Hi Olivier,
Thanks for the clarifications. Q1 and Q2 are not yet clear to me. Please
find my response embedded in mail below.
Best Regards,
Asha
On Tue, Mar 27, 2012 at 12:15 AM, Olivier Martin <olivier.martin(a)arm.com>wrote:
> Hi Asha,****
>
> ** **
>
> *Q1: Is UEFI flashed as single binary i.e., PlatformPkg.fd or different
> efi's?*
>
> It really depends of your configuration/platform. UEFI is only a portion
> of your boot firmware. Prior to UEFI, you might have one (or more)
> additional boot code/firmware.****
>
> The EDK2 FDF file will help you to generate a Firmware Device file you can
> flash onto your platform Non-Volatile memory (eg: ROM, NAND)****
>
> But you can also load single EFI application/driver (EFI binary) from a
> UEFI firmware. For instance, if you develop a EFI driver, you do not need
> to rebuild the full Platform UEFI firmware. You can load the single EFI
> driver.****
>
> ** [Asha]
>
Assuming UEFI binary is flashed to NAND Flash, the Firmware (some
bootloader) prior to UEFI will load UEFI
binary (xxx.fd) from volatile memory like ROM, NAND to RAM. In this
scenario, will the binary xxx.fd be copied to
RAM or just the SEC phase of UEFI?
> **
>
> *Q2. As the code is in ROM (flash), I understand that it must be copied
> to RAM for execution.
> **
>
> Will complete UEFI be copied to RAM or some phase or section wise?
> *
>
> RAM provides much better performance than some other memory storage. Code
> can also be run from RAM (at the difference of NAND for instance).****
>
> But there is no requirement to copy the code into RAM for execution. It
> depends of your platform specification and the choice of the firmware
> engineer.****
>
> If we take the example of EDK2 code base, the PEI Core (Pre EFI phase)
> copies the firmware into DRAM as soon as the System Memory (DRAM) is setup.
> ****
>
> ** [Asha]
>
Considering EDK2 as example, if PEI core must initialize DRAM, from
where will PEI core be executed? From
flash? you said "PEI Core (Pre EFI phase) copies the firmware into
DRAM "
what or which "firmware" is this? Is it DXE?
> **
>
> *Q3. How does operating system invoke the UEFI Runtime services? *
>
> The UEFI specification covers this part . See the section “2.3 Calling
> Convention“ of the Specification.****
>
> ** **
>
> *Q4. I read in one of the internet link that other than Timer, there are
> no hardware interrupts.
> Without hardware interrupt how will the USB kind of device be
> recognized? *
>
> The UEFI drivers can register functions to be called periodically (see the
> ‘Boot Services’ Event functions). The USB drivers register periodic
> functions that scan if new devices are connected.****
>
> ** **
>
> I hope that will help you,****
>
> Olivier****
>
> ** **
>
> ** **
>
> ** **
>
> *From:* boot-architecture-bounces(a)lists.linaro.org [mailto:
> boot-architecture-bounces(a)lists.linaro.org] *On Behalf Of *Ryan Harkin
> *Sent:* 26 March 2012 16:17
> *To:* Asha R
> *Cc:* boot-architecture(a)lists.linaro.org
> *Subject:* Re: UEFI basic questions****
>
> ** **
>
> Hi Asha,
>
> I don't know the proper answers to all of your questions. I would suggest
> that the EDK2 mailing list might give you more detailed answers:
> ****
>
> edk2-devel(a)lists.sourceforge.net
>
> I think you have to sign up for the list first:
>
> https://lists.sourceforge.net/lists/listinfo/edk2-devel
>
> ****
>
> 2012/3/26 Asha R <rd.asha22(a)gmail.com>****
>
> I am new to UEFI and trying to understand working of UEFI on ARM platform.
> ****
>
> On building UEFI, there are different binaries generated for each phase
> and also for drivers
> eg: platformSec.efi, platformBds.efi and UsbDxe.efi,etc****
>
> Q1. Is UEFI flashed as single binary i.e., PlatformPkg.fd or different
> efi's? ****
>
> I flash a single .fd onto my platforms.****
>
> Q1. As the code is in ROM (flash), I understand that it must be copied to
> RAM for execution.
> Will complete UEFI be copied to RAM or some phase or section wise? *
> ***
>
> Q3. How does operating sysetm invoke the UEFI Runtime services? ****
>
> Q4. I read in one of the internet link that other than Timer, there are no
> hardware interrupts.
> Without hardware interrupt how will the USB kind of device be
> recognized? ****
>
> UEFI has interrupt support for ARM platforms.****
>
>
> Regards,
> Ryan.****
>
I am new to UEFI and trying to understand working of UEFI on ARM platform.
On building UEFI, there are different binaries generated for each phase and
also for drivers
eg: platformSec.efi, platformBds.efi and UsbDxe.efi,etc
Q1. Is UEFI flashed as single binary i.e., PlatformPkg.fd or different
efi's?
Q1. As the code is in ROM (flash), I understand that it must be copied to
RAM for execution.
Will complete UEFI be copied to RAM or some phase or section wise?
Q3. How does operating sysetm invoke the UEFI Runtime services?
Q4. I read in one of the internet link that other than Timer, there are no
hardware interrupts.
Without hardware interrupt how will the USB kind of device be
recognized?
Kindly clarify my doubts to help me in understanding UEFI better. Thanks in
advance.
Regards,
Asha R
Hi Guys ,
I am working on a samsung sp5v310 development kit.
I wanted to read a peripheral register (battery charger) in u-boot.
This device is interfaced on i2c0. I am working with the u-boot in
git://git.linaro.org/boot/u-boot-linaro-stable.git
I wanted to do i2c_read in the code, but as i understand the i2c
support is not there in u-boot for s5pv310 SOCs. Please confirm, if this
is correct or i am wrong.
If there is support to do i2c operations on s5pv310 SOCs, please point
me to the right file, where i can find i2c configurations for s5pv310 SOCs.
thanks,
- Avinash
PS:
1. There are no i2c macros in include/configs/smdkv310.h.
2. After adding,below macros to enable i2c support for smdkv310,
/* I1C */
#define CONFIG_HARD_I2C 1
#define CONFIG_SYS_I2C_SPEED 100000
#define CONFIG_SYS_I2C_SLAVE 1
#define CONFIG_SYS_I2C_BUS 0
#define CONFIG_SYS_I2C_BUS_SELECT 1
#define CONFIG_I2C_MULTI_BUS 1
i get below error:
/media/data/avinash/c.labs/
git.clones/bootloader/uboot/arch/arm/lib/board.c:189: undefined reference
to `i2c_init'
common/libcommon.o: In function `stdio_init':
/media/data/avinash/c.labs/git.clones/bootloader/uboot/common/stdio.c:215:
undefined reference to `i2c_init'
And there are no i2c related support inside the u-boot for samsung smdkv310.
avinash@avinash-desktop:/media/data/avinash/c.labs/git.clones/bootloader-new/uboot$
grep -inlr 'v310' . > v310.files.txt
avinash@avinash-desktop:/media/data/avinash/c.labs/git.clones/bootloader-new/uboot$
cat v310.files.txt | xargs grep -inr 'i2c_init'
Binary file ./uboot.PS matches
avinash@avinash-desktop:/media/data/avinash/c.labs/git.clones/bootloader-new/uboot$
Update: Linaro has signed the paperwork, and as soon as everything is
resolved, I will be joining the ARM bindings sub-team
> - [CARRIEDOVER] David Rusling: - Linaro UEFI membership - target Linaro
> Connect
> - [CARRIEDOVER] Grant to draft a position on secure boot, from the
> Linaro perspective, which can be reviewed by Linaro TSC
Kristen Eisenberg
Billige Flüge
Marketing GmbH
Emanuelstr. 3,
10317 Berlin
Deutschland
Telefon: +49 (33)
5310967
Email:
utebachmeier at
gmail.com
Site:
http://flug.airego.de
- Billige Flüge vergleichen
Hi all,
I have just created a page on the Tianocore wikipage to explain the steps to
build EDK2 for the Samsung Origen low cost development board:
https://sourceforge.net/apps/mediawiki/tianocore/index.php?title=SamsungPlat
formPkg
These instructions are based on the Samsung Origen EDK2 support written by
Girish K S (Samsung / Linaro) and target the Linaro Linux 'distribution'
(Linux kernel + file system provided by Linaro).
Please find the attached patches to fix the current head of the Samsung
Origen EDK2 port (ed456b2d64f3 (22/11/2011)).
Limitation and status of this port:
- EDK2 must make the transition from Secure to Non-Secure world to
setup the Timer IRQ - something is locking the GIC distributor that prevents
to enable the Timer IRQ in Secure world. I have not investigated this issue.
- Cannot start the Linaro's kernel - I have not investigated this
issue yet.
- Display support (not tested).
- Can start a EFI Shell or an EFI application.
Kind Regards,
Olivier
Hello
yes that is correct, I have now cancelled the meeting and sent you all a notification.
I was talking to Grant yesterday about how to proceed with the meeting, and there will be more information communicated once we have finished Connect. In short, the meetings should resume once we have enough progress in the porting of grub for uboot at least.
In the meantime we could use the mailing list for any comments or news.
Are there any serious objections on this? I fear that without progress on the code side the meetings would merely discuss what we have read, we can do that on the mailing list.
Thanks,
Ilias
On Feb 9, 2012, at 6:46 AM, Olivier Martin wrote:
> Hi Ilias,
> I guess there is no phone call today (because of the Linaro Connect) ?
> Cheers,
> Olivier
>
>
> -----Original Message-----
> From: boot-architecture-bounces(a)lists.linaro.org [mailto:boot-architecture-bounces@lists.linaro.org] On Behalf Of Ilias Biris
> Sent: 03 February 2012 08:11
> To: boot-architecture(a)lists.linaro.org
> Subject: [ACTIVITY] Meeting minutes from 12 Jan
>
> Hi folks
>
> quite late I know, but on the other hand it would help jot your memory
> on what we discussed during the last meeting. Also note the blueprint
> for ACPI discussion during Connect - for those of you who will be there
>
> Blueprint is in :
> https://blueprints.launchpad.net/linaro/+spec/linaro-general-q112-acpi-disc…
>
>
> The meeting minutes are in:
> https://wiki.linaro.org/OfficeofCTO/BootArchitecture/2012-01-12
>
> Cheers
> --
> Ilias Biris ilias.biris(a)linaro.org
> Project Manager, Linaro
> M: +358504839608, IRC: ibiris Skype: ilias_biris
> Linaro.org│ Open source software for ARM SoCs
>
> _______________________________________________
> boot-architecture mailing list
> boot-architecture(a)lists.linaro.org
> http://lists.linaro.org/mailman/listinfo/boot-architecture
>
>
>
Hi folks
quite late I know, but on the other hand it would help jot your memory
on what we discussed during the last meeting. Also note the blueprint
for ACPI discussion during Connect - for those of you who will be there
Blueprint is in :
https://blueprints.launchpad.net/linaro/+spec/linaro-general-q112-acpi-disc…
The meeting minutes are in:
https://wiki.linaro.org/OfficeofCTO/BootArchitecture/2012-01-12
Cheers
--
Ilias Biris ilias.biris(a)linaro.org
Project Manager, Linaro
M: +358504839608, IRC: ibiris Skype: ilias_biris
Linaro.org│ Open source software for ARM SoCs
Hello folks
with apologies for sending this late here are the meeting notes from the
latest boot architecture meeting:
https://wiki.linaro.org/OfficeofCTO/BootArchitecture/2011-12-08
There was only 1 action recorded and that is
ACTION: Jean-Christophe to look at boot-cycle from the DT side
Also JC asked what would be the wishlist in terms of HW and IO for a
reference platform. Please send your ideas straight to the list or if
you prefer to JC directly.
Cheers,
--
Ilias Biris ilias.biris(a)linaro.org
Project Manager, Linaro
M: +358504839608, IRC: ibiris Skype: ilias_biris
Linaro.org│ Open source software for ARM SoCs
Hi,
I'm wondering if folks in the boot architecture team have any input on this?
~Deepak
---------- Forwarded message ----------
From: Venkatraman S <venkat(a)linaro.org>
Date: 25 November 2011 04:19
Subject: eMMC4.5 Boot partition support -expectations
To: Deepak Saxena <dsaxena(a)linaro.org>, Saugata Das <saugata.das(a)linaro.org>
Cc: Arnd Bergmann <arnd.bergmann(a)linaro.org>, Yejin.moon(a)samsung.com
Hi Deepak, all,
According to this cycle's eMMC4.5 requirements in [1], there is no
need to provide any specific support in the boot loader for Boot
partition, right ?
As of 3.1, the Kernel recognizes the boot partitions and exposes
them as read-only. Any user space tool can update the data in the boot
partition by overriding the sysfs entry, and writing the appropriate
data.
Let me know if there are any specific expectations on this.
Thanks and regards,
Venkat.
[1] https://linaro-public.papyrs.com/public/4086/KWG2011-EMMC-4.5-SUPPORT/
