From patchwork Mon Jul 15 21:06:55 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Phil Dennis-Jordan X-Patchwork-Id: 13735393 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 5AB09C3DA42 for ; Wed, 17 Jul 2024 11:14:31 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1sU2TL-0004ho-1Z; Wed, 17 Jul 2024 07:05:23 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1sU2SP-0000be-Ma for qemu-devel@nongnu.org; Wed, 17 Jul 2024 07:04:29 -0400 Received: from 89-104-8-17.customer.bnet.at ([89.104.8.17] helo=intel-mbp.local) by eggs.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1sU2SN-0006zq-Dw for qemu-devel@nongnu.org; Wed, 17 Jul 2024 07:04:25 -0400 Received: by intel-mbp.local (Postfix, from userid 501) id 40836379637; Mon, 15 Jul 2024 23:07:38 +0200 (CEST) From: Phil Dennis-Jordan To: qemu-devel@nongnu.org, pbonzini@redhat.com, agraf@csgraf.de, graf@amazon.com, marcandre.lureau@redhat.com, berrange@redhat.com, thuth@redhat.com, philmd@linaro.org, peter.maydell@linaro.org, akihiko.odaki@daynix.com, phil@philjordan.eu, lists@philjordan.eu Subject: [PATCH 16/26] hw/display/apple-gfx: Asynchronous MMIO writes on x86-64 Date: Mon, 15 Jul 2024 23:06:55 +0200 Message-Id: <20240715210705.32365-17-phil@philjordan.eu> X-Mailer: git-send-email 2.39.3 (Apple Git-146) In-Reply-To: <20240715210705.32365-1-phil@philjordan.eu> References: <20240715210705.32365-1-phil@philjordan.eu> MIME-Version: 1.0 Received-SPF: none client-ip=89.104.8.17; envelope-from=phil@intel-mbp.local; helo=intel-mbp.local X-Spam_score_int: -6 X-Spam_score: -0.7 X-Spam_bar: / X-Spam_report: (-0.7 / 5.0 requ) BAYES_00=-1.9, HEADER_FROM_DIFFERENT_DOMAINS=0.001, KHOP_HELO_FCRDNS=0.261, NO_DNS_FOR_FROM=0.001, RDNS_DYNAMIC=0.982, SPF_HELO_NONE=0.001, SPF_NONE=0.001, TVD_RCVD_IP=0.001 autolearn=no autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org This change ensures that the MMIO write calls into the PVG framework are performed asynchronously on a background dispatch queue. Without this, we rapidly run into re-entrant MMIO issues. This problem only seems to exist on x86-64 hosts. Conversely, doing it async on arm64/vmapple causes other issues, so we're left with 2 different implementations. Signed-off-by: Phil Dennis-Jordan --- hw/display/apple-gfx.m | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/hw/display/apple-gfx.m b/hw/display/apple-gfx.m index 806feb58fa..48463e5a1f 100644 --- a/hw/display/apple-gfx.m +++ b/hw/display/apple-gfx.m @@ -67,15 +67,28 @@ static uint64_t apple_gfx_read(void *opaque, hwaddr offset, unsigned size) return res; } -static void apple_gfx_write(void *opaque, hwaddr offset, uint64_t val, unsigned size) +static void apple_gfx_write(void *opaque, hwaddr offset, uint64_t val, + unsigned size) { AppleGFXState *s = opaque; trace_apple_gfx_write(offset, val); +#ifdef __x86_64__ + id dev = s->pgdev; + dispatch_queue_t bg_queue = NULL; + + bg_queue = dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_DEFAULT, 0ul); + [dev retain]; + dispatch_async(bg_queue, ^{ + [dev mmioWriteAtOffset:offset value:val]; + [dev release]; + }); +#else bql_unlock(); [s->pgdev mmioWriteAtOffset:offset value:val]; bql_lock(); +#endif } static const MemoryRegionOps apple_gfx_ops = {