public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Laszlo Ersek" <lersek@redhat.com>
To: devel@edk2.groups.io, rebecca@bsdio.com
Cc: Jordan Justen <jordan.l.justen@intel.com>,
	Ard Biesheuvel <ard.biesheuvel@arm.com>,
	Leif Lindholm <leif.lindholm@linaro.org>,
	Michael Kinney <michael.d.kinney@intel.com>,
	Andrew Fish <afish@apple.com>, Peter Grehan <grehan@freebsd.org>
Subject: Re: [edk2-devel] [PATCH v3 2/6] OvmfPkg: Add QemuFwCfgLibNull
Date: Thu, 23 Apr 2020 10:21:53 +0200	[thread overview]
Message-ID: <b3bd046e-1762-243f-e1c6-01660ad6a8d4@redhat.com> (raw)
In-Reply-To: <20200421030955.114850-3-rebecca@bsdio.com>

On 04/21/20 05:09, Rebecca Cran wrote:
> Add a null implementation library for QemuFwCfgLib, in order to
> support building PciHostBridgeLib for bhyve.
> 
> Signed-off-by: Rebecca Cran <rebecca@bsdio.com>
> ---
>  .../Library/QemuFwCfgLib/QemuFwCfgLibNull.inf |  37 ++++
>  OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgNull.c  | 209 ++++++++++++++++++
>  2 files changed, 246 insertions(+)
>  create mode 100644 OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgLibNull.inf
>  create mode 100644 OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgNull.c
> 
> diff --git a/OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgLibNull.inf b/OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgLibNull.inf
> new file mode 100644
> index 0000000000..09f86c2b02
> --- /dev/null
> +++ b/OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgLibNull.inf
> @@ -0,0 +1,37 @@
> +## @file
> +#
> +#  Stateful, implicitly initialized fw_cfg library.

(1) This comment should be updated -- it should say (for example):

"Null implementation of the fw_cfg library".

(I could update this for you myself, but below there is another change
that I cannot do for you.)

> +#
> +#  Copyright (C) 2013, Red Hat, Inc.
> +#  Copyright (c) 2008 - 2012, Intel Corporation. All rights reserved.<BR>
> +#  Copyright (c) 2017, AMD Incorporated. All rights reserved.<BR>

(2) This is a new file; you should add your (C) on top (like you do in
patch#1). I can't add this for you.

> +#
> +#  SPDX-License-Identifier: BSD-2-Clause-Patent
> +#
> +##
> +
> +[Defines]
> +  INF_VERSION                    = 0x00010005
> +  BASE_NAME                      = QemuFwCfgLibNull
> +  FILE_GUID                      = B9D1A1F2-01E2-4732-982D-C7F9ED51AC6B
> +  MODULE_TYPE                    = BASE
> +  VERSION_STRING                 = 1.0
> +  LIBRARY_CLASS                  = QemuFwCfgLib
> +
> +#
> +# The following information is for reference only and not required by the build tools.
> +#
> +#  VALID_ARCHITECTURES           = IA32 X64
> +#

(3) Technically speaking, the null instance is suitable for all
architectures supported by edk2. I suggest simply dropping the
VALID_ARCHITECTURES comment block.

> +
> +[Sources]
> +  QemuFwCfgLibInternal.h

(4) We don't depend on this header file, please don't list it here.

> +  QemuFwCfgNull.c
> +
> +[Packages]
> +  MdePkg/MdePkg.dec
> +  MdeModulePkg/MdeModulePkg.dec

(5) I *think* MdeModulePkg.dec is not needed, as we don't consume any
content from MdeModulePkg. (The DebugLib class comes from MdePkg.)

> +  OvmfPkg/OvmfPkg.dec

Yes, this is necessary (because the lib class header that we'll include
comes from OvmfPkg).

> +
> +[LibraryClasses]
> +  DebugLib
> diff --git a/OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgNull.c b/OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgNull.c
> new file mode 100644
> index 0000000000..e2cc5f3406
> --- /dev/null
> +++ b/OvmfPkg/Library/QemuFwCfgLib/QemuFwCfgNull.c
> @@ -0,0 +1,209 @@
> +/** @file
> +
> +  Stateful and implicitly initialized fw_cfg library implementation.

(6) Same as (1).

> +
> +  Copyright (C) 2013, Red Hat, Inc.
> +  Copyright (c) 2011 - 2013, Intel Corporation. All rights reserved.<BR>
> +  Copyright (c) 2017, Advanced Micro Devices. All rights reserved.<BR>
> +
> +  SPDX-License-Identifier: BSD-2-Clause-Patent
> +**/

(7) Same as (2).

> +
> +#include <Uefi.h>
> +#include <Library/DebugLib.h>
> +#include <IndustryStandard/QemuFwCfg.h>
> +

(8) I believe <Uefi.h> and <IndustryStandard/QemuFwCfg.h> can be dropped.

In exchange, <Library/QemuFwCfgLib.h> should be included. That's because
the lib class header declares the functions that this lib instance will
define (= implement). The lib class header also pulls in all
dependencies that are needed for the API declarations.

When you include <Library/QemuFwCfgLib.h>, please add it under
DebugLib.h, to keep the #include list sorted.


> +/**
> +  Returns a boolean indicating if the firmware configuration interface
> +  is available or not.
> +
> +  This function may change fw_cfg state.
> +
> +  @retval    TRUE   The interface is available
> +  @retval    FALSE  The interface is not available
> +
> +**/
> +BOOLEAN
> +EFIAPI
> +QemuFwCfgIsAvailable (
> +  VOID
> +  )
> +{
> +  return FALSE;
> +}
> +
> +
> +/**
> +  Selects a firmware configuration item for reading.
> +
> +  Following this call, any data read from this item will start from
> +  the beginning of the configuration item's data.
> +
> +  @param[in] QemuFwCfgItem - Firmware Configuration item to read
> +
> +**/
> +VOID
> +EFIAPI
> +QemuFwCfgSelectItem (
> +  IN FIRMWARE_CONFIG_ITEM   QemuFwCfgItem
> +  )
> +{
> +  ASSERT (FALSE);
> +}
> +
> +
> +/**
> +  Reads firmware configuration bytes into a buffer
> +
> +  If called multiple times, then the data read will
> +  continue at the offset of the firmware configuration
> +  item where the previous read ended.
> +
> +  @param[in] Size - Size in bytes to read
> +  @param[in] Buffer - Buffer to store data into
> +
> +**/
> +VOID
> +EFIAPI
> +QemuFwCfgReadBytes (
> +  IN UINTN                  Size,
> +  IN VOID                   *Buffer  OPTIONAL
> +  )
> +{
> +  ASSERT (FALSE);
> +}
> +
> +
> +/**
> +  Writes firmware configuration bytes from a buffer
> +
> +  If called multiple times, then the data written will
> +  continue at the offset of the firmware configuration
> +  item where the previous write ended.
> +
> +  @param[in] Size - Size in bytes to write
> +  @param[in] Buffer - Buffer to read data from
> +
> +**/
> +VOID
> +EFIAPI
> +QemuFwCfgWriteBytes (
> +  IN UINTN                  Size,
> +  IN VOID                   *Buffer
> +  )
> +{
> +  ASSERT (FALSE);
> +}
> +
> +
> +/**
> +  Skip bytes in the firmware configuration item.
> +
> +  Increase the offset of the firmware configuration item without transferring
> +  bytes between the item and a caller-provided buffer. Subsequent read, write
> +  or skip operations will commence at the increased offset.
> +
> +  @param[in] Size  Number of bytes to skip.
> +**/
> +VOID
> +EFIAPI
> +QemuFwCfgSkipBytes (
> +  IN UINTN                  Size
> +  )
> +{
> +  ASSERT (FALSE);
> +}
> +
> +
> +/**
> +  Reads a UINT8 firmware configuration value
> +
> +  @return    Value of Firmware Configuration item read
> +
> +**/
> +UINT8
> +EFIAPI
> +QemuFwCfgRead8 (
> +  VOID
> +  )
> +{
> +  ASSERT (FALSE);
> +  return 0;
> +}
> +
> +
> +/**
> +  Reads a UINT16 firmware configuration value
> +
> +  @return    Value of Firmware Configuration item read
> +
> +**/
> +UINT16
> +EFIAPI
> +QemuFwCfgRead16 (
> +  VOID
> +  )
> +{
> +  ASSERT (FALSE);
> +  return 0;
> +}
> +
> +
> +/**
> +  Reads a UINT32 firmware configuration value
> +
> +  @return    Value of Firmware Configuration item read
> +
> +**/
> +UINT32
> +EFIAPI
> +QemuFwCfgRead32 (
> +  VOID
> +  )
> +{
> +  ASSERT (FALSE);
> +  return 0;
> +}
> +
> +
> +/**
> +  Reads a UINT64 firmware configuration value
> +
> +  @return    Value of Firmware Configuration item read
> +
> +**/
> +UINT64
> +EFIAPI
> +QemuFwCfgRead64 (
> +  VOID
> +  )
> +{
> +  ASSERT (FALSE);
> +  return 0;
> +}
> +
> +
> +/**
> +  Find the configuration item corresponding to the firmware configuration file.
> +
> +  @param[in]  Name - Name of file to look up.
> +  @param[out] Item - Configuration item corresponding to the file, to be passed
> +                     to QemuFwCfgSelectItem ().
> +  @param[out] Size - Number of bytes in the file.
> +
> +  @return    RETURN_SUCCESS       If file is found.
> +             RETURN_NOT_FOUND     If file is not found.
> +             RETURN_UNSUPPORTED   If firmware configuration is unavailable.
> +
> +**/
> +RETURN_STATUS
> +EFIAPI
> +QemuFwCfgFindFile (
> +  IN   CONST CHAR8           *Name,
> +  OUT  FIRMWARE_CONFIG_ITEM  *Item,
> +  OUT  UINTN                 *Size
> +  )
> +{
> +  return RETURN_UNSUPPORTED;
> +}
> +
> 

These parts are fine.

Thanks!
Laszlo


  reply	other threads:[~2020-04-23  8:22 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-21  3:09 [PATCH v3 0/6] Add BhyvePkg, to support the bhyve hypervisor Rebecca Cran
2020-04-21  3:09 ` [PATCH v3 1/6] OvmfPkg: Add bhyve support into AcpiTimerLib Rebecca Cran
2020-04-23  7:56   ` [edk2-devel] " Laszlo Ersek
2020-04-21  3:09 ` [PATCH v3 2/6] OvmfPkg: Add QemuFwCfgLibNull Rebecca Cran
2020-04-23  8:21   ` Laszlo Ersek [this message]
2020-04-21  3:09 ` [PATCH v3 3/6] OvmfPkg: Add VBE2 mode info structure to LegacyVgaBios.h Rebecca Cran
2020-04-23  8:05   ` [edk2-devel] " Laszlo Ersek
2020-04-21  3:09 ` [PATCH v3 4/6] Add BhyvePkg, to support the bhyve hypervisor Rebecca Cran
2020-04-23  9:19   ` [edk2-devel] " Laszlo Ersek
2020-04-23  9:42     ` Laszlo Ersek
2020-04-24  5:54       ` Rebecca Cran
2020-04-24 12:22         ` Laszlo Ersek
2020-04-23 20:08     ` Rebecca Cran
2020-04-24 10:11       ` Laszlo Ersek
2020-04-24 16:00         ` Rebecca Cran
2020-04-21  3:09 ` [PATCH v3 5/6] BhyvePkg: Add PlatformPei Rebecca Cran
2020-04-23  9:24   ` [edk2-devel] " Laszlo Ersek
2020-04-21  3:09 ` [PATCH v3 6/6] BhyvePkg: Add AcpiPlatformDxe Rebecca Cran
2020-04-23  9:44   ` [edk2-devel] " Laszlo Ersek
2020-04-21 15:27 ` [PATCH v3 0/6] Add BhyvePkg, to support the bhyve hypervisor Laszlo Ersek
2020-04-21 15:38   ` Rebecca Cran
2020-04-22 15:21     ` Laszlo Ersek
2020-04-22 16:48       ` Rebecca Cran
2020-04-24 10:11         ` Laszlo Ersek

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=b3bd046e-1762-243f-e1c6-01660ad6a8d4@redhat.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