From: "Ard Biesheuvel" <ard.biesheuvel@linaro.org>
To: Pete Batard <pete@akeo.ie>
Cc: edk2-devel-groups-io <devel@edk2.groups.io>,
"Leif Lindholm" <leif@nuviainc.com>,
"Philippe Mathieu-Daudé" <philmd@redhat.com>,
"Andrei Warkentin" <awarkentin@vmware.com>
Subject: Re: [edk2-devel][PATCH 2/5] Platform/RPi: Separate RAM descriptors between 0-3 GB and 3+ GB
Date: Tue, 3 Mar 2020 11:39:19 +0100 [thread overview]
Message-ID: <CAKv+Gu9-1x0MMj4VZmJiXEspCks1kQgDnPG-zHeOdZSDcRdwBg@mail.gmail.com> (raw)
In-Reply-To: <20200303103339.7468-3-pete@akeo.ie>
On Tue, 3 Mar 2020 at 11:33, Pete Batard <pete@akeo.ie> wrote:
>
> From: Andrei Warkentin <andrey.warkentin@gmail.com>
>
> This is required so that we can make the 3 to 4 GB segment of
> the Raspberry Pi 4 a user-configurable option.
>
> This also removes the need for the PcdAcpiBasicMode PCD.
>
> Signed-off-by: Pete Batard <pete@akeo.ie>
Please fix the commit log so it is clear what 'this' means
> ---
> Platform/RaspberryPi/Library/PlatformLib/PlatformLib.inf | 3 --
> Platform/RaspberryPi/Library/PlatformLib/RaspberryPiMem.c | 33 ++++++++++++++------
> 2 files changed, 23 insertions(+), 13 deletions(-)
>
> diff --git a/Platform/RaspberryPi/Library/PlatformLib/PlatformLib.inf b/Platform/RaspberryPi/Library/PlatformLib/PlatformLib.inf
> index 9abcc2cc0075..77cdbe399a06 100644
> --- a/Platform/RaspberryPi/Library/PlatformLib/PlatformLib.inf
> +++ b/Platform/RaspberryPi/Library/PlatformLib/PlatformLib.inf
> @@ -59,8 +59,5 @@ [FixedPcd]
> gEfiMdeModulePkgTokenSpaceGuid.PcdFlashNvStorageFtwWorkingSize
> gEfiMdeModulePkgTokenSpaceGuid.PcdFlashNvStorageFtwSpareSize
>
> -[FeaturePcd]
> - gRaspberryPiTokenSpaceGuid.PcdAcpiBasicMode
> -
> [Ppis]
> gArmMpCoreInfoPpiGuid
> diff --git a/Platform/RaspberryPi/Library/PlatformLib/RaspberryPiMem.c b/Platform/RaspberryPi/Library/PlatformLib/RaspberryPiMem.c
> index 4b388465cdde..3f257d4fa528 100644
> --- a/Platform/RaspberryPi/Library/PlatformLib/RaspberryPiMem.c
> +++ b/Platform/RaspberryPi/Library/PlatformLib/RaspberryPiMem.c
> @@ -25,7 +25,7 @@ UINT32 mBoardRevision;
>
>
> // The total number of descriptors, including the final "end-of-table" descriptor.
> -#define MAX_VIRTUAL_MEMORY_MAP_DESCRIPTORS 10
> +#define MAX_VIRTUAL_MEMORY_MAP_DESCRIPTORS 11
>
> STATIC BOOLEAN VirtualMemoryInfoInitialized = FALSE;
> STATIC RPI_MEMORY_REGION_INFO VirtualMemoryInfo[MAX_VIRTUAL_MEMORY_MAP_DESCRIPTORS];
> @@ -60,6 +60,7 @@ ArmPlatformGetVirtualMemoryMap (
> {
> UINTN Index = 0;
> UINTN GpuIndex;
> + INT64 OrigMemorySize;
> INT64 SystemMemorySize;
> ARM_MEMORY_REGION_DESCRIPTOR *VirtualMemoryTable;
>
> @@ -78,7 +79,6 @@ ArmPlatformGetVirtualMemoryMap (
> return;
> }
>
> -
> // Firmware Volume
> VirtualMemoryTable[Index].PhysicalBase = FixedPcdGet64 (PcdFdBaseAddress);
> VirtualMemoryTable[Index].VirtualBase = VirtualMemoryTable[Index].PhysicalBase;
Drop this hunk please
> @@ -155,13 +155,13 @@ ArmPlatformGetVirtualMemoryMap (
> VirtualMemoryInfo[Index].Type = RPI_MEM_UNMAPPED_REGION;
> VirtualMemoryInfo[Index++].Name = L"SoC Reserved (283x)";
>
> - if (FeaturePcdGet (PcdAcpiBasicMode)) {
> - //
> - // Limit the memory to 3 GB to work around the DMA bugs in the SoC without
> - // having to rely on IORT or _DMA descriptions.
> - //
> - SystemMemorySize = MIN(SystemMemorySize, 3U * SIZE_1GB);
> - }
> + //
> + // By default we limit the memory to 3 GB to work around the DMA bugs in the SoC, for
> + // OSes that don't support _DMA range descriptors. On 4GB boards, it's runtime
> + // setting to boot with 4GB, and the additional 1GB is added by ConfigDxe.
> + //
> + OrigMemorySize = SystemMemorySize;
> + SystemMemorySize = MIN(SystemMemorySize, 3U * SIZE_1GB);
>
> // If we have RAM above the 1 GB mark, declare it
> if (SystemMemorySize - SIZE_1GB > 0) {
> @@ -170,7 +170,20 @@ ArmPlatformGetVirtualMemoryMap (
> VirtualMemoryTable[Index].Length = SystemMemorySize - SIZE_1GB;
> VirtualMemoryTable[Index].Attributes = ARM_MEMORY_REGION_ATTRIBUTE_WRITE_BACK;
> VirtualMemoryInfo[Index].Type = RPI_MEM_BASIC_REGION;
> - VirtualMemoryInfo[Index++].Name = L"Extended System RAM";
> + VirtualMemoryInfo[Index++].Name = L"Extended System RAM below 3GB";
> + }
> +
> + //
> + // If we have RAM above 3 GB mark, declare it so it's mapped, but
> + // don't add it to the memory map. This is done later by ConfigDxe if necessary.
> + //
> + if (OrigMemorySize > (3UL * SIZE_1GB)) {
> + VirtualMemoryTable[Index].PhysicalBase = 3UL * SIZE_1GB;
> + VirtualMemoryTable[Index].VirtualBase = VirtualMemoryTable[Index].PhysicalBase;
> + VirtualMemoryTable[Index].Length = OrigMemorySize - VirtualMemoryTable[Index].PhysicalBase;
> + VirtualMemoryTable[Index].Attributes = ARM_MEMORY_REGION_ATTRIBUTE_WRITE_BACK;
> + VirtualMemoryInfo[Index].Type = RPI_MEM_UNMAPPED_REGION;
> + VirtualMemoryInfo[Index++].Name = L"Extended System RAM above 3GB";
> }
>
> // End of Table
> --
> 2.21.0.windows.1
>
next prev parent reply other threads:[~2020-03-03 10:39 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-03 10:33 [edk2-devel][PATCH 0/5] Platform/RPi: User config improvements Pete Batard
2020-03-03 10:33 ` [edk2-devel][PATCH 1/5] Platform/RPi: Add firmware call to read installed memory size Pete Batard
2020-03-03 10:37 ` Ard Biesheuvel
2020-03-03 10:33 ` [edk2-devel][PATCH 2/5] Platform/RPi: Separate RAM descriptors between 0-3 GB and 3+ GB Pete Batard
2020-03-03 10:39 ` Ard Biesheuvel [this message]
2020-03-03 10:33 ` [edk2-devel][PATCH 3/5] Platform/RPi: Make 3GB/4GB a runtime (BIOS setup) choice Pete Batard
2020-03-03 10:42 ` Ard Biesheuvel
2020-03-03 10:33 ` [edk2-devel][PATCH 4/5] Platform/RPi: Make Device Tree provision " Pete Batard
2020-03-03 10:45 ` Ard Biesheuvel
2020-03-03 10:33 ` [edk2-devel][PATCH 5/5] Platform/RPi/ConfigDxe: Improve RPi configuration forms Pete Batard
2020-03-03 10:45 ` Ard Biesheuvel
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-list from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAKv+Gu9-1x0MMj4VZmJiXEspCks1kQgDnPG-zHeOdZSDcRdwBg@mail.gmail.com \
--to=devel@edk2.groups.io \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox