From: "Ard Biesheuvel" <ard.biesheuvel@linaro.org>
To: Leif Lindholm <leif.lindholm@linaro.org>
Cc: edk2-devel-groups-io <devel@edk2.groups.io>
Subject: Re: [PATCH edk2-platforms v2 5/8] Silicon/AMD/StyxDtbLoaderLib: add interrupt-affinity property to PMU node
Date: Thu, 28 Nov 2019 14:29:52 +0100 [thread overview]
Message-ID: <CAKv+Gu_w_ZVuUDzwc7JxCprOzKb06KnscY6EqM7t-VPXmaJkSg@mail.gmail.com> (raw)
In-Reply-To: <20191128132244.GO7359@bivouac.eciton.net>
On Thu, 28 Nov 2019 at 14:22, Leif Lindholm <leif.lindholm@linaro.org> wrote:
>
> On Wed, Nov 27, 2019 at 19:44:36 +0100, Ard Biesheuvel wrote:
> > AMD Seattle uses a range of SPIs to signal PMU events, and this requires
> > a description in the DT which SPI maps to which CPU. This requires us to
> > defer the generation of the PMU node to a point where the CPU phandles
> > have been allocated.
>
> Ah, so these were previously populated with garbage? Oops.
>
No, not really. Moving the code obfuscates it a bit, but the new code
has an additional fdt_appendprop_cell() to set "interrupt-affinity"
which was missing entirely before.
> > Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
>
> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
>
Thanks,
> > ---
> > Silicon/AMD/Styx/Library/StyxDtbLoaderLib/StyxDtbLoaderLib.c | 55 ++++++++++----------
> > 1 file changed, 28 insertions(+), 27 deletions(-)
> >
> > diff --git a/Silicon/AMD/Styx/Library/StyxDtbLoaderLib/StyxDtbLoaderLib.c b/Silicon/AMD/Styx/Library/StyxDtbLoaderLib/StyxDtbLoaderLib.c
> > index 261b5f59c8df..2f7b5e2a7b25 100644
> > --- a/Silicon/AMD/Styx/Library/StyxDtbLoaderLib/StyxDtbLoaderLib.c
> > +++ b/Silicon/AMD/Styx/Library/StyxDtbLoaderLib/StyxDtbLoaderLib.c
> > @@ -331,33 +331,6 @@ PrepareFdt (
> > // Get Id from primary CPU
> > MpId = (UINTN)ArmReadMpidr ();
> >
> > - // Create /pmu node
> > - PmuNode = fdt_add_subnode(Fdt, 0, "pmu");
> > - if (PmuNode >= 0) {
> > - fdt_setprop_string (Fdt, PmuNode, "compatible", "arm,armv8-pmuv3");
> > -
> > - // append PMU interrupts
> > - for (Index = 0; Index < ArmCoreCount; Index++) {
> > - MpId = (UINTN)GET_MPID (ArmCoreInfoTable[Index].ClusterId,
> > - ArmCoreInfoTable[Index].CoreId);
> > -
> > - Status = AmdMpCoreInfoProtocol->GetPmuSpiFromMpId (MpId, &PmuInt.IntId);
> > - if (EFI_ERROR (Status)) {
> > - DEBUG ((DEBUG_ERROR,
> > - "FDT: Error getting PMU interrupt for MpId '0x%x'\n", MpId));
> > - return Status;
> > - }
> > -
> > - PmuInt.Flag = cpu_to_fdt32 (PMU_INT_FLAG_SPI);
> > - PmuInt.IntId = cpu_to_fdt32 (PmuInt.IntId);
> > - PmuInt.Type = cpu_to_fdt32 (PMU_INT_TYPE_HIGH_LEVEL);
> > - fdt_appendprop (Fdt, PmuNode, "interrupts", &PmuInt, sizeof(PmuInt));
> > - }
> > - } else {
> > - DEBUG ((DEBUG_ERROR, "FDT: Error creating 'pmu' node\n"));
> > - return EFI_INVALID_PARAMETER;
> > - }
> > -
> > // Create /cpus noide
> > Node = fdt_add_subnode (Fdt, 0, "cpus");
> > if (Node >= 0) {
> > @@ -449,6 +422,34 @@ PrepareFdt (
> > return EFI_INVALID_PARAMETER;
> > }
> >
> > + // Create /pmu node
> > + PmuNode = fdt_add_subnode(Fdt, 0, "pmu");
> > + if (PmuNode >= 0) {
> > + fdt_setprop_string (Fdt, PmuNode, "compatible", "arm,armv8-pmuv3");
> > +
> > + // append PMU interrupts
> > + for (Index = 0; Index < ArmCoreCount; Index++) {
> > + MpId = (UINTN)GET_MPID (ArmCoreInfoTable[Index].ClusterId,
> > + ArmCoreInfoTable[Index].CoreId);
> > +
> > + Status = AmdMpCoreInfoProtocol->GetPmuSpiFromMpId (MpId, &PmuInt.IntId);
> > + if (EFI_ERROR (Status)) {
> > + DEBUG ((DEBUG_ERROR,
> > + "FDT: Error getting PMU interrupt for MpId '0x%x'\n", MpId));
> > + return Status;
> > + }
> > +
> > + PmuInt.Flag = cpu_to_fdt32 (PMU_INT_FLAG_SPI);
> > + PmuInt.IntId = cpu_to_fdt32 (PmuInt.IntId);
> > + PmuInt.Type = cpu_to_fdt32 (PMU_INT_TYPE_HIGH_LEVEL);
> > + fdt_appendprop (Fdt, PmuNode, "interrupts", &PmuInt, sizeof(PmuInt));
> > + fdt_appendprop_cell (Fdt, PmuNode, "interrupt-affinity", Phandle[Index]);
> > + }
> > + } else {
> > + DEBUG ((DEBUG_ERROR, "FDT: Error creating 'pmu' node\n"));
> > + return EFI_INVALID_PARAMETER;
> > + }
> > +
> > SetSocIdStatus (Fdt);
> > SetXgbeStatus (Fdt);
> >
> > --
> > 2.17.1
> >
next prev parent reply other threads:[~2019-11-28 13:29 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-11-27 18:44 [PATCH edk2-platforms v2 0/8] fixes and updates for AMD OverDrive Ard Biesheuvel
2019-11-27 18:44 ` [PATCH edk2-platforms v2 1/8] Platform/Overdrive: add missing resolution for FileHandleLib Ard Biesheuvel
2019-11-27 18:44 ` [PATCH edk2-platforms v2 2/8] Platform/Overdrive: clean up stream ID descriptions in DT Ard Biesheuvel
2019-11-27 18:44 ` [PATCH edk2-platforms v2 3/8] Platform/Overdrive: fix a typo in the DT Ard Biesheuvel
2019-11-28 12:32 ` Leif Lindholm
2019-11-28 13:32 ` Ard Biesheuvel
2019-11-27 18:44 ` [PATCH edk2-platforms v2 4/8] Silicon/AMD/Styx: clean up stream ID mappings for SMMU Ard Biesheuvel
2019-11-27 18:44 ` [PATCH edk2-platforms v2 5/8] Silicon/AMD/StyxDtbLoaderLib: add interrupt-affinity property to PMU node Ard Biesheuvel
2019-11-28 13:22 ` Leif Lindholm
2019-11-28 13:29 ` Ard Biesheuvel [this message]
2019-11-27 18:44 ` [PATCH edk2-platforms v2 6/8] Silicon/AMD/StyxDtbLoaderLib: add description of the cache topology Ard Biesheuvel
2019-11-28 13:24 ` Leif Lindholm
2019-11-27 18:44 ` [PATCH edk2-platforms v2 7/8] Silicon/AMD/StyxDtbLoaderLib: use Cortex-A57 IDs instead of generic ARMv8 Ard Biesheuvel
2019-11-28 13:37 ` Leif Lindholm
2019-11-28 13:39 ` Ard Biesheuvel
2019-11-28 13:40 ` Leif Lindholm
2019-11-28 15:07 ` Ard Biesheuvel
2019-11-27 18:44 ` [PATCH edk2-platforms v2 8/8] Silicon/AMD/StyxDtbLoaderLib: omit linux,phandle properties Ard Biesheuvel
2019-11-28 13:38 ` Leif Lindholm
2019-11-28 15:48 ` [PATCH edk2-platforms v2 0/8] fixes and updates for AMD OverDrive 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+Gu_w_ZVuUDzwc7JxCprOzKb06KnscY6EqM7t-VPXmaJkSg@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