public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Ni, Ruiyu" <ruiyu.ni@Intel.com>
To: Laszlo Ersek <lersek@redhat.com>,
	edk2-devel-01 <edk2-devel@lists.01.org>
Cc: Chao Zhang <chao.b.zhang@intel.com>,
	Eric Dong <eric.dong@intel.com>,
	Jaben Carsey <jaben.carsey@intel.com>,
	Jiaxin Wu <jiaxin.wu@intel.com>,
	Jiewen Yao <jiewen.yao@intel.com>,
	Liming Gao <liming.gao@intel.com>,
	Michael D Kinney <michael.d.kinney@intel.com>,
	Roman Bacik <roman.bacik@broadcom.com>,
	Siyuan Fu <siyuan.fu@intel.com>, Star Zeng <star.zeng@intel.com>
Subject: Re: [PATCH 1/6] MdePkg/UefiLib: introduce EfiOpenFileByDevicePath()
Date: Fri, 27 Jul 2018 17:28:47 +0800	[thread overview]
Message-ID: <425d0b30-a322-23c7-6d56-8f23696b86cd@Intel.com> (raw)
In-Reply-To: <20180718205043.17574-2-lersek@redhat.com>

On 7/19/2018 4:50 AM, Laszlo Ersek wrote:

> +  //
> +  // Traverse the device path nodes relative to the filesystem.
> +  //
> +  while (!IsDevicePathEnd (*FilePath)) {
> +    //
> +    // Keep local variables that relate to the current device path node tightly
> +    // scoped.
> +    //
> +    FILEPATH_DEVICE_PATH *FilePathNode;
> +    CHAR16               *AlignedPathName;
> +    CHAR16               *PathName;
> +    EFI_FILE_PROTOCOL    *NextFile;
1. Not sure if it follows the coding style. I would prefer to move the 
definition to the beginning of the function.

> +
> +    if (DevicePathType (*FilePath) != MEDIA_DEVICE_PATH ||
> +        DevicePathSubType (*FilePath) != MEDIA_FILEPATH_DP) {
> +      Status = EFI_INVALID_PARAMETER;
> +      goto CloseLastFile;
> +    }
> +    FilePathNode = (FILEPATH_DEVICE_PATH *)*FilePath;
> +
> +    //
> +    // FilePathNode->PathName may be unaligned, and the UEFI specification
> +    // requires pointers that are passed to protocol member functions to be
> +    // aligned. Create an aligned copy of the pathname if necessary.
> +    //
> +    if ((UINTN)FilePathNode->PathName % sizeof *FilePathNode->PathName == 0) {
> +      AlignedPathName = NULL;
> +      PathName = FilePathNode->PathName;
> +    } else {
> +      AlignedPathName = AllocateCopyPool (
> +                          (DevicePathNodeLength (FilePathNode) -
> +                           SIZE_OF_FILEPATH_DEVICE_PATH),
> +                          FilePathNode->PathName
> +                          );
> +      if (AlignedPathName == NULL) {
> +        Status = EFI_OUT_OF_RESOURCES;
> +        goto CloseLastFile;
> +      }
> +      PathName = AlignedPathName;
> +    }
> +
> +    //
> +    // Open the next pathname fragment with EFI_FILE_MODE_CREATE masked out and
> +    // with Attributes set to 0.
> +    //
> +    Status = LastFile->Open (
> +                         LastFile,
> +                         &NextFile,
> +                         PathName,
> +                         OpenMode & ~(UINT64)EFI_FILE_MODE_CREATE,
> +                         0
> +                         );
2. As I said in previous mail, is it really needed?
Per spec it's not required. Per FAT driver implementation, it's also not 
required.

> +
> +    //
> +    // Retry with EFI_FILE_MODE_CREATE and the original Attributes if the first
> +    // attempt failed, and the caller specified EFI_FILE_MODE_CREATE.
> +    //
> +    if (EFI_ERROR (Status) && (OpenMode & EFI_FILE_MODE_CREATE) != 0) {
> +      Status = LastFile->Open (
> +                           LastFile,
> +                           &NextFile,
> +                           PathName,
> +                           OpenMode,
> +                           Attributes
> +                           );
> +    }
> +
> +    //
> +    // Release any AlignedPathName on both error and success paths; PathName is
> +    // no longer needed.
> +    //
> +    if (AlignedPathName != NULL) {
> +      FreePool (AlignedPathName);
> +    }
> +    if (EFI_ERROR (Status)) {
> +      goto CloseLastFile;
> +    }
> +
> +    //
> +    // Advance to the next device path node.
> +    //
> +    LastFile->Close (LastFile);
> +    LastFile = NextFile;
> +    *FilePath = NextDevicePathNode (FilePathNode);
> +  }
> +
> +  *File = LastFile;
> +  return EFI_SUCCESS;
> +
> +CloseLastFile:
> +  LastFile->Close (LastFile);
> +
> +  ASSERT (EFI_ERROR (Status));
3. ASSERT_EFI_ERROR (Status);

> +  return Status;
> +}
> 


-- 
Thanks,
Ray


  parent reply	other threads:[~2018-07-27  9:28 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-07-18 20:50 [PATCH 0/6] UefiLib: centralize OpenFileByDevicePath() and fix its bugs Laszlo Ersek
2018-07-18 20:50 ` [PATCH 1/6] MdePkg/UefiLib: introduce EfiOpenFileByDevicePath() Laszlo Ersek
2018-07-18 23:10   ` Yao, Jiewen
2018-07-19 10:47     ` Laszlo Ersek
2018-07-19 13:03       ` Yao, Jiewen
2018-07-24 17:20   ` Laszlo Ersek
2018-07-27  9:15   ` Ni, Ruiyu
2018-07-27  9:28   ` Ni, Ruiyu [this message]
2018-07-27 12:06     ` Laszlo Ersek
2018-07-30  1:54       ` Ni, Ruiyu
2018-07-30 14:13         ` Laszlo Ersek
2018-08-02  4:06           ` Gao, Liming
2018-08-02 14:45             ` Laszlo Ersek
2018-07-18 20:50 ` [PATCH 2/6] MdeModulePkg/RamDiskDxe: replace OpenFileByDevicePath() with UefiLib API Laszlo Ersek
2018-07-19 10:36   ` Zeng, Star
2018-07-19 13:20     ` Laszlo Ersek
2018-07-20 10:22       ` Zeng, Star
2018-07-18 20:50 ` [PATCH 3/6] NetworkPkg/TlsAuthConfigDxe: " Laszlo Ersek
2018-07-24 17:20   ` Laszlo Ersek
2018-07-25  0:30   ` Wu, Jiaxin
2018-07-18 20:50 ` [PATCH 4/6] SecurityPkg/SecureBootConfigDxe: " Laszlo Ersek
2018-07-24  5:09   ` Zhang, Chao B
2018-07-18 20:50 ` [PATCH 5/6] ShellPkg/UefiShellLib: drop DeviceHandle param of ShellOpenFileByDevicePath() Laszlo Ersek
2018-07-18 20:50 ` [PATCH 6/6] ShellPkg/UefiShellLib: rebase ShellOpenFileByDevicePath() to UefiLib API Laszlo Ersek
2018-07-18 21:15 ` [PATCH 0/6] UefiLib: centralize OpenFileByDevicePath() and fix its bugs Carsey, Jaben
2018-07-19  0:07   ` Ard Biesheuvel
2018-07-19 10:38     ` 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=425d0b30-a322-23c7-6d56-8f23696b86cd@Intel.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