qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Alexander Graf <graf@amazon.com>
To: "Akihiko Odaki" <akihiko.odaki@daynix.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	qemu-devel@nongnu.org
Cc: qemu-block@nongnu.org, qemu-arm@nongnu.org,
	"Cameron Esfahani" <dirty@apple.com>,
	"Stefan Hajnoczi" <stefanha@redhat.com>,
	"Michael S . Tsirkin" <mst@redhat.com>,
	"Kevin Wolf" <kwolf@redhat.com>,
	"Hanna Reitz" <hreitz@redhat.com>,
	"Marcel Apfelbaum" <marcel.apfelbaum@gmail.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Peter Maydell" <peter.maydell@linaro.org>,
	"Mads Ynddal" <mads@ynddal.dk>,
	"Daniel P . Berrangé" <berrange@redhat.com>,
	"Bernhard Beschow" <shentey@gmail.com>,
	"Gerd Hoffmann" <kraxel@redhat.com>
Subject: Re: [PATCH v2 01/12] build: Only define OS_OBJECT_USE_OBJC with gcc
Date: Thu, 31 Aug 2023 10:59:02 +0200	[thread overview]
Message-ID: <6e98c4e8-89ea-4fd1-a6cf-e9cdcf043482@amazon.com> (raw)
In-Reply-To: <c933da5b-92fe-47f6-ade1-9c1e2770aa23@daynix.com>


On 31.08.23 10:53, Akihiko Odaki wrote:
>
>
> On 2023/08/31 17:12, Philippe Mathieu-Daudé wrote:
>> On 30/8/23 18:14, Alexander Graf wrote:
>>> Recent versions of macOS use clang instead of gcc. The 
>>> OS_OBJECT_USE_OBJC
>>> define is only necessary when building with gcc. Let's not define it 
>>> when
>>> building with clang.
>>>
>>> With this patch, I can successfully include GCD headers in QEMU when
>>> building with clang.
>>>
>>> Signed-off-by: Alexander Graf <graf@amazon.com>
>>> ---
>>>   meson.build | 4 +++-
>>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/meson.build b/meson.build
>>> index 98e68ef0b1..0d6a0015a1 100644
>>> --- a/meson.build
>>> +++ b/meson.build
>>> @@ -224,7 +224,9 @@ qemu_ldflags = []
>>>   if targetos == 'darwin'
>>>     # Disable attempts to use ObjectiveC features in os/object.h since
>>> they
>>>     # won't work when we're compiling with gcc as a C compiler.
>>> -  qemu_common_flags += '-DOS_OBJECT_USE_OBJC=0'
>>> +  if compiler.get_id() == 'gcc'
>>> +    qemu_common_flags += '-DOS_OBJECT_USE_OBJC=0'
>>> +  endif
>>>   elif targetos == 'solaris'
>>>     # needed for CMSG_ macros in sys/socket.h
>>>     qemu_common_flags += '-D_XOPEN_SOURCE=600'
>>
>> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
>>
>
> Defining OS_OBJECT_USE_OBJC does not look like a proper solution.
> Looking at os/object.h, it seems OS_OBJECT_USE_OBJC is defined as 0 when:
> !defined(OS_OBJECT_HAVE_OBJC_SUPPORT) && (!defined(__OBJC__) ||
> defined(__OBJC_GC__))
>
> This means OS_OBJECT_USE_OBJC is always 0 if Objective-C is disabled. I
> also confirmed os/object.h will not use Objective-C features when
> compiled as C code on clang with the following command:
>
> clang -E -x -c - <<EOF
> #include <os/object.h>
> EOF
>
> If compilation fails with GCC when not defining OS_OBJECT_USE_OBJC, it
> probably means GCC incorrectly treats C code as Objective-C and that is
> the problem we should solve. I cannot confirm this theory however since
> I have only an Apple Silicon Mac that is incompatible with GCC.


My take on this was to make the gcc hack be a "legacy" thing that we put 
into its own corner, so that in a few years we can just drop it 
altogether. I don't really think it's worth wasting much time on this 
workaround and its potential compatibility with old macOS versions.


Alex





Amazon Development Center Germany GmbH
Krausenstr. 38
10117 Berlin
Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss
Eingetragen am Amtsgericht Charlottenburg unter HRB 149173 B
Sitz: Berlin
Ust-ID: DE 289 237 879



  reply	other threads:[~2023-08-31  8:59 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-30 16:14 [PATCH v2 00/12] Introduce new vmapple machine type Alexander Graf
2023-08-30 16:14 ` [PATCH v2 01/12] build: Only define OS_OBJECT_USE_OBJC with gcc Alexander Graf
2023-08-31  8:12   ` Philippe Mathieu-Daudé
2023-08-31  8:53     ` Akihiko Odaki
2023-08-31  8:59       ` Alexander Graf [this message]
2023-08-31 10:45         ` Akihiko Odaki
2023-08-30 16:14 ` [PATCH v2 02/12] hw/misc/pvpanic: Add MMIO interface Alexander Graf
2023-09-01  5:19   ` Mark Cave-Ayland
2023-08-30 16:14 ` [PATCH v2 03/12] hvf: Increase number of possible memory slots Alexander Graf
2023-08-30 16:14 ` [PATCH v2 04/12] hvf: arm: Ignore writes to CNTP_CTL_EL0 Alexander Graf
2023-08-31  8:13   ` Philippe Mathieu-Daudé
2023-08-30 16:14 ` [PATCH v2 05/12] hw: Add vmapple subdir Alexander Graf
2023-08-30 16:14 ` [PATCH v2 06/12] gpex: Allow more than 4 legacy IRQs Alexander Graf
2023-08-30 16:14 ` [PATCH v2 07/12] hw/vmapple/aes: Introduce aes engine Alexander Graf
2023-09-01  5:34   ` Mark Cave-Ayland
2023-08-30 16:14 ` [PATCH v2 08/12] hw/vmapple/bdif: Introduce vmapple backdoor interface Alexander Graf
2023-08-31 19:46   ` Stefan Hajnoczi
2023-09-01  5:40   ` Mark Cave-Ayland
2023-08-30 16:14 ` [PATCH v2 09/12] hw/vmapple/cfg: Introduce vmapple cfg region Alexander Graf
2023-09-01  5:46   ` Mark Cave-Ayland
2023-08-30 16:14 ` [PATCH v2 10/12] hw/vmapple/apple-gfx: Introduce ParavirtualizedGraphics.Framework support Alexander Graf
2023-09-23 21:04   ` Phil Dennis-Jordan
2023-08-30 16:14 ` [PATCH v2 11/12] hw/vmapple/virtio-blk: Add support for apple virtio-blk Alexander Graf
2023-08-31 20:03   ` Stefan Hajnoczi
2023-08-31 20:34     ` Philippe Mathieu-Daudé
2023-09-01  6:53   ` Mark Cave-Ayland
2023-08-30 16:14 ` [PATCH v2 12/12] hw/vmapple/vmapple: Add vmapple machine type Alexander Graf
2023-09-01  7:07   ` Mark Cave-Ayland
2023-10-12 11:46   ` Francesco Cagnin

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-all 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=6e98c4e8-89ea-4fd1-a6cf-e9cdcf043482@amazon.com \
    --to=graf@amazon.com \
    --cc=akihiko.odaki@daynix.com \
    --cc=berrange@redhat.com \
    --cc=dirty@apple.com \
    --cc=hreitz@redhat.com \
    --cc=kraxel@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mads@ynddal.dk \
    --cc=marcel.apfelbaum@gmail.com \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=philmd@linaro.org \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=shentey@gmail.com \
    --cc=stefanha@redhat.com \
    /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;
as well as URLs for NNTP newsgroup(s).