From mboxrd@z Thu Jan 1 00:00:00 1970 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 lists1p.gnu.org (lists1p.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 1D1A6C61DFD for ; Mon, 31 Aug 2026 18:36:42 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists1p.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1x16rl-0003iL-Kn; Mon, 31 Aug 2026 14:36:21 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists1p.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1x16rh-0003hQ-04; Mon, 31 Aug 2026 14:36:17 -0400 Received: from mx0a-001b2d01.pphosted.com ([148.163.156.1]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1x16ra-0002zG-UM; Mon, 31 Aug 2026 14:36:13 -0400 Received: from pps.filterd (m0353729.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.18.1.11/8.18.1.11) with ESMTP id 67VIVbIG1235188; Mon, 31 Aug 2026 18:36:09 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ibm.com; h=cc :content-transfer-encoding:content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to; s=pp1; bh=XpASYh qHXy7bJtHUKM9K69uQsmwkvvRshP/kZFHb6fU=; b=dAtUUha3ZLAAaO32RK4nXk 8fZW7x2A/QTF8gkWL2sNbuvJMWYFGpReWGU2Q7Vj51itXuPLOuZNx1pCgQpFB5l8 3bSV+EiWr3m/hz5hNRn8lxdarOHntCA+/5Qn/uGTjm+VzC2SsAEf11eBXDcGXWFo 2przvvReCR6cnfMIddNrKufGh/u+tR+Kf7sLgA+qNJpSVFvZnX0pSQli8I0bF0E/ HD5xydDSXYnSGYmaGviKsqx8aKioPx4j9iFsE+zJQg8iqQGvCbIB/cxvSAhlJ3wr 2pOtjvA9zkMrEz6KhL/AWQei60xw0TDoe1cA1u3SzE4bK7aWuce1nhBRaQ9oMb4w == Received: from ppma23.wdc07v.mail.ibm.com (5d.69.3da9.ip4.static.sl-reverse.com [169.61.105.93]) by mx0a-001b2d01.pphosted.com (PPS) with ESMTPS id 4gbq3r34ya-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 31 Aug 2026 18:36:08 +0000 (GMT) Received: from pps.filterd (ppma23.wdc07v.mail.ibm.com [127.0.0.1]) by ppma23.wdc07v.mail.ibm.com (8.18.1.7/8.18.1.7) with ESMTP id 67VIU9IN022170; Mon, 31 Aug 2026 18:36:07 GMT Received: from smtprelay06.wdc07v.mail.ibm.com ([172.16.1.73]) by ppma23.wdc07v.mail.ibm.com (PPS) with ESMTPS id 4gcb8h7da0-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 31 Aug 2026 18:36:07 +0000 (GMT) Received: from smtpav02.wdc07v.mail.ibm.com (smtpav02.wdc07v.mail.ibm.com [10.39.53.229]) by smtprelay06.wdc07v.mail.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 67VIa6pR32965150 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 31 Aug 2026 18:36:06 GMT Received: from smtpav02.wdc07v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 4481358059; Mon, 31 Aug 2026 18:36:06 +0000 (GMT) Received: from smtpav02.wdc07v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 0216058058; Mon, 31 Aug 2026 18:36:05 +0000 (GMT) Received: from [9.61.73.66] (unknown [9.61.73.66]) by smtpav02.wdc07v.mail.ibm.com (Postfix) with ESMTP; Mon, 31 Aug 2026 18:36:04 +0000 (GMT) Message-ID: Date: Mon, 31 Aug 2026 14:36:04 -0400 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH v1 4/7] pc-bios/s390-ccw: Add support for virtio-net-pci IPL To: Jared Rossi , qemu-devel@nongnu.org, qemu-s390x@nongnu.org Cc: mst@redhat.com, borntraeger@linux.ibm.com, jjherne@linux.ibm.com, cohuck@redhat.com, farman@linux.ibm.com, mjrosato@linux.ibm.com, pasic@linux.ibm.com, farosas@suse.de, lvivier@redhat.com, pbonzini@redhat.com References: <20260818205324.580199-1-zycai@linux.ibm.com> <20260818205324.580199-5-zycai@linux.ibm.com> <5a7ed440-ed57-448a-9baa-4626464ee9f8@linux.ibm.com> Content-Language: en-US From: Zhuoying Cai In-Reply-To: <5a7ed440-ed57-448a-9baa-4626464ee9f8@linux.ibm.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-TM-AS-GCONF: 00 X-Authority-Analysis: v=2.4 cv=EIc2FVZC c=1 sm=1 tr=0 ts=6a95c998 cx=c_pps a=3Bg1Hr4SwmMryq2xdFQyZA==:117 a=3Bg1Hr4SwmMryq2xdFQyZA==:17 a=IkcTkHD0fZMA:10 a=Sv0fKeRqtYgA:10 a=VkNPw1HP01LnGYTKEx00:22 a=RnoormkPH1_aCDwRdu11:22 a=uAbxVGIbfxUO_5tXvNgY:22 a=VnNF1IyMAAAA:8 a=ctsVLiZ_AAAA:8 a=kIP0pMTcak2oO-wSoEIA:9 a=3ZKOabzyN94A:10 a=QEXdDO2ut3YA:10 a=3VCIHQLkQthbvTRuo98s:22 X-Proofpoint-Spam-Details-Enc: AW1haW4tMjYwODMxMDE2MCBTYWx0ZWRfX/wD/24MZvNQ5 F81n0NxlvXAjWhGzC2pRAi2tlMVaRo+DSDmgU3iJT0vrCBAx0rwKF8IUiOhnxx74bfxHsTiV1B+ AopeIB2FFtrXgMy7282S77mRyklT6+89daoo5KZFDwCbRa4lPUboWBfW5ahIqrWLcAGuNKU5kRT gNEHZ2pYKw9blXbRC4Je/Ib5WJDwJ9f/uTFMA+Q/V2IkGC9VvPz2ODVwf0nFVLJ6WRuh+dfE/51 2U5EqfZDdhXFAgFrrMGMKKSbYJTeeVkfTp0fqow3n9jqXYn+ItY6u7Sq+KZozxJGEqbNtxleQNe ofPUR1WJx513FUo7V85Sap7v17N5uHsI0ykdSEv8olu/pWeDvgHoXCk1nYfGlqlDMpnHkTpk7AK rps2+g0AGdUxt/fJ4RhDVn3ZPDY1mvX6Qn9kErdnuBotdXHCbzD3i1rPDbDrtin68cTc3iV1lAs iq/f1ulOX5fjcwf96Pw== X-Proofpoint-GUID: tJFC-WnqB4mQkan6klJlynnR6JGq17I5 X-Proofpoint-ORIG-GUID: tJFC-WnqB4mQkan6klJlynnR6JGq17I5 X-Proofpoint-Spam-Info: AW1haW4tMjYwODMxMDE2MCBTYWx0ZWRfX8lc0CBbc5LfM baZ+Fcv7luYqwKEepgr4EHlfhqxOYoN+GiOeCHwBdi0ByhzvQWnsdzUqfW87iL2TXGHCKYB9EBk fk3MV0f9wMYmbI/KKfWo4mF2Rk626HI= X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.293,Aquarius:18.0.1176,Hydra:6.1.134,FMLib:17.12.100.49 definitions=2026-08-31_06,2026-08-31_01,2025-10-01_01 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 bulkscore=0 impostorscore=0 suspectscore=0 priorityscore=1501 clxscore=1015 phishscore=0 spamscore=0 adultscore=0 lowpriorityscore=0 malwarescore=0 classifier=typeunknown authscore=0 authtc= authcc= route=outbound adjust=0 reason=mlx scancount=1 engine=8.22.0-2606150000 definitions=main-2608310160 Received-SPF: pass client-ip=148.163.156.1; envelope-from=zycai@linux.ibm.com; helo=mx0a-001b2d01.pphosted.com X-Spam_score_int: -26 X-Spam_score: -2.7 X-Spam_bar: -- X-Spam_report: (-2.7 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_LOW=-0.7, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: qemu development 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 Thanks for the feedback! On 8/26/26 12:44 PM, Jared Rossi wrote: > > > On 8/18/26 4:53 PM, Zhuoying Cai wrote: >> Enable network booting via virtio-net-pci by implementing PCI transport >> support for virtio-net. >> >> This patch also adds endianness handling for virtio PCI ring operations in >> little-endian to ensure correct behavior on s390x. > > Endianness handling was actually added in the previous patch.  Is this a > misplaced message? Or perhaps it should say extends, rather than adds? > Good catch. It should be "extends" rather than "adds". I'll update the commit message. >> Signed-off-by: Zhuoying Cai >> --- >> pc-bios/s390-ccw/main.c | 1 + >> pc-bios/s390-ccw/virtio-net.c | 37 +++++++++++++++++++++++------- >> pc-bios/s390-ccw/virtio-pci.c | 43 +++++++++++++++++++++++++++++++++++ >> pc-bios/s390-ccw/virtio-pci.h | 2 ++ >> 4 files changed, 75 insertions(+), 8 deletions(-) >> >> diff --git a/pc-bios/s390-ccw/main.c b/pc-bios/s390-ccw/main.c >> index 8bc6e8eaa3..0eeb53a2bf 100644 >> --- a/pc-bios/s390-ccw/main.c >> +++ b/pc-bios/s390-ccw/main.c >> @@ -328,6 +328,7 @@ static void ipl_pci_device(void) >> } >> >> switch (vdev->dev_type) { >> + case VIRTIO_ID_NET: >> case VIRTIO_ID_BLOCK: >> if (virtio_setup() == 0) { >> zipl_load(); /* only return on error */ >> diff --git a/pc-bios/s390-ccw/virtio-net.c b/pc-bios/s390-ccw/virtio-net.c >> index 3a9ae789cf..209f98a0e3 100644 >> --- a/pc-bios/s390-ccw/virtio-net.c >> +++ b/pc-bios/s390-ccw/virtio-net.c >> @@ -62,7 +62,16 @@ int virtio_net_init(void *mac_addr) >> rx_last_idx = 0; >> >> vdev->guest_features[0] = VIRTIO_NET_F_MAC_BIT; >> - virtio_ccw_setup(vdev); >> + switch (virtio_get_device()->ipl_type) { >> + case S390_IPL_TYPE_CCW: >> + virtio_ccw_setup(vdev); >> + break; >> + case S390_IPL_TYPE_PCI: >> + virtio_pci_setup(vdev); >> + break; >> + default: >> + return -1; >> + } >> >> if (!(vdev->guest_features[0] & VIRTIO_NET_F_MAC_BIT)) { >> puts("virtio-net device does not support the MAC address feature"); >> @@ -115,22 +124,30 @@ int recv(int fd, void *buf, int maxlen, int flags) >> VRing *rxvq = &vdev->vrings[VQ_RX]; >> int len, id; >> uint8_t *pkt; >> + uint16_t rx_used_idx, rx_avail_idx; >> + uint32_t rx_used_len, rx_used_id; >> + uint64_t rx_desc_addr; >> >> - if (rx_last_idx == rxvq->used->idx) { >> + rx_used_idx = virtio_tswap16(rxvq->used->idx); >> + if (rx_last_idx == rx_used_idx) { >> return 0; >> } >> >> - len = rxvq->used->ring[rx_last_idx % rxvq->num].len - virtio_net_hdr_size; >> + rx_used_len = virtio_tswap32(rxvq->used->ring[rx_last_idx % rxvq->num].len); >> + rx_used_id = virtio_tswap32(rxvq->used->ring[rx_last_idx % rxvq->num].id); >> + >> + len = rx_used_len - virtio_net_hdr_size; >> if (len > maxlen) { >> puts("virtio-net: Receive buffer too small"); >> len = maxlen; >> } >> - id = rxvq->used->ring[rx_last_idx % rxvq->num].id % rxvq->num; >> - pkt = (uint8_t *)(rxvq->desc[id].addr + virtio_net_hdr_size); >> + id = rx_used_id % rxvq->num; >> + rx_desc_addr = virtio_tswap64(rxvq->desc[id].addr); >> + pkt = (uint8_t *)(rx_desc_addr + virtio_net_hdr_size); >> >> #if DEBUG_VIRTIO_NET /* Dump packet */ >> int i; >> - printf("\nbuf %p: len=%i\n", (void *)rxvq->desc[id].addr, len); >> + printf("\nbuf %p: len=%i\n", (void *)rx_desc_addr, len); >> for (i = 0; i < 64; i++) { >> printf(" %02x", pkt[i]); >> if ((i % 16) == 15) { >> @@ -144,8 +161,10 @@ int recv(int fd, void *buf, int maxlen, int flags) >> memcpy(buf, pkt, len); >> >> /* Mark buffer as available to the host again */ >> - rxvq->avail->ring[rxvq->avail->idx % rxvq->num] = id; >> - rxvq->avail->idx = rxvq->avail->idx + 1; >> + rx_avail_idx = virtio_tswap16(rxvq->avail->idx); >> + rxvq->avail->ring[rx_avail_idx % rxvq->num] = virtio_tswap16(id); >> + rx_avail_idx++; >> + rxvq->avail->idx = virtio_tswap16(rx_avail_idx); >> vring_notify(rxvq); >> >> /* Move index to next entry */ >> @@ -164,6 +183,8 @@ bool virtio_net_setup(void) >> switch (virtio_get_device()->ipl_type) { >> case S390_IPL_TYPE_CCW: >> return virtio_ccw_net_setup(); >> + case S390_IPL_TYPE_PCI: >> + return virtio_pci_net_setup(); >> default: >> return false; >> } >> diff --git a/pc-bios/s390-ccw/virtio-pci.c b/pc-bios/s390-ccw/virtio-pci.c >> index f501252c81..c7d62766b8 100644 >> --- a/pc-bios/s390-ccw/virtio-pci.c >> +++ b/pc-bios/s390-ccw/virtio-pci.c >> @@ -52,6 +52,10 @@ void virtio_pci_id2type(VDev *vdev, uint16_t device_id) >> case 0x1001: >> vdev->dev_type = VIRTIO_ID_BLOCK; >> break; >> + case 0x1041: >> + case 0x1000: >> + vdev->dev_type = VIRTIO_ID_NET; >> + break; >> default: >> vdev->dev_type = 0; >> } >> @@ -199,6 +203,14 @@ static int virtio_pci_get_blk_config(void) >> return rc; >> } >> >> +static int virtio_pci_get_net_config(void) >> +{ >> + VirtioNetConfig *cfg = &virtio_get_device()->config.net; >> + int rc = vpci_read_flex(d_cap.off, d_cap.bar, cfg, sizeof(VirtioNetConfig)); >> + >> + return rc; >> +} >> + >> static int virtio_pci_negotiate(void) >> { >> int i, rc; >> @@ -330,6 +342,7 @@ bool virtio_pci_is_supported(VDev *vdev) >> if (vdev->vendor_id == PCI_VENDOR_VIRTIO) { >> switch (vdev->dev_type) { >> case VIRTIO_ID_BLOCK: >> + case VIRTIO_ID_NET: >> return true; >> default: >> return false; >> @@ -384,6 +397,11 @@ int virtio_pci_setup(VDev *vdev) >> vdev->cmd_vr_idx = 0; >> virtio_pci_get_blk_config(); >> break; >> + case VIRTIO_ID_NET: >> + vdev->nr_vqs = 2; >> + vdev->cmd_vr_idx = 0; >> + virtio_pci_get_net_config(); >> + break; >> default: >> puts("Unsupported virtio device"); >> return -ENODEV; >> @@ -458,3 +476,28 @@ int virtio_pci_setup_device(void) >> >> return 0; >> } >> + >> +static bool find_pci_net_dev(void) >> +{ >> + if (!virtio_is_supported(virtio_get_device())) { >> + return false; >> + } >> + >> + if (virtio_get_device_type() != VIRTIO_ID_NET) { >> + return false; >> + } >> + >> + return true; >> +} >> + >> +bool virtio_pci_net_setup(void) >> +{ >> + bool found = false; >> + >> + if (have_iplb || store_iplb(&iplb)) { >> + IPL_assert(iplb.pbt == S390_IPL_TYPE_PCI, "IPL_TYPE_PCI expected"); >> + found = find_pci_net_dev(); >> + } >> + >> + return found; >> +} > > As far as I can tell the net specific setup/find are not needed for PCI > devices.  By the time we reach the the call to virtio_pci_net_setup() we > know vdev->ipl_type == S390_IPL_TYPE_PCI because that is the prerequisite > to call the function in the first place, and vdev->ipl_type was previously > set based on the PBT.  We also already know that our device type is > VIRTIO_ID_NET because otherwise the call to virtio_setup() in main.c > would have fallen through to the default case and failed and/or netmain() > would not be called from bootmap.c. > > Although we haven't checked virtio_is_supported() yet, we will do that > verification in virtio_net_init(), which is the very next step. > > For PCI devices, everything checked from virtio_net_setup() appears to be > guaranteed by the call chain or validated by subsequent net code. Maybe > virtio_net_setup() only needs to be called for CCW devices? > I was trying to align the CCW setup flow with the PCI devices, but I agree that it isn't really necessary. If virtio_net_setup() only needs to be called for CCW devices, is the generic virtio_net_setup() wrapper still needed? Something like the following seems sufficient: if (virtio_get_device()->ipl_type == S390_IPL_TYPE_CCW && !virtio_ccw_net_setup()) { puts("No valid virtio ccw net device found."); } Alternatively, we could simply add return true for the S390_IPL_TYPE_PCI case in virtio_net_setup(). One thing worth noting is that virtio_net_init() calls virtio_*_setup(), which performs virtio_is_supported() as its first step. However, the return value of virtio_*_setup() is currently ignored in virtio_net_init(). This is pre-existing behavior and applies to both the CCW and PCI paths. If setup fails, execution still proceeds to MAC negotiation and buffer allocation. Since we are removing the network-specific setup from the PCI path, would it make sense to also fix virtio_net_init() to check and propagate the return value of virtio_*_setup()? The error is already reported by virtio_*_setup(), but execution currently continues into MAC negotiation and buffer allocation even when device initialization has failed. > Regards, > Jared Rossi >> diff --git a/pc-bios/s390-ccw/virtio-pci.h b/pc-bios/s390-ccw/virtio-pci.h >> index 33b683bd92..2c3136fd05 100644 >> --- a/pc-bios/s390-ccw/virtio-pci.h >> +++ b/pc-bios/s390-ccw/virtio-pci.h >> @@ -70,6 +70,8 @@ long virtio_pci_notify(VRing *vr); >> bool virtio_pci_is_supported(VDev *vdev); >> int virtio_pci_setup(VDev *vdev); >> int virtio_pci_setup_device(void); >> +bool virtio_pci_is_supported(VDev *vdev); >> +bool virtio_pci_net_setup(void); >> >> int vpci_read_flex(uint64_t offset, uint8_t pcias, void *buf, int len); >> int vpci_read_bswap64(uint64_t offset, uint8_t pcias, uint64_t *buf); >