From: "Laszlo Ersek" <lersek@redhat.com>
To: devel@edk2.groups.io, ard.biesheuvel@arm.com, virtio-fs@redhat.com
Cc: "Jordan Justen" <jordan.l.justen@intel.com>,
"Leif Lindholm" <leif@nuviainc.com>,
"Philippe Mathieu-Daudé" <philmd@redhat.com>,
"Liming Gao (Byosoft address)" <gaoliming@byosoft.com.cn>
Subject: Re: [edk2-devel] [edk2 PATCH 00/48] ArmVirtPkg, OvmfPkg: virtio filesystem driver
Date: Mon, 21 Dec 2020 19:02:07 +0100 [thread overview]
Message-ID: <1f510693-832b-8b57-fce8-e415aaf4e238@redhat.com> (raw)
In-Reply-To: <ba471de0-0744-9d65-2107-29471ddc8d20@arm.com>
On 12/21/20 11:10, Ard Biesheuvel wrote:
> On 12/21/20 2:46 AM, Laszlo Ersek wrote:
>> On 12/20/20 11:15, Ard Biesheuvel wrote:
>>> On 12/20/20 1:09 AM, Laszlo Ersek wrote:
>>
>>>> The only thing I couldn't test that way was (obviously) building on
>>>> Windows. So clearly now that I've tried merging the series at
>>>> <https://github.com/tianocore/edk2/pull/1248>, that's what fails.
>>>> Because, this is the first time that EmbeddedPkg/TimeBaseLib is seen
>>>> by the Visual Studio compiler, and Visual Studio complains:
>>>>
>>>>> ERROR - Compiler #2220 from
>>>>> d:\a\1\s\EmbeddedPkg\Library\TimeBaseLib\TimeBaseLib.c(111): the
>>>>> following warning is treated as an error
>>>>> WARNING - Compiler #4244 from
>>>>> d:\a\1\s\EmbeddedPkg\Library\TimeBaseLib\TimeBaseLib.c(111): '=':
>>>>> conversion from 'UINTN' to 'UINT32', possible loss of data
>>>>
>>>> It happens to be correct:
>>>>
>>>> 99 /**
>>>> 100 Converts EFI_TIME to Epoch seconds (elapsed since 1970 JANUARY 01, 00:00:00 UTC)
>>>> 101 **/
>>>> 102 UINT32
>>>> 103 EFIAPI
>>>> 104 EfiTimeToEpoch (
>>>> 105 IN EFI_TIME *Time
>>>> 106 )
>>>> 107 {
>>>> 108 UINT32 EpochDays; // Number of days elapsed since EPOCH_JULIAN_DAY
>>>> 109 UINT32 EpochSeconds;
>>>> 110
>>>> 111 EpochDays = EfiGetEpochDays (Time);
>>>> 112
>>>> 113 EpochSeconds = (EpochDays * SEC_PER_DAY) + ((UINTN)Time->Hour * SEC_PER_HOUR) + (Time->Minute * SEC_PER_MIN) + Time->Second;
>>>> 114
>>>> 115 return EpochSeconds;
>>>> 116 }
>>>>
>>>> This was discussed on the list earlier, when the function was duplicated
>>>> for the HTTP dynamic command:
>>>>
>>>> - https://edk2.groups.io/g/devel/message/64933
>>>> - https://edk2.groups.io/g/devel/message/65186
>>>>
>>>> Compare EfiTimeToEpoch() in
>>>> "ShellPkg/DynamicCommand/HttpDynamicCommand/Http.c".
>>>>
>>>> 430 STATIC
>>>> 431 UINTN
>>>> 432 EFIAPI
>>>> 433 EfiTimeToEpoch (
>>>> 434 IN EFI_TIME *Time
>>>> 435 )
>>>> 436 {
>>>> 437 //
>>>> 438 // Number of days elapsed since EPOCH_JULIAN_DAY.
>>>> 439 //
>>>> 440 UINTN EpochDays;
>>>> 441 UINTN EpochSeconds;
>>>> 442
>>>> 443 EpochDays = EfiGetEpochDays (Time);
>>>> 444
>>>> 445 EpochSeconds = (EpochDays * SEC_PER_DAY) +
>>>> 446 ((UINTN)Time->Hour * SEC_PER_HOUR) +
>>>> 447 (Time->Minute * SEC_PER_MIN) + Time->Second;
>>>> 448
>>>> 449 return EpochSeconds;
>>>> 450 }
>>>>
>>>> So, in order to merge this series, I'll first have to fix
>>>> EfiTimeToEpoch() in EmbeddedPkg :(
>>>>
>>>> I wish the fixed version in
>>>> "ShellPkg/DynamicCommand/HttpDynamicCommand/Http.c" had been contributed
>>>> back to EmbeddedPkg.
>>>>
>>>> Anyway, that is for 2021.
>>
>> (given the fantastic opportunities provided by the COVID-19 pandemic, to
>> catch up with family and friends around Christmas /s, I might as well
>> treat the discussion of this patch series during my PTO as something
>> welcome that takes my mind off of how much I miss the people I can't
>> meet this year)
>>
>>> Thanks for clearing up the outstanding questions.
>>>
>>> For the EmbeddedPkg change from UINT32 to UINTN
>>>
>>> Acked-by: Ard Biesheuvel <ard.biesheuvel@arm.com>
>>>
>>> although I suppose you will have to preserve the prototype, so adding
>>> a (UINT32) cast to line 111 is probably the best approach here.
>>
>> I ended up installing a brand new Windows 10 + VS2019 virtual machine,
>> as I got fed up with the nasty surprises in the PRs (and I refuse to
>> publish my work-in-progress branch just for the sake of setting off CI
>> on it).
>>
>> Two consequences:
>>
>> (1) I'll squash the following trivial updates into two of the patches,
>> for my next merge request attempt:
>>
>> 5: bb254f89067a ! 7: 17a76bbec317 OvmfPkg/VirtioFsDxe: add a scatter-gather list data type
>> @@ -20,6 +20,8 @@
>> Signed-off-by: Laszlo Ersek <lersek@redhat.com>
>> Message-Id: <20201216211125.19496-6-lersek@redhat.com>
>> Acked-by: Ard Biesheuvel <ard.biesheuvel@arm.com>
>> + [lersek@redhat.com: suppress useless VS2019 warning about signed/unsigned
>> + comparison in VirtioFsSgListsValidate()]
>>
>> diff --git a/OvmfPkg/VirtioFsDxe/VirtioFsDxe.h b/OvmfPkg/VirtioFsDxe/VirtioFsDxe.h
>> --- a/OvmfPkg/VirtioFsDxe/VirtioFsDxe.h
>> @@ -213,7 +215,7 @@
>> + // can be added to the virtio queue, after the other descriptors added
>> + // previously.
>> + //
>> -+ if (SgList->NumVec > MAX_UINT16 - DescriptorsNeeded ||
>> ++ if (SgList->NumVec > (UINTN)(MAX_UINT16 - DescriptorsNeeded) ||
>> + DescriptorsNeeded + SgList->NumVec > VirtioFs->QueueSize) {
>> + return EFI_UNSUPPORTED;
>> + }
>>
>> and
>>
>> 46: 4f42ecc2d9bb ! 48: b807d3c0b54b OvmfPkg/VirtioFsDxe: add helper for determining access time updates
>> @@ -12,6 +12,8 @@
>> Signed-off-by: Laszlo Ersek <lersek@redhat.com>
>> Message-Id: <20201216211125.19496-47-lersek@redhat.com>
>> Acked-by: Ard Biesheuvel <ard.biesheuvel@arm.com>
>> + [lersek@redhat.com: suppress bogus VS2019 warning about lack of
>> + initialization for ZeroTime]
>>
>> diff --git a/OvmfPkg/VirtioFsDxe/VirtioFsDxe.h b/OvmfPkg/VirtioFsDxe/VirtioFsDxe.h
>> --- a/OvmfPkg/VirtioFsDxe/VirtioFsDxe.h
>> @@ -92,7 +94,7 @@
>> + EFI_TIME *Time[3];
>> + EFI_TIME *NewTime[ARRAY_SIZE (Time)];
>> + UINTN Idx;
>> -+ STATIC CONST EFI_TIME ZeroTime;
>> ++ STATIC CONST EFI_TIME ZeroTime = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
>> + BOOLEAN Change[ARRAY_SIZE (Time)];
>> + UINT64 Seconds[ARRAY_SIZE (Time)];
>> +
>>
>> (2) I've written three patches in total for fixing EfiTimeToEpoch(),
>> including the prototype:
>>
>> (2a) edk2-platforms:
>>
>> 1 Silicon/Marvell/RealTimeClockLib: make EpochSeconds, WakeupSeconds UINTN
>>
>> Silicon/Marvell/Armada7k8k/Library/RealTimeClockLib/RealTimeClockLib.c | 14 ++++++++++----
>> 1 file changed, 10 insertions(+), 4 deletions(-)
>>
>> (other RealTimeClockLib instances / EfiTimeToEpoch() callers in
>> edk2-platforms need no updates; furthermore, edk2-non-osi contains no
>> EfiTimeToEpoch() calls at all)
>>
>> (2b) edk2:
>>
>> 1 ArmPlatformPkg/PL031RealTimeClockLib: cast EfiTimeToEpoch() val. to UINT32
>>
>> ArmPlatformPkg/Library/PL031RealTimeClockLib/PL031RealTimeClockLib.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> 2 EmbeddedPkg/TimeBaseLib: remove useless truncation to 32-bit
>>
>> EmbeddedPkg/Include/Library/TimeBaseLib.h | 2 +-
>> EmbeddedPkg/Library/TimeBaseLib/TimeBaseLib.c | 6 +++---
>> 2 files changed, 4 insertions(+), 4 deletions(-)
>>
>> If you're reading this before 2021, please let me know if you'd tolerate
>> receiving these patches for approval still in 2020. (The edk2-platforms
>> patch theoretically belongs to Leif and Marcin, but if Leif has stopped
>> consuming work email (which we all should have by now, I guess...), then
>> I believe you could ACK that patch in Leif's stead.)
>>
>
> No problem.
>
Merged as commit range c06635ea3f4b..35ed29f207fd, via
<https://github.com/tianocore/edk2/pull/1256>.
I've also added <https://bugzilla.tianocore.org/show_bug.cgi?id=3097> to
<https://github.com/tianocore/tianocore.github.io/wiki/EDK-II-Release-Planning#proposed-features>.
Thank you, Ard!
Laszlo
prev parent reply other threads:[~2020-12-21 18:02 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-16 21:10 [edk2 PATCH 00/48] ArmVirtPkg, OvmfPkg: virtio filesystem driver Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 01/48] OvmfPkg: introduce VirtioFsDxe Laszlo Ersek
2020-12-18 17:42 ` Ard Biesheuvel
2020-12-18 18:13 ` [Virtio-fs] " Dr. David Alan Gilbert
2020-12-19 21:16 ` Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 02/48] ArmVirtPkg: include VirtioFsDxe in the ArmVirtQemu* platforms Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 03/48] OvmfPkg/VirtioFsDxe: DriverBinding: open VirtioDevice, install SimpleFs Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 04/48] OvmfPkg/VirtioFsDxe: implement virtio device (un)initialization Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 05/48] OvmfPkg/VirtioFsDxe: add a scatter-gather list data type Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 06/48] OvmfPkg/VirtioFsDxe: introduce the basic FUSE request/response headers Laszlo Ersek
2020-12-17 11:49 ` [Virtio-fs] " Dr. David Alan Gilbert
2020-12-17 13:57 ` Laszlo Ersek
2020-12-17 14:06 ` Dr. David Alan Gilbert
2020-12-17 14:32 ` Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 07/48] OvmfPkg/VirtioFsDxe: map "errno" values to EFI_STATUS Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 08/48] OvmfPkg/VirtioFsDxe: submit the FUSE_INIT request to the device Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 09/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_OPENDIR Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 10/48] OvmfPkg/VirtioFsDxe: add shared wrapper for FUSE_RELEASE / FUSE_RELEASEDIR Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 11/48] OvmfPkg/VirtioFsDxe: implement EFI_SIMPLE_FILE_SYSTEM_PROTOCOL.OpenVolume() Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 12/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_FORGET Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 13/48] OvmfPkg/VirtioFsDxe: add a shared wrapper for FUSE_FSYNC / FUSE_FSYNCDIR Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 14/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_FLUSH Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 15/48] OvmfPkg/VirtioFsDxe: flush, sync, release and forget in Close() / Delete() Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 16/48] OvmfPkg/VirtioFsDxe: add helper for appending and sanitizing paths Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 17/48] OvmfPkg/VirtioFsDxe: manage path lifecycle in OpenVolume, Close, Delete Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 18/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_OPEN Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 19/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_MKDIR Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 20/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_CREATE Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 21/48] OvmfPkg/VirtioFsDxe: convert FUSE inode attributes to EFI_FILE_INFO Laszlo Ersek
2020-12-16 21:10 ` [edk2 PATCH 22/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_LOOKUP Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 23/48] OvmfPkg/VirtioFsDxe: split canon. path into last parent + last component Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 24/48] OvmfPkg/VirtioFsDxe: add a shared wrapper for FUSE_UNLINK / FUSE_RMDIR Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 25/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_GETATTR Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 26/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.Open() Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 27/48] OvmfPkg/VirtioFsDxe: erase the dir. entry in EFI_FILE_PROTOCOL.Delete() Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 28/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_STATFS Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 29/48] OvmfPkg/VirtioFsDxe: add helper for formatting UEFI basenames Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 30/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.GetInfo() Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 31/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.GetPosition, .SetPosition Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 32/48] OvmfPkg/VirtioFsDxe: add a shared wrapper for FUSE_READ / FUSE_READDIRPLUS Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 33/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.Read() for regular files Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 34/48] OvmfPkg/VirtioFsDxe: convert FUSE dirent filename to EFI_FILE_INFO Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 35/48] OvmfPkg/VirtioFsDxe: add EFI_FILE_INFO cache fields to VIRTIO_FS_FILE Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 36/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.Read() for directories Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 37/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.Flush() Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 38/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_WRITE Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 39/48] OvmfPkg/VirtioFsDxe: implement EFI_FILE_PROTOCOL.Write() Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 40/48] OvmfPkg/VirtioFsDxe: handle the volume label in EFI_FILE_PROTOCOL.SetInfo Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 41/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_RENAME2 Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 42/48] OvmfPkg/VirtioFsDxe: add helper for composing rename/move destination path Laszlo Ersek
2020-12-18 17:39 ` Ard Biesheuvel
2020-12-19 22:40 ` Laszlo Ersek
2020-12-19 22:54 ` Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 43/48] OvmfPkg/VirtioFsDxe: handle file rename/move in EFI_FILE_PROTOCOL.SetInfo Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 44/48] OvmfPkg/VirtioFsDxe: implement the wrapper function for FUSE_SETATTR Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 45/48] OvmfPkg/VirtioFsDxe: add helper for determining file size update Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 46/48] OvmfPkg/VirtioFsDxe: add helper for determining access time updates Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 47/48] OvmfPkg/VirtioFsDxe: add helper for determining file mode bits update Laszlo Ersek
2020-12-16 21:11 ` [edk2 PATCH 48/48] OvmfPkg/VirtioFsDxe: handle attribute updates in EFI_FILE_PROTOCOL.SetInfo Laszlo Ersek
2020-12-18 17:44 ` [edk2 PATCH 00/48] ArmVirtPkg, OvmfPkg: virtio filesystem driver Ard Biesheuvel
2020-12-20 0:09 ` Laszlo Ersek
2020-12-20 10:15 ` Ard Biesheuvel
2020-12-21 1:46 ` Laszlo Ersek
2020-12-21 10:10 ` Ard Biesheuvel
2020-12-21 18:02 ` Laszlo Ersek [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=1f510693-832b-8b57-fce8-e415aaf4e238@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