From: "Marcin Wojtas" <mw@semihalf.com>
To: Ard Biesheuvel <ard.biesheuvel@arm.com>
Cc: edk2-devel-groups-io <devel@edk2.groups.io>,
Leif Lindholm <leif@nuviainc.com>
Subject: Re: [RFT PATCH edk2-platforms] Silicon/Marvell/MvI2cDxe: connect all I2C masters at EndOfDxe
Date: Fri, 5 Jun 2020 17:07:46 +0200 [thread overview]
Message-ID: <CAPv3WKdCnXhQkuNt27HFZXhzh5MA8GociB_knmygwX5dAuZ4HA@mail.gmail.com> (raw)
In-Reply-To: <5903c3ba-8869-5b7e-73e0-cd3a90217e98@arm.com>
pt., 5 cze 2020 o 16:46 Ard Biesheuvel <ard.biesheuvel@arm.com> napisał(a):
>
> On 6/5/20 3:27 PM, Marcin Wojtas wrote:
> > pt., 5 cze 2020 o 15:08 Ard Biesheuvel <ard.biesheuvel@arm.com> napisał(a):
> >>
> >> On 6/5/20 1:54 PM, Marcin Wojtas wrote:
> >>> Hi Ard,
> >>>
> >>>
> >>> czw., 4 cze 2020 o 23:35 Ard Biesheuvel <ard.biesheuvel@arm.com> napisał(a):
> >>>>
> >>>> To ensure that platforms incorporating MvI2cDxe will keep working
> >>>> as intended once the platform BDS code stops calling ConnectAll(),
> >>>> connect the I2C masters explicitly at EndOfDxe.
> >>>>
> >>>> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@arm.com>
> >>>> ---
> >>>> Build tested only.
> >>>>
> >>>> Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.inf | 3 ++
> >>>> Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.h | 1 +
> >>>> Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.c | 54 ++++++++++++++++++--
> >>>> 3 files changed, 55 insertions(+), 3 deletions(-)
> >>>>
> >>>> diff --git a/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.inf b/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.inf
> >>>> index e59fee0ac1b5..f631fbe797fc 100755
> >>>> --- a/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.inf
> >>>> +++ b/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.inf
> >>>> @@ -45,5 +45,8 @@ [Pcd]
> >>>> gMarvellTokenSpaceGuid.PcdI2cBaudRate
> >>>> gMarvellTokenSpaceGuid.PcdI2cBusCount
> >>>>
> >>>> +[Guids]
> >>>> + gEfiEndOfDxeEventGroupGuid
> >>>> +
> >>>> [Depex]
> >>>> TRUE
> >>>> diff --git a/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.h b/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.h
> >>>> index f5c2cdd8ab3a..6caaa45cece0 100644
> >>>> --- a/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.h
> >>>> +++ b/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.h
> >>>> @@ -80,6 +80,7 @@ typedef struct {
> >>>>
> >>>> typedef struct {
> >>>> VENDOR_DEVICE_PATH Guid;
> >>>> + UINTN Instance;
> >>>> EFI_DEVICE_PATH_PROTOCOL End;
> >>>> } MV_I2C_DEVICE_PATH;
> >>>>
> >>>> diff --git a/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.c b/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.c
> >>>> index b13ab8f02c99..dfe8da9891a5 100755
> >>>> --- a/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.c
> >>>> +++ b/Silicon/Marvell/Drivers/I2c/MvI2cDxe/MvI2cDxe.c
> >>>> @@ -30,12 +30,13 @@ STATIC MV_I2C_DEVICE_PATH MvI2cDevicePathProtocol = {
> >>>> HARDWARE_DEVICE_PATH,
> >>>> HW_VENDOR_DP,
> >>>> {
> >>>> - (UINT8) (sizeof(VENDOR_DEVICE_PATH)),
> >>>> - (UINT8) (sizeof(VENDOR_DEVICE_PATH) >> 8),
> >>>> + (UINT8) (OFFSET_OF (MV_I2C_DEVICE_PATH, End)),
> >>>> + (UINT8) (OFFSET_OF (MV_I2C_DEVICE_PATH, End) >> 8),
> >>>> },
> >>>> },
> >>>> EFI_CALLER_ID_GUID
> >>>> },
> >>>> + 0, // Instance
> >>>> {
> >>>> END_DEVICE_PATH_TYPE,
> >>>> END_ENTIRE_DEVICE_PATH_SUBTYPE,
> >>>> @@ -86,7 +87,7 @@ MvI2cInitialiseController (
> >>>> DEBUG((DEBUG_ERROR, "MvI2cDxe: I2C device path allocation failed\n"));
> >>>> return EFI_OUT_OF_RESOURCES;
> >>>> }
> >>>> - DevicePath->Guid.Guid.Data4[0] = Bus;
> >>>> + DevicePath->Instance = Bus;
> >>>>
> >>>> /* if attachment succeeds, this gets freed at ExitBootServices */
> >>>> I2cMasterContext = AllocateZeroPool (sizeof (I2C_MASTER_CONTEXT));
> >>>> @@ -139,6 +140,47 @@ MvI2cInitialiseController (
> >>>> return Status;
> >>>> }
> >>>>
> >>>> +STATIC
> >>>> +VOID
> >>>> +EFIAPI
> >>>> +OnEndOfDxe (
> >>>> + IN EFI_EVENT Event,
> >>>> + IN VOID *Context
> >>>> + )
> >>>> +{
> >>>> + MV_I2C_DEVICE_PATH *DevicePath;
> >>>> + EFI_DEVICE_PATH_PROTOCOL *DevicePathPointer;
> >>>> + EFI_HANDLE DeviceHandle;
> >>>> + EFI_STATUS Status;
> >>>> +
> >>>> + gBS->CloseEvent (Event);
> >>>> +
> >>>> + DevicePath = AllocateCopyPool (sizeof (MvI2cDevicePathProtocol),
> >>>> + &MvI2cDevicePathProtocol);
> >>>> + if (DevicePath == NULL) {
> >>>> + DEBUG ((DEBUG_ERROR, "MvI2cDxe: I2C device path allocation failed\n"));
> >>>> + return;
> >>>> + }
> >>>> +
> >>>> + do {
> >>>> + DevicePathPointer = (EFI_DEVICE_PATH_PROTOCOL *)DevicePath;
> >>>> + Status = gBS->LocateDevicePath (&gEfiI2cMasterProtocolGuid,
> >>>> + &DevicePathPointer, &DeviceHandle);
> >>>> + if (EFI_ERROR (Status)) {
> >>>> + break;
> >>>> + }
> >>>> +
> >>>> + Status = gBS->ConnectController (DeviceHandle, NULL, NULL, TRUE);
> >>>> + DEBUG ((DEBUG_INFO, "%a: ConnectController () returned %r\n",
> >>>> + __FUNCTION__, Status));
> >>>> +
> >>>> + DevicePath->Instance++;
> >>>> + } while (TRUE);
> >>>> +
> >>>> + gBS->FreePool (DevicePath);
> >>>> +}
> >>>> +
> >>>> +
> >>>> EFI_STATUS
> >>>> EFIAPI
> >>>> MvI2cInitialise (
> >>>> @@ -150,6 +192,8 @@ MvI2cInitialise (
> >>>> MV_BOARD_I2C_DESC *Desc;
> >>>> EFI_STATUS Status;
> >>>> UINTN Index;
> >>>> + EFI_EVENT EndOfDxeEvent;
> >>>> +
> >>>>
> >>>> /* Obtain list of available controllers */
> >>>> Status = gBS->LocateProtocol (&gMarvellBoardDescProtocolGuid,
> >>>> @@ -177,6 +221,10 @@ MvI2cInitialise (
> >>>>
> >>>> BoardDescProtocol->BoardDescFree (Desc);
> >>>>
> >>>> + Status = gBS->CreateEventEx (EVT_NOTIFY_SIGNAL, TPL_CALLBACK, OnEndOfDxe,
> >>>> + NULL, &gEfiEndOfDxeEventGroupGuid, &EndOfDxeEvent);
> >>>> + ASSERT_EFI_ERROR (Status);
> >>>> +
> >>>> return EFI_SUCCESS;
> >>>> }
> >>>>
> >>>> --
> >>>> 2.26.2
> >>>>
> >>>
> >>> This change works as planned, so you can add my:
> >>> Tested-by: Marcin Wojtas <mw@semihalf.com>
> >>>
> >>> I pushed this patch with slightly improved style to:
> >>> https://github.com/Semihalf/edk2-platforms/commits/i2c-end-of-dxe
> >>>
> >>
> >> Thanks Marcin.
> >>
> >> I'll cherry pick some of those improvement, although I don't see the
> >> point of replace 'while (TRUE)' with 'while (1)'
> >>
> >
> > In the diff I did not change that:
> > https://github.com/Semihalf/edk2-platforms/commit/99b5775e9b1692d309ca25eac2ea86b0ca71437b#diff-044fee521a7731faf141e6c6d74fd83bR181
> >
>
> Fair enough.
>
> I have just grabbed your version, and applied it as
> c8000ecccc83..ed4cc8059ec5
>
Great, thanks!
Marcin
prev parent reply other threads:[~2020-06-05 15:07 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-04 21:35 [RFT PATCH edk2-platforms] Silicon/Marvell/MvI2cDxe: connect all I2C masters at EndOfDxe Ard Biesheuvel
2020-06-05 11:54 ` Marcin Wojtas
2020-06-05 13:07 ` Ard Biesheuvel
2020-06-05 13:27 ` Marcin Wojtas
2020-06-05 14:46 ` Ard Biesheuvel
2020-06-05 15:07 ` Marcin Wojtas [this message]
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=CAPv3WKdCnXhQkuNt27HFZXhzh5MA8GociB_knmygwX5dAuZ4HA@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