this permits to select one of efi stub bootm or efi handover protocol
tested on dell latitude and qemu.
Signed-off-by: Chali Anis
---
efi/payload/Kconfig | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/efi/payload/Kconfig b/efi/payload/Kconfig
index
8f76b00744e7453ddde4a
Since an EFI malloc callback is available to provide additional memory
on demand, limit early allocations to what barebox strictly requires.
This ensures more deterministic behavior and reduces the risk of OOM
conditions.
Signed-off-by: Chali Anis
---
efi/payload/early-mem.c| 18
this will fix the OOM issues for when barebox is running on top of efi.
the OOM were unpredictble since we can't determine the size required in
advance for initrd or kernel. with this patch we don't need anymore to
increase memory in efi early mem, we allocate only what barebox needs
to relocate it
From: Ahmad Fatoum
This patch has more stock, between implementing EFI STUB boot, refactor to
reuse the code and finaly support the fit image format.
This code is tested on many qemu EFI compilations comming from ovmf ubuntu
package, tianocore efi for qemu, local edk2 build, and also tested on RP
+
include/efi/efi-payload.h | 5 +-
include/filetype.h| 6 +
include/malloc.h | 19 ++
25 files changed, 1352 insertions(+), 246 deletions(-)
---
base-commit: 3366e9e04a284b7fed71c4b8729fe331b6547d6b
change-id: 20250919-efi-loader-cdff44ad2d40
Best regards
The EFI payload loader used to copy OS images, initrds and FDTs into
freshly allocated EFI pages before handing them to LoadImage() or
installing configuration tables. This resulted in unnecessary memory
moves, extra allocations and longer boot times.
Rework the loader to:
- pass buffers from re
implement the efi stub boot on x86, this code is tested on qemu,
and dell latitude 7490. the EFI_ALLOCATE_MAX_PAGES cause an xfuncs
issue on real hardware so I transformed it to allocate any pages.
Signed-off-by: Chali Anis
---
efi/payload/bootm.c | 139 ++
add x86 linux filetype to be used to boot with efi stub
bootm image handler.
Signed-off-by: Chali Anis
---
common/filetype.c | 4
include/filetype.h | 6 ++
2 files changed, 10 insertions(+)
diff --git a/common/filetype.c b/common/filetype.c
index
eb397a175cb63df06e4426d2b99e4fde360c
From: Ahmad Fatoum
We currently lack a way to build a barebox EFI payload on ARM without
enabling some other subarchitecture. As CONFIG_EFI_PAYLOAD has a global
effect and adds EFI stubs to all enabled board, an EFI subarchitecture
doesn't fit into our current model.
Instead, let's make it possi
Signed-off-by: Chali Anis
---
arch/arm/configs/efi_v8_defconfig | 266 ++
1 file changed, 266 insertions(+)
diff --git a/arch/arm/configs/efi_v8_defconfig
b/arch/arm/configs/efi_v8_defconfig
new file mode 100644
index
From: Ahmad Fatoum
Signed-off-by: Ahmad Fatoum
Signed-off-by: Chali Anis
---
efi/Kconfig| 4 +
efi/payload/Kconfig| 4 +
efi/payload/Makefile | 2 +
efi/payload/handover.c | 195 ++
efi/payload/image.c| 228 ++-
Add the ability to install an initrd media protocol from an initrd file
or a fitImage, support both the initrd media protocol or installing the
initramfs directly to efi like what linux do.
Signed-off-by: Chali Anis
---
efi/guid.c| 7 +++
efi/payload/efi-initrd.c | 123 +++
From: Ahmad Fatoum
Signed-off-by: Ahmad Fatoum
Signed-off-by: Chali Anis
---
efi/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/efi/Kconfig b/efi/Kconfig
index
5f4c6713d539abc88e3244f7ff3933274a8c4ebb..5ae7ec92e5f8d12f72842c5c97bc9ed924b6dff9
100644
--- a/efi/Kco
remove the allocate EFI_ALLOCATE_MAX_ADDRESSES and use a different
memsize for x86 since the initrd could be bigger that those used in
arm. it might be refactored in the future with a more generic allocation
strategy (i.e barebox malloc memory vs an allocator for the boot loader data)
something lik
From: Ahmad Fatoum
This patch has more stock, between implementing EFI STUB boot, refactor to
reuse the code and finaly support the fit image format.
This code is tested on many qemu EFI compilations comming from ovmf ubuntu
package, tianocore efi for qemu, local edk2 build, and also tested on RP
On Mon, Sep 15, 2025 at 11:02:12AM +0200, Ahmad Fatoum wrote:
> Hi,
>
> On 9/14/25 9:32 PM, chalian...@gmail.com wrote:
> > From: Chali Anis
> >
> > This patch is an esthetic cleanup, rename the dma_ops to dma_device_ops
> > to be sure that the struct in drivers/dma is different from one used in
Based on this commit found in the TI downstream repository:
| commit 20778f6c125dc24e62767dc4837d793a0f18c319
| Author: Siddharth Vadapalli
| Date: Wed Jun 4 15:41:53 2025 +0530
|
|PENDING: arm64: dts: ti: k3-am62l-main: Fix MDIO clock
|
|Based on the Clock ID documentation for AM62L [0
17 matches
Mail list logo