From nobody Sun Jun 15 15:58:17 2025 Delivered-To: importer2@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer2=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=gmail.com ARC-Seal: i=1; a=rsa-sha256; t=1716991763; cv=none; d=zohomail.com; s=zohoarc; b=E6LdbOdcVEDt5X5NE4Ej303wqfgNEjYZnG87gL2QOoL8psPfGY0Ls5isUlAWBNHEPD+H2JLbx80qfxPbuQ4LRKTpxWw49A9VtXd+/jMxeLZRXzu+a/K3LfulxYp5z1tNbmm3a7bWL20QbfQtPLFaMz2/SMwhw7600UWVkw+3wl4= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1716991763; h=Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=6KnYV1+9ySu8xm9Hqqrupes7IMG1mFuyj5Dx0g125FI=; b=balNpVYZ78EmZGv84giH+5MMw2sZbGBnFuuyo/CnUUidJJYWEJng91Xp2KfW8i7NR0JEQjL3bZaQs9jL0ZRI4y0hyjWJF8DJ6rqDRg7vDBUnm3k2tKB22blZ743nsJ5+sYrs6LDsk6LC+d5e2UmXKW6xay/zUjhGxKCVtuEt+0A= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer2=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1716991763082675.86110350143; Wed, 29 May 2024 07:09:23 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1sCJyC-0006Bc-0N; Wed, 29 May 2024 10:08:00 -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 1sCJyA-0006B6-By for qemu-devel@nongnu.org; Wed, 29 May 2024 10:07:58 -0400 Received: from mail-lj1-x235.google.com ([2a00:1450:4864:20::235]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1sCJy7-0005jJ-T2 for qemu-devel@nongnu.org; Wed, 29 May 2024 10:07:58 -0400 Received: by mail-lj1-x235.google.com with SMTP id 38308e7fff4ca-2e724bc466fso22554611fa.3 for ; Wed, 29 May 2024 07:07:55 -0700 (PDT) Received: from gmail.com (213-67-3-247-no600.tbcn.telia.com. [213.67.3.247]) by smtp.gmail.com with ESMTPSA id 38308e7fff4ca-2e95bcf47a4sm25715181fa.57.2024.05.29.07.07.50 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 29 May 2024 07:07:50 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1716991673; x=1717596473; darn=nongnu.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=6KnYV1+9ySu8xm9Hqqrupes7IMG1mFuyj5Dx0g125FI=; b=eFmYNES3/MFZ6rSgSIAvqbZ92EF5aey3bMXzCqUkGtscIciiobJVqM1CcfIi9XQFpT EEZTNQZUEAQd4inDNJS17hvlso7HN0yz5wNc0zlH6JaPBASE7EFwTHDd+WIcFmGmWqpU aR6XEUqp54ycjlVTqiWBjEisNjnzxE95PNYBD5fplgOC7+RX2gyQ5rLcTH+zftyi0mAG URN3TKvxxoyFeNjpKYmP597MIGZVsjVI1akx9n5ghrVChzpb2FWwB0t+qyXbZTqABRn7 U36hPw/Cz7TPsFrziLh7J4ZkoK2fk19BMS6deNNRTnM7PsQbG8HoH8BMiFOfl5Z9nbcv qNSQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1716991673; x=1717596473; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=6KnYV1+9ySu8xm9Hqqrupes7IMG1mFuyj5Dx0g125FI=; b=WSjf+WK0HCuGm4JjMBI+vOH6GJDsC8dYy3bSQ+T6jYOtJPX1DdAzJR+ycjcBmNLz95 gjdEVCaHDDTI7qVjiDCDgMK4qn+/WKI26x1WfLPcsKPa7jsxR+YpQP77NMw2D00PS1NL wxekC+gM3b9K7ZodpXpR86hXo19r/O94afkVaTvys9qzLMl4cOoUy0359PBNOHRXrqTu uhJXYXJ8dc+3tGKS+pZ0oLBUZQ2Fhf5apWPHINGpRMCSAPjYVe+WdBvDCEmWItms+6Fi aq5j8aBFCyfablA0heXeAIq8VVrjkxTOSWgWiI4+CFw7EM7bgHaCYmUq8dJpVdNSH3FQ c2Fw== X-Gm-Message-State: AOJu0Ywqjc7tRlCyoRFLt3VXW5u3UshT3gYfpZYN313JUuQObXwPILIG I9CIrEEdRtBCnCgmVBk+fLwszSgmgGxUaWBPrSzrNzt9rTh+H843jcoaAlNb X-Google-Smtp-Source: AGHT+IGD9ZwNPNUv35c+UJ5GB2OUukDkocTjVhprwhJ89cISo2uRDy4+PmwLngnnEeG1Una9nQbZqw== X-Received: by 2002:a2e:a685:0:b0:2ea:80b7:c464 with SMTP id 38308e7fff4ca-2ea80b7c7acmr4569281fa.22.1716991672097; Wed, 29 May 2024 07:07:52 -0700 (PDT) From: "Edgar E. Iglesias" To: qemu-devel@nongnu.org Cc: edgar.iglesias@gmail.com, sstabellini@kernel.org, jgross@suse.com, "Edgar E. Iglesias" , Manos Pitsidianakis , Anthony PERARD , Paul Durrant , xen-devel@lists.xenproject.org Subject: [PATCH v8 7/8] xen: mapcache: Add support for grant mappings Date: Wed, 29 May 2024 16:07:38 +0200 Message-Id: <20240529140739.1387692-8-edgar.iglesias@gmail.com> X-Mailer: git-send-email 2.40.1 In-Reply-To: <20240529140739.1387692-1-edgar.iglesias@gmail.com> References: <20240529140739.1387692-1-edgar.iglesias@gmail.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer2=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=2a00:1450:4864:20::235; envelope-from=edgar.iglesias@gmail.com; helo=mail-lj1-x235.google.com X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 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: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer2=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer2=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @gmail.com) X-ZM-MESSAGEID: 1716991764954100009 Content-Type: text/plain; charset="utf-8" From: "Edgar E. Iglesias" Add a second mapcache for grant mappings. The mapcache for grants needs to work with XC_PAGE_SIZE granularity since we can't map larger ranges than what has been granted to us. Like with foreign mappings (xen_memory), machines using grants are expected to initialize the xen_grants MR and map it into their address-map accordingly. CC: Manos Pitsidianakis Signed-off-by: Edgar E. Iglesias Reviewed-by: Stefano Stabellini --- hw/xen/xen-hvm-common.c | 12 ++- hw/xen/xen-mapcache.c | 165 +++++++++++++++++++++++++------- include/hw/xen/xen-hvm-common.h | 3 + include/sysemu/xen.h | 1 + 4 files changed, 144 insertions(+), 37 deletions(-) diff --git a/hw/xen/xen-hvm-common.c b/hw/xen/xen-hvm-common.c index a0a0252da0..b8ace1c368 100644 --- a/hw/xen/xen-hvm-common.c +++ b/hw/xen/xen-hvm-common.c @@ -10,12 +10,18 @@ #include "hw/boards.h" #include "hw/xen/arch_hvm.h" =20 -MemoryRegion xen_memory; +MemoryRegion xen_memory, xen_grants; =20 -/* Check for xen memory. */ +/* Check for any kind of xen memory, foreign mappings or grants. */ bool xen_mr_is_memory(MemoryRegion *mr) { - return mr =3D=3D &xen_memory; + return mr =3D=3D &xen_memory || mr =3D=3D &xen_grants; +} + +/* Check specifically for grants. */ +bool xen_mr_is_grants(MemoryRegion *mr) +{ + return mr =3D=3D &xen_grants; } =20 void xen_ram_alloc(ram_addr_t ram_addr, ram_addr_t size, MemoryRegion *mr, diff --git a/hw/xen/xen-mapcache.c b/hw/xen/xen-mapcache.c index a07c47b0b1..5f23b0adbe 100644 --- a/hw/xen/xen-mapcache.c +++ b/hw/xen/xen-mapcache.c @@ -14,6 +14,7 @@ =20 #include =20 +#include "hw/xen/xen-hvm-common.h" #include "hw/xen/xen_native.h" #include "qemu/bitmap.h" =20 @@ -21,6 +22,8 @@ #include "sysemu/xen-mapcache.h" #include "trace.h" =20 +#include +#include =20 #if HOST_LONG_BITS =3D=3D 32 # define MCACHE_MAX_SIZE (1UL<<31) /* 2GB Cap */ @@ -41,6 +44,7 @@ typedef struct MapCacheEntry { unsigned long *valid_mapping; uint32_t lock; #define XEN_MAPCACHE_ENTRY_DUMMY (1 << 0) +#define XEN_MAPCACHE_ENTRY_GRANT (1 << 1) uint8_t flags; hwaddr size; struct MapCacheEntry *next; @@ -71,6 +75,8 @@ typedef struct MapCache { } MapCache; =20 static MapCache *mapcache; +static MapCache *mapcache_grants; +static xengnttab_handle *xen_region_gnttabdev; =20 static inline void mapcache_lock(MapCache *mc) { @@ -131,6 +137,12 @@ void xen_map_cache_init(phys_offset_to_gaddr_t f, void= *opaque) unsigned long max_mcache_size; unsigned int bucket_shift; =20 + xen_region_gnttabdev =3D xengnttab_open(NULL, 0); + if (xen_region_gnttabdev =3D=3D NULL) { + error_report("mapcache: Failed to open gnttab device"); + exit(EXIT_FAILURE); + } + if (HOST_LONG_BITS =3D=3D 32) { bucket_shift =3D 16; } else { @@ -159,6 +171,15 @@ void xen_map_cache_init(phys_offset_to_gaddr_t f, void= *opaque) mapcache =3D xen_map_cache_init_single(f, opaque, bucket_shift, max_mcache_size); + + /* + * Grant mappings must use XC_PAGE_SIZE granularity since we can't + * map anything beyond the number of pages granted to us. + */ + mapcache_grants =3D xen_map_cache_init_single(f, opaque, + XC_PAGE_SHIFT, + max_mcache_size); + setrlimit(RLIMIT_AS, &rlimit_as); } =20 @@ -168,17 +189,24 @@ static void xen_remap_bucket(MapCache *mc, hwaddr size, hwaddr address_index, bool dummy, + bool grant, + bool is_write, ram_addr_t ram_offset) { uint8_t *vaddr_base; - xen_pfn_t *pfns; - int *err; + g_autofree uint32_t *refs =3D NULL; + g_autofree xen_pfn_t *pfns =3D NULL; + g_autofree int *err; unsigned int i; hwaddr nb_pfn =3D size >> XC_PAGE_SHIFT; =20 trace_xen_remap_bucket(address_index); =20 - pfns =3D g_new0(xen_pfn_t, nb_pfn); + if (grant) { + refs =3D g_new0(uint32_t, nb_pfn); + } else { + pfns =3D g_new0(xen_pfn_t, nb_pfn); + } err =3D g_new0(int, nb_pfn); =20 if (entry->vaddr_base !=3D NULL) { @@ -207,21 +235,51 @@ static void xen_remap_bucket(MapCache *mc, g_free(entry->valid_mapping); entry->valid_mapping =3D NULL; =20 - for (i =3D 0; i < nb_pfn; i++) { - pfns[i] =3D (address_index << (mc->bucket_shift - XC_PAGE_SHIFT)) = + i; + if (grant) { + hwaddr grant_base =3D address_index - (ram_offset >> XC_PAGE_SHIFT= ); + + for (i =3D 0; i < nb_pfn; i++) { + refs[i] =3D grant_base + i; + } + } else { + for (i =3D 0; i < nb_pfn; i++) { + pfns[i] =3D (address_index << (mc->bucket_shift - XC_PAGE_SHIF= T)) + i; + } } =20 - /* - * If the caller has requested the mapping at a specific address use - * MAP_FIXED to make sure it's honored. - */ + entry->flags &=3D ~XEN_MAPCACHE_ENTRY_GRANT; + if (!dummy) { - vaddr_base =3D xenforeignmemory_map2(xen_fmem, xen_domid, vaddr, - PROT_READ | PROT_WRITE, - vaddr ? MAP_FIXED : 0, - nb_pfn, pfns, err); + if (grant) { + int prot =3D PROT_READ; + + if (is_write) { + prot |=3D PROT_WRITE; + } + + entry->flags |=3D XEN_MAPCACHE_ENTRY_GRANT; + assert(vaddr =3D=3D NULL); + vaddr_base =3D xengnttab_map_domain_grant_refs(xen_region_gntt= abdev, + nb_pfn, + xen_domid, refs, + prot); + } else { + /* + * If the caller has requested the mapping at a specific addre= ss use + * MAP_FIXED to make sure it's honored. + * + * We don't yet support upgrading mappings from RO to RW, to h= andle + * models using ordinary address_space_rw(), foreign mappings = ignore + * is_write and are always mapped RW. + */ + vaddr_base =3D xenforeignmemory_map2(xen_fmem, xen_domid, vadd= r, + PROT_READ | PROT_WRITE, + vaddr ? MAP_FIXED : 0, + nb_pfn, pfns, err); + } if (vaddr_base =3D=3D NULL) { - perror("xenforeignmemory_map2"); + perror(grant ? "xengnttab_map_domain_grant_refs" + : "xenforeignmemory_map2"); exit(-1); } } else { @@ -260,15 +318,13 @@ static void xen_remap_bucket(MapCache *mc, bitmap_set(entry->valid_mapping, i, 1); } } - - g_free(pfns); - g_free(err); } =20 static uint8_t *xen_map_cache_unlocked(MapCache *mc, hwaddr phys_addr, hwaddr size, ram_addr_t ram_offset, - uint8_t lock, bool dma, bool is_wri= te) + uint8_t lock, bool dma, + bool grant, bool is_write) { MapCacheEntry *entry, *pentry =3D NULL, *free_entry =3D NULL, *free_pentry =3D NULL; @@ -340,7 +396,7 @@ tryagain: entry =3D g_new0(MapCacheEntry, 1); pentry->next =3D entry; xen_remap_bucket(mc, entry, NULL, cache_size, address_index, dummy, - ram_offset); + grant, is_write, ram_offset); } else if (!entry->lock) { if (!entry->vaddr_base || entry->paddr_index !=3D address_index || entry->size !=3D cache_size || @@ -348,7 +404,7 @@ tryagain: test_bit_size >> XC_PAGE_SHIFT, entry->valid_mapping)) { xen_remap_bucket(mc, entry, NULL, cache_size, address_index, d= ummy, - ram_offset); + grant, is_write, ram_offset); } } =20 @@ -399,12 +455,26 @@ uint8_t *xen_map_cache(MemoryRegion *mr, uint8_t lock, bool dma, bool is_write) { + bool grant =3D xen_mr_is_grants(mr); + MapCache *mc =3D grant ? mapcache_grants : mapcache; uint8_t *p; =20 - mapcache_lock(mapcache); - p =3D xen_map_cache_unlocked(mapcache, phys_addr, size, ram_addr_offse= t, - lock, dma, is_write); - mapcache_unlock(mapcache); + if (grant && !lock) { + /* + * Grants are only supported via address_space_map(). Anything + * else is considered a user/guest error. + * + * QEMU generally doesn't expect these mappings to ever fail, so + * if this happens we report an error message and abort(). + */ + error_report("Tried to access a grant reference without mapping it= ."); + abort(); + } + + mapcache_lock(mc); + p =3D xen_map_cache_unlocked(mc, phys_addr, size, ram_addr_offset, + lock, dma, grant, is_write); + mapcache_unlock(mc); return p; } =20 @@ -449,7 +519,14 @@ static ram_addr_t xen_ram_addr_from_mapcache_single(Ma= pCache *mc, void *ptr) =20 ram_addr_t xen_ram_addr_from_mapcache(void *ptr) { - return xen_ram_addr_from_mapcache_single(mapcache, ptr); + ram_addr_t addr; + + addr =3D xen_ram_addr_from_mapcache_single(mapcache, ptr); + if (addr =3D=3D RAM_ADDR_INVALID) { + addr =3D xen_ram_addr_from_mapcache_single(mapcache_grants, ptr); + } + + return addr; } =20 static void xen_invalidate_map_cache_entry_unlocked(MapCache *mc, @@ -460,6 +537,7 @@ static void xen_invalidate_map_cache_entry_unlocked(Map= Cache *mc, hwaddr paddr_index; hwaddr size; int found =3D 0; + int rc; =20 QTAILQ_FOREACH(reventry, &mc->locked_entries, next) { if (reventry->vaddr_req =3D=3D buffer) { @@ -502,7 +580,14 @@ static void xen_invalidate_map_cache_entry_unlocked(Ma= pCache *mc, } =20 ram_block_notify_remove(entry->vaddr_base, entry->size, entry->size); - if (munmap(entry->vaddr_base, entry->size) !=3D 0) { + if (entry->flags & XEN_MAPCACHE_ENTRY_GRANT) { + rc =3D xengnttab_unmap(xen_region_gnttabdev, entry->vaddr_base, + entry->size >> mc->bucket_shift); + } else { + rc =3D munmap(entry->vaddr_base, entry->size); + } + + if (rc) { perror("unmap fails"); exit(-1); } @@ -521,14 +606,24 @@ typedef struct XenMapCacheData { uint8_t *buffer; } XenMapCacheData; =20 +static void xen_invalidate_map_cache_entry_single(MapCache *mc, uint8_t *b= uffer) +{ + mapcache_lock(mc); + xen_invalidate_map_cache_entry_unlocked(mc, buffer); + mapcache_unlock(mc); +} + +static void xen_invalidate_map_cache_entry_all(uint8_t *buffer) +{ + xen_invalidate_map_cache_entry_single(mapcache, buffer); + xen_invalidate_map_cache_entry_single(mapcache_grants, buffer); +} + static void xen_invalidate_map_cache_entry_bh(void *opaque) { XenMapCacheData *data =3D opaque; =20 - mapcache_lock(mapcache); - xen_invalidate_map_cache_entry_unlocked(mapcache, data->buffer); - mapcache_unlock(mapcache); - + xen_invalidate_map_cache_entry_all(data->buffer); aio_co_wake(data->co); } =20 @@ -543,9 +638,7 @@ void coroutine_mixed_fn xen_invalidate_map_cache_entry(= uint8_t *buffer) xen_invalidate_map_cache_entry_bh, &data); qemu_coroutine_yield(); } else { - mapcache_lock(mapcache); - xen_invalidate_map_cache_entry_unlocked(mapcache, buffer); - mapcache_unlock(mapcache); + xen_invalidate_map_cache_entry_all(buffer); } } =20 @@ -597,6 +690,7 @@ void xen_invalidate_map_cache(void) bdrv_drain_all(); =20 xen_invalidate_map_cache_single(mapcache); + xen_invalidate_map_cache_single(mapcache_grants); } =20 static uint8_t *xen_replace_cache_entry_unlocked(MapCache *mc, @@ -632,13 +726,16 @@ static uint8_t *xen_replace_cache_entry_unlocked(MapC= ache *mc, return NULL; } =20 + assert((entry->flags & XEN_MAPCACHE_ENTRY_GRANT) =3D=3D 0); + address_index =3D new_phys_addr >> mc->bucket_shift; address_offset =3D new_phys_addr & (mc->bucket_size - 1); =20 trace_xen_replace_cache_entry_dummy(old_phys_addr, new_phys_addr); =20 xen_remap_bucket(mc, entry, entry->vaddr_base, - cache_size, address_index, false, old_phys_addr); + cache_size, address_index, false, + false, false, old_phys_addr); if (!test_bits(address_offset >> XC_PAGE_SHIFT, test_bit_size >> XC_PAGE_SHIFT, entry->valid_mapping)) { diff --git a/include/hw/xen/xen-hvm-common.h b/include/hw/xen/xen-hvm-commo= n.h index 65a51aac2e..3d796235dc 100644 --- a/include/hw/xen/xen-hvm-common.h +++ b/include/hw/xen/xen-hvm-common.h @@ -16,6 +16,7 @@ #include =20 extern MemoryRegion xen_memory; +extern MemoryRegion xen_grants; extern MemoryListener xen_io_listener; extern DeviceListener xen_device_listener; =20 @@ -29,6 +30,8 @@ extern DeviceListener xen_device_listener; do { } while (0) #endif =20 +#define XEN_GRANT_ADDR_OFF (1ULL << 63) + static inline uint32_t xen_vcpu_eport(shared_iopage_t *shared_page, int i) { return shared_page->vcpu_ioreq[i].vp_eport; diff --git a/include/sysemu/xen.h b/include/sysemu/xen.h index 3445888e39..d70eacfbe2 100644 --- a/include/sysemu/xen.h +++ b/include/sysemu/xen.h @@ -50,4 +50,5 @@ static inline void xen_ram_alloc(ram_addr_t ram_addr, ram= _addr_t size, #endif /* CONFIG_XEN_IS_POSSIBLE */ =20 bool xen_mr_is_memory(MemoryRegion *mr); +bool xen_mr_is_grants(MemoryRegion *mr); #endif --=20 2.40.1