serial_find (const char *name)
> >
It can be difficult to determine what device backs a particular (cryptoN)
device because they are automatically generated. This allows users to see
in the grub shell which devices back which crypto devices.
Signed-off-by: Glenn Washburn
---
grub-core/disk/cryptodisk.c | 22 ++
Signed-off-by: Glenn Washburn
---
grub-core/fs/proc.c | 91 +
1 file changed, 91 insertions(+)
diff --git a/grub-core/fs/proc.c b/grub-core/fs/proc.c
index 5f516502d..9baabb7d4 100644
--- a/grub-core/fs/proc.c
+++ b/grub-core/fs/proc.c
@@ -190,10 +190,
Its not obvious from the grub shell, nor is it documented anywhere, what
the fields in each line of (procfs)/luks_script represent. This makes it
easier for the user to comprehend the meaning of fields directly from the
shell.
Signed-off-by: Glenn Washburn
---
grub-core/disk/cryptodisk.c | 6 +++
In addition, the sector size in bytes is added to each line and it is
allowed to be 4 decimal digits long, which covers the most common cases of
512 and 4096 byte sectors. The size allocation is updated to reflect this
additional field, allow up to 4 characters and 1 space added.
Signed-off-by: Gl
Signed-off-by: Glenn Washburn
---
grub-core/disk/loopback.c | 57 +++
1 file changed, 57 insertions(+)
diff --git a/grub-core/disk/loopback.c b/grub-core/disk/loopback.c
index 41bebd14f..5957f58df 100644
--- a/grub-core/disk/loopback.c
+++ b/grub-core/disk/loo
This patch series adds proc entries to show the user some of GRUB's internal
state on devices, loopback devices, and luks devices. This can be valuable
when debugging grub, either as a developer, or as a user trying to determine
why their config is not working as expected. Some of the same info can
On Fri, 19 Mar 2021 09:07:26 +1100
Benjamin Herrenschmidt wrote:
> This adds the ability for the driver to access UARTs via MMIO instead
> of PIO selectively at runtime, and exposes a new function to add an
> MMIO port.
>
> Signed-off-by: Benjamin Herrenschmidt
> ---
> docs/grub.texi
On Fri, 19 Mar 2021 12:27:15 +1100
Benjamin Herrenschmidt wrote:
> On Thu, 2021-03-18 at 19:03 -0500, Glenn Washburn wrote:
> > > This was tested using SPCR on an actual c5.metal instance, and
> > > using explicit instanciation via serial -p mmio on a
> > > modified qemu hacked to create
22 matches
Mail list logo