diff --git a/MAINTAINERS b/MAINTAINERS index 87f90721b9..2a1520dee7 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -407,6 +407,7 @@ Guest CPU Cores (Xen): X86 M: Stefano Stabellini M: Anthony Perard +M: Paul Durrant L: xen-devel@lists.xenproject.org S: Supported F: */xen* @@ -414,10 +415,12 @@ F: hw/9pfs/xen-9p-backend.c F: hw/char/xen_console.c F: hw/display/xenfb.c F: hw/net/xen_nic.c -F: hw/block/xen_* +F: hw/block/xen* +F: hw/block/dataplane/xen* F: hw/xen/ F: hw/xenpv/ F: hw/i386/xen/ +F: include/hw/block/dataplane/xen* F: include/hw/xen/ F: include/sysemu/xen-mapcache.h diff --git a/configure b/configure index 4ea3f14883..f992709b89 100755 --- a/configure +++ b/configure @@ -357,7 +357,6 @@ vnc_png="" xkbcommon="" xen="" xen_ctrl_version="" -xen_pv_domain_build="no" xen_pci_passthrough="" linux_aio="" cap_ng="" @@ -1119,10 +1118,6 @@ for opt do ;; --enable-xen-pci-passthrough) xen_pci_passthrough="yes" ;; - --disable-xen-pv-domain-build) xen_pv_domain_build="no" - ;; - --enable-xen-pv-domain-build) xen_pv_domain_build="yes" - ;; --disable-brlapi) brlapi="no" ;; --enable-brlapi) brlapi="yes" @@ -1685,8 +1680,6 @@ Advanced options (experts only): --tls-priority default TLS protocol/cipher priority string --enable-gprof QEMU profiling with gprof --enable-profiler profiler support - --enable-xen-pv-domain-build - xen pv domain builder --enable-debug-stack-usage track the maximum stack usage of stacks created by qemu_alloc_stack @@ -2678,12 +2671,6 @@ if test "$xen_pci_passthrough" != "no"; then fi fi -if test "$xen_pv_domain_build" = "yes" && - test "$xen" != "yes"; then - error_exit "User requested Xen PV domain builder support" \ - "which requires Xen support." -fi - ########################################## # Windows Hypervisor Platform accelerator (WHPX) check if test "$whpx" != "no" ; then @@ -6069,7 +6056,6 @@ fi echo "xen support $xen" if test "$xen" = "yes" ; then echo "xen ctrl version $xen_ctrl_version" - echo "pv dom build $xen_pv_domain_build" fi echo "brlapi support $brlapi" echo "bluez support $bluez" @@ -6539,9 +6525,6 @@ fi if test "$xen" = "yes" ; then echo "CONFIG_XEN_BACKEND=y" >> $config_host_mak echo "CONFIG_XEN_CTRL_INTERFACE_VERSION=$xen_ctrl_version" >> $config_host_mak - if test "$xen_pv_domain_build" = "yes" ; then - echo "CONFIG_XEN_PV_DOMAIN_BUILD=y" >> $config_host_mak - fi fi if test "$linux_aio" = "yes" ; then echo "CONFIG_LINUX_AIO=y" >> $config_host_mak diff --git a/hw/9pfs/xen-9p-backend.c b/hw/9pfs/xen-9p-backend.c index 9015fe7773..25ab04d95a 100644 --- a/hw/9pfs/xen-9p-backend.c +++ b/hw/9pfs/xen-9p-backend.c @@ -12,7 +12,7 @@ #include "hw/hw.h" #include "hw/9pfs/9p.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "hw/9pfs/xen-9pfs.h" #include "qapi/error.h" #include "qemu/config-file.h" @@ -45,7 +45,7 @@ typedef struct Xen9pfsRing { } Xen9pfsRing; typedef struct Xen9pfsDev { - struct XenDevice xendev; /* must be first */ + struct XenLegacyDevice xendev; /* must be first */ V9fsState state; char *path; char *security_model; @@ -56,7 +56,7 @@ typedef struct Xen9pfsDev { Xen9pfsRing *rings; } Xen9pfsDev; -static void xen_9pfs_disconnect(struct XenDevice *xendev); +static void xen_9pfs_disconnect(struct XenLegacyDevice *xendev); static void xen_9pfs_in_sg(Xen9pfsRing *ring, struct iovec *in_sg, @@ -243,7 +243,7 @@ static const V9fsTransport xen_9p_transport = { .push_and_notify = xen_9pfs_push_and_notify, }; -static int xen_9pfs_init(struct XenDevice *xendev) +static int xen_9pfs_init(struct XenLegacyDevice *xendev) { return 0; } @@ -305,7 +305,7 @@ static void xen_9pfs_evtchn_event(void *opaque) qemu_bh_schedule(ring->bh); } -static void xen_9pfs_disconnect(struct XenDevice *xendev) +static void xen_9pfs_disconnect(struct XenLegacyDevice *xendev) { Xen9pfsDev *xen_9pdev = container_of(xendev, Xen9pfsDev, xendev); int i; @@ -321,7 +321,7 @@ static void xen_9pfs_disconnect(struct XenDevice *xendev) } } -static int xen_9pfs_free(struct XenDevice *xendev) +static int xen_9pfs_free(struct XenLegacyDevice *xendev) { Xen9pfsDev *xen_9pdev = container_of(xendev, Xen9pfsDev, xendev); int i; @@ -354,7 +354,7 @@ static int xen_9pfs_free(struct XenDevice *xendev) return 0; } -static int xen_9pfs_connect(struct XenDevice *xendev) +static int xen_9pfs_connect(struct XenLegacyDevice *xendev) { Error *err = NULL; int i; @@ -467,7 +467,7 @@ out: return -1; } -static void xen_9pfs_alloc(struct XenDevice *xendev) +static void xen_9pfs_alloc(struct XenLegacyDevice *xendev) { xenstore_write_be_str(xendev, "versions", VERSIONS); xenstore_write_be_int(xendev, "max-rings", MAX_RINGS); diff --git a/hw/block/Makefile.objs b/hw/block/Makefile.objs index 53ce5751ae..e206b8e712 100644 --- a/hw/block/Makefile.objs +++ b/hw/block/Makefile.objs @@ -4,7 +4,7 @@ common-obj-$(CONFIG_SSI_M25P80) += m25p80.o common-obj-$(CONFIG_NAND) += nand.o common-obj-$(CONFIG_PFLASH_CFI01) += pflash_cfi01.o common-obj-$(CONFIG_PFLASH_CFI02) += pflash_cfi02.o -common-obj-$(CONFIG_XEN) += xen_disk.o +common-obj-$(CONFIG_XEN) += xen-block.o common-obj-$(CONFIG_ECC) += ecc.o common-obj-$(CONFIG_ONENAND) += onenand.o common-obj-$(CONFIG_NVME_PCI) += nvme.o diff --git a/hw/block/dataplane/Makefile.objs b/hw/block/dataplane/Makefile.objs index e786f66421..c6c68dbc00 100644 --- a/hw/block/dataplane/Makefile.objs +++ b/hw/block/dataplane/Makefile.objs @@ -1 +1,2 @@ obj-y += virtio-blk.o +obj-$(CONFIG_XEN) += xen-block.o diff --git a/hw/block/dataplane/xen-block.c b/hw/block/dataplane/xen-block.c new file mode 100644 index 0000000000..d0d8905a33 --- /dev/null +++ b/hw/block/dataplane/xen-block.c @@ -0,0 +1,827 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * (c) Gerd Hoffmann + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; under version 2 of the License. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, see . + * + * Contributions after 2012-01-13 are licensed under the terms of the + * GNU GPL, version 2 or (at your option) any later version. + */ + +#include "qemu/osdep.h" +#include "qemu/error-report.h" +#include "qapi/error.h" +#include "hw/hw.h" +#include "hw/xen/xen_common.h" +#include "hw/block/xen_blkif.h" +#include "sysemu/block-backend.h" +#include "sysemu/iothread.h" +#include "xen-block.h" + +typedef struct XenBlockRequest { + blkif_request_t req; + int16_t status; + off_t start; + QEMUIOVector v; + void *buf; + size_t size; + int presync; + int aio_inflight; + int aio_errors; + XenBlockDataPlane *dataplane; + QLIST_ENTRY(XenBlockRequest) list; + BlockAcctCookie acct; +} XenBlockRequest; + +struct XenBlockDataPlane { + XenDevice *xendev; + XenEventChannel *event_channel; + unsigned int *ring_ref; + unsigned int nr_ring_ref; + void *sring; + int64_t file_blk; + int64_t file_size; + int protocol; + blkif_back_rings_t rings; + int more_work; + QLIST_HEAD(inflight_head, XenBlockRequest) inflight; + QLIST_HEAD(freelist_head, XenBlockRequest) freelist; + int requests_total; + int requests_inflight; + unsigned int max_requests; + BlockBackend *blk; + QEMUBH *bh; + IOThread *iothread; + AioContext *ctx; +}; + +static void reset_request(XenBlockRequest *request) +{ + memset(&request->req, 0, sizeof(request->req)); + request->status = 0; + request->start = 0; + request->size = 0; + request->presync = 0; + + request->aio_inflight = 0; + request->aio_errors = 0; + + request->dataplane = NULL; + memset(&request->list, 0, sizeof(request->list)); + memset(&request->acct, 0, sizeof(request->acct)); + + qemu_iovec_reset(&request->v); +} + +static XenBlockRequest *xen_block_start_request(XenBlockDataPlane *dataplane) +{ + XenBlockRequest *request = NULL; + + if (QLIST_EMPTY(&dataplane->freelist)) { + if (dataplane->requests_total >= dataplane->max_requests) { + goto out; + } + /* allocate new struct */ + request = g_malloc0(sizeof(*request)); + request->dataplane = dataplane; + /* + * We cannot need more pages per requests than this, and since we + * re-use requests, allocate the memory once here. It will be freed + * xen_block_dataplane_destroy() when the request list is freed. + */ + request->buf = qemu_memalign(XC_PAGE_SIZE, + BLKIF_MAX_SEGMENTS_PER_REQUEST * + XC_PAGE_SIZE); + dataplane->requests_total++; + qemu_iovec_init(&request->v, 1); + } else { + /* get one from freelist */ + request = QLIST_FIRST(&dataplane->freelist); + QLIST_REMOVE(request, list); + } + QLIST_INSERT_HEAD(&dataplane->inflight, request, list); + dataplane->requests_inflight++; + +out: + return request; +} + +static void xen_block_finish_request(XenBlockRequest *request) +{ + XenBlockDataPlane *dataplane = request->dataplane; + + QLIST_REMOVE(request, list); + dataplane->requests_inflight--; +} + +static void xen_block_release_request(XenBlockRequest *request) +{ + XenBlockDataPlane *dataplane = request->dataplane; + + QLIST_REMOVE(request, list); + reset_request(request); + request->dataplane = dataplane; + QLIST_INSERT_HEAD(&dataplane->freelist, request, list); + dataplane->requests_inflight--; +} + +/* + * translate request into iovec + start offset + * do sanity checks along the way + */ +static int xen_block_parse_request(XenBlockRequest *request) +{ + XenBlockDataPlane *dataplane = request->dataplane; + size_t len; + int i; + + switch (request->req.operation) { + case BLKIF_OP_READ: + break; + case BLKIF_OP_FLUSH_DISKCACHE: + request->presync = 1; + if (!request->req.nr_segments) { + return 0; + } + /* fall through */ + case BLKIF_OP_WRITE: + break; + case BLKIF_OP_DISCARD: + return 0; + default: + error_report("error: unknown operation (%d)", request->req.operation); + goto err; + }; + + if (request->req.operation != BLKIF_OP_READ && + blk_is_read_only(dataplane->blk)) { + error_report("error: write req for ro device"); + goto err; + } + + request->start = request->req.sector_number * dataplane->file_blk; + for (i = 0; i < request->req.nr_segments; i++) { + if (i == BLKIF_MAX_SEGMENTS_PER_REQUEST) { + error_report("error: nr_segments too big"); + goto err; + } + if (request->req.seg[i].first_sect > request->req.seg[i].last_sect) { + error_report("error: first > last sector"); + goto err; + } + if (request->req.seg[i].last_sect * dataplane->file_blk >= + XC_PAGE_SIZE) { + error_report("error: page crossing"); + goto err; + } + + len = (request->req.seg[i].last_sect - + request->req.seg[i].first_sect + 1) * dataplane->file_blk; + request->size += len; + } + if (request->start + request->size > dataplane->file_size) { + error_report("error: access beyond end of file"); + goto err; + } + return 0; + +err: + request->status = BLKIF_RSP_ERROR; + return -1; +} + +static int xen_block_copy_request(XenBlockRequest *request) +{ + XenBlockDataPlane *dataplane = request->dataplane; + XenDevice *xendev = dataplane->xendev; + XenDeviceGrantCopySegment segs[BLKIF_MAX_SEGMENTS_PER_REQUEST]; + int i, count; + int64_t file_blk = dataplane->file_blk; + bool to_domain = (request->req.operation == BLKIF_OP_READ); + void *virt = request->buf; + Error *local_err = NULL; + + if (request->req.nr_segments == 0) { + return 0; + } + + count = request->req.nr_segments; + + for (i = 0; i < count; i++) { + if (to_domain) { + segs[i].dest.foreign.ref = request->req.seg[i].gref; + segs[i].dest.foreign.offset = request->req.seg[i].first_sect * + file_blk; + segs[i].source.virt = virt; + } else { + segs[i].source.foreign.ref = request->req.seg[i].gref; + segs[i].source.foreign.offset = request->req.seg[i].first_sect * + file_blk; + segs[i].dest.virt = virt; + } + segs[i].len = (request->req.seg[i].last_sect - + request->req.seg[i].first_sect + 1) * file_blk; + virt += segs[i].len; + } + + xen_device_copy_grant_refs(xendev, to_domain, segs, count, &local_err); + + if (local_err) { + error_reportf_err(local_err, "failed to copy data: "); + + request->aio_errors++; + return -1; + } + + return 0; +} + +static int xen_block_do_aio(XenBlockRequest *request); +static int xen_block_send_response(XenBlockRequest *request); + +static void xen_block_complete_aio(void *opaque, int ret) +{ + XenBlockRequest *request = opaque; + XenBlockDataPlane *dataplane = request->dataplane; + + aio_context_acquire(dataplane->ctx); + + if (ret != 0) { + error_report("%s I/O error", + request->req.operation == BLKIF_OP_READ ? + "read" : "write"); + request->aio_errors++; + } + + request->aio_inflight--; + if (request->presync) { + request->presync = 0; + xen_block_do_aio(request); + goto done; + } + if (request->aio_inflight > 0) { + goto done; + } + + switch (request->req.operation) { + case BLKIF_OP_READ: + /* in case of failure request->aio_errors is increased */ + if (ret == 0) { + xen_block_copy_request(request); + } + break; + case BLKIF_OP_WRITE: + case BLKIF_OP_FLUSH_DISKCACHE: + if (!request->req.nr_segments) { + break; + } + break; + default: + break; + } + + request->status = request->aio_errors ? BLKIF_RSP_ERROR : BLKIF_RSP_OKAY; + xen_block_finish_request(request); + + switch (request->req.operation) { + case BLKIF_OP_WRITE: + case BLKIF_OP_FLUSH_DISKCACHE: + if (!request->req.nr_segments) { + break; + } + case BLKIF_OP_READ: + if (request->status == BLKIF_RSP_OKAY) { + block_acct_done(blk_get_stats(dataplane->blk), &request->acct); + } else { + block_acct_failed(blk_get_stats(dataplane->blk), &request->acct); + } + break; + case BLKIF_OP_DISCARD: + default: + break; + } + if (xen_block_send_response(request)) { + Error *local_err = NULL; + + xen_device_notify_event_channel(dataplane->xendev, + dataplane->event_channel, + &local_err); + if (local_err) { + error_report_err(local_err); + } + } + xen_block_release_request(request); + + qemu_bh_schedule(dataplane->bh); + +done: + aio_context_release(dataplane->ctx); +} + +static bool xen_block_split_discard(XenBlockRequest *request, + blkif_sector_t sector_number, + uint64_t nr_sectors) +{ + XenBlockDataPlane *dataplane = request->dataplane; + int64_t byte_offset; + int byte_chunk; + uint64_t byte_remaining, limit; + uint64_t sec_start = sector_number; + uint64_t sec_count = nr_sectors; + + /* Wrap around, or overflowing byte limit? */ + if (sec_start + sec_count < sec_count || + sec_start + sec_count > INT64_MAX / dataplane->file_blk) { + return false; + } + + limit = BDRV_REQUEST_MAX_SECTORS * dataplane->file_blk; + byte_offset = sec_start * dataplane->file_blk; + byte_remaining = sec_count * dataplane->file_blk; + + do { + byte_chunk = byte_remaining > limit ? limit : byte_remaining; + request->aio_inflight++; + blk_aio_pdiscard(dataplane->blk, byte_offset, byte_chunk, + xen_block_complete_aio, request); + byte_remaining -= byte_chunk; + byte_offset += byte_chunk; + } while (byte_remaining > 0); + + return true; +} + +static int xen_block_do_aio(XenBlockRequest *request) +{ + XenBlockDataPlane *dataplane = request->dataplane; + + if (request->req.nr_segments && + (request->req.operation == BLKIF_OP_WRITE || + request->req.operation == BLKIF_OP_FLUSH_DISKCACHE) && + xen_block_copy_request(request)) { + goto err; + } + + request->aio_inflight++; + if (request->presync) { + blk_aio_flush(request->dataplane->blk, xen_block_complete_aio, + request); + return 0; + } + + switch (request->req.operation) { + case BLKIF_OP_READ: + qemu_iovec_add(&request->v, request->buf, request->size); + block_acct_start(blk_get_stats(dataplane->blk), &request->acct, + request->v.size, BLOCK_ACCT_READ); + request->aio_inflight++; + blk_aio_preadv(dataplane->blk, request->start, &request->v, 0, + xen_block_complete_aio, request); + break; + case BLKIF_OP_WRITE: + case BLKIF_OP_FLUSH_DISKCACHE: + if (!request->req.nr_segments) { + break; + } + + qemu_iovec_add(&request->v, request->buf, request->size); + block_acct_start(blk_get_stats(dataplane->blk), &request->acct, + request->v.size, + request->req.operation == BLKIF_OP_WRITE ? + BLOCK_ACCT_WRITE : BLOCK_ACCT_FLUSH); + request->aio_inflight++; + blk_aio_pwritev(dataplane->blk, request->start, &request->v, 0, + xen_block_complete_aio, request); + break; + case BLKIF_OP_DISCARD: + { + struct blkif_request_discard *req = (void *)&request->req; + if (!xen_block_split_discard(request, req->sector_number, + req->nr_sectors)) { + goto err; + } + break; + } + default: + /* unknown operation (shouldn't happen -- parse catches this) */ + goto err; + } + + xen_block_complete_aio(request, 0); + + return 0; + +err: + xen_block_finish_request(request); + request->status = BLKIF_RSP_ERROR; + return -1; +} + +static int xen_block_send_response(XenBlockRequest *request) +{ + XenBlockDataPlane *dataplane = request->dataplane; + int send_notify = 0; + int have_requests = 0; + blkif_response_t *resp; + + /* Place on the response ring for the relevant domain. */ + switch (dataplane->protocol) { + case BLKIF_PROTOCOL_NATIVE: + resp = (blkif_response_t *)RING_GET_RESPONSE( + &dataplane->rings.native, + dataplane->rings.native.rsp_prod_pvt); + break; + case BLKIF_PROTOCOL_X86_32: + resp = (blkif_response_t *)RING_GET_RESPONSE( + &dataplane->rings.x86_32_part, + dataplane->rings.x86_32_part.rsp_prod_pvt); + break; + case BLKIF_PROTOCOL_X86_64: + resp = (blkif_response_t *)RING_GET_RESPONSE( + &dataplane->rings.x86_64_part, + dataplane->rings.x86_64_part.rsp_prod_pvt); + break; + default: + return 0; + } + + resp->id = request->req.id; + resp->operation = request->req.operation; + resp->status = request->status; + + dataplane->rings.common.rsp_prod_pvt++; + + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&dataplane->rings.common, + send_notify); + if (dataplane->rings.common.rsp_prod_pvt == + dataplane->rings.common.req_cons) { + /* + * Tail check for pending requests. Allows frontend to avoid + * notifications if requests are already in flight (lower + * overheads and promotes batching). + */ + RING_FINAL_CHECK_FOR_REQUESTS(&dataplane->rings.common, + have_requests); + } else if (RING_HAS_UNCONSUMED_REQUESTS(&dataplane->rings.common)) { + have_requests = 1; + } + + if (have_requests) { + dataplane->more_work++; + } + return send_notify; +} + +static int xen_block_get_request(XenBlockDataPlane *dataplane, + XenBlockRequest *request, RING_IDX rc) +{ + switch (dataplane->protocol) { + case BLKIF_PROTOCOL_NATIVE: { + blkif_request_t *req = + RING_GET_REQUEST(&dataplane->rings.native, rc); + + memcpy(&request->req, req, sizeof(request->req)); + break; + } + case BLKIF_PROTOCOL_X86_32: { + blkif_x86_32_request_t *req = + RING_GET_REQUEST(&dataplane->rings.x86_32_part, rc); + + blkif_get_x86_32_req(&request->req, req); + break; + } + case BLKIF_PROTOCOL_X86_64: { + blkif_x86_64_request_t *req = + RING_GET_REQUEST(&dataplane->rings.x86_64_part, rc); + + blkif_get_x86_64_req(&request->req, req); + break; + } + } + /* Prevent the compiler from accessing the on-ring fields instead. */ + barrier(); + return 0; +} + +/* + * Threshold of in-flight requests above which we will start using + * blk_io_plug()/blk_io_unplug() to batch requests. + */ +#define IO_PLUG_THRESHOLD 1 + +static void xen_block_handle_requests(XenBlockDataPlane *dataplane) +{ + RING_IDX rc, rp; + XenBlockRequest *request; + int inflight_atstart = dataplane->requests_inflight; + int batched = 0; + + dataplane->more_work = 0; + + rc = dataplane->rings.common.req_cons; + rp = dataplane->rings.common.sring->req_prod; + xen_rmb(); /* Ensure we see queued requests up to 'rp'. */ + + /* + * If there was more than IO_PLUG_THRESHOLD requests in flight + * when we got here, this is an indication that there the bottleneck + * is below us, so it's worth beginning to batch up I/O requests + * rather than submitting them immediately. The maximum number + * of requests we're willing to batch is the number already in + * flight, so it can grow up to max_requests when the bottleneck + * is below us. + */ + if (inflight_atstart > IO_PLUG_THRESHOLD) { + blk_io_plug(dataplane->blk); + } + while (rc != rp) { + /* pull request from ring */ + if (RING_REQUEST_CONS_OVERFLOW(&dataplane->rings.common, rc)) { + break; + } + request = xen_block_start_request(dataplane); + if (request == NULL) { + dataplane->more_work++; + break; + } + xen_block_get_request(dataplane, request, rc); + dataplane->rings.common.req_cons = ++rc; + + /* parse them */ + if (xen_block_parse_request(request) != 0) { + switch (request->req.operation) { + case BLKIF_OP_READ: + block_acct_invalid(blk_get_stats(dataplane->blk), + BLOCK_ACCT_READ); + break; + case BLKIF_OP_WRITE: + block_acct_invalid(blk_get_stats(dataplane->blk), + BLOCK_ACCT_WRITE); + break; + case BLKIF_OP_FLUSH_DISKCACHE: + block_acct_invalid(blk_get_stats(dataplane->blk), + BLOCK_ACCT_FLUSH); + default: + break; + }; + + if (xen_block_send_response(request)) { + Error *local_err = NULL; + + xen_device_notify_event_channel(dataplane->xendev, + dataplane->event_channel, + &local_err); + if (local_err) { + error_report_err(local_err); + } + } + xen_block_release_request(request); + continue; + } + + if (inflight_atstart > IO_PLUG_THRESHOLD && + batched >= inflight_atstart) { + blk_io_unplug(dataplane->blk); + } + xen_block_do_aio(request); + if (inflight_atstart > IO_PLUG_THRESHOLD) { + if (batched >= inflight_atstart) { + blk_io_plug(dataplane->blk); + batched = 0; + } else { + batched++; + } + } + } + if (inflight_atstart > IO_PLUG_THRESHOLD) { + blk_io_unplug(dataplane->blk); + } + + if (dataplane->more_work && + dataplane->requests_inflight < dataplane->max_requests) { + qemu_bh_schedule(dataplane->bh); + } +} + +static void xen_block_dataplane_bh(void *opaque) +{ + XenBlockDataPlane *dataplane = opaque; + + aio_context_acquire(dataplane->ctx); + xen_block_handle_requests(dataplane); + aio_context_release(dataplane->ctx); +} + +static void xen_block_dataplane_event(void *opaque) +{ + XenBlockDataPlane *dataplane = opaque; + + qemu_bh_schedule(dataplane->bh); +} + +XenBlockDataPlane *xen_block_dataplane_create(XenDevice *xendev, + BlockConf *conf, + IOThread *iothread) +{ + XenBlockDataPlane *dataplane = g_new0(XenBlockDataPlane, 1); + + dataplane->xendev = xendev; + dataplane->file_blk = conf->logical_block_size; + dataplane->blk = conf->blk; + dataplane->file_size = blk_getlength(dataplane->blk); + + QLIST_INIT(&dataplane->inflight); + QLIST_INIT(&dataplane->freelist); + + if (iothread) { + dataplane->iothread = iothread; + object_ref(OBJECT(dataplane->iothread)); + dataplane->ctx = iothread_get_aio_context(dataplane->iothread); + } else { + dataplane->ctx = qemu_get_aio_context(); + } + dataplane->bh = aio_bh_new(dataplane->ctx, xen_block_dataplane_bh, + dataplane); + + return dataplane; +} + +void xen_block_dataplane_destroy(XenBlockDataPlane *dataplane) +{ + XenBlockRequest *request; + + if (!dataplane) { + return; + } + + while (!QLIST_EMPTY(&dataplane->freelist)) { + request = QLIST_FIRST(&dataplane->freelist); + QLIST_REMOVE(request, list); + qemu_iovec_destroy(&request->v); + qemu_vfree(request->buf); + g_free(request); + } + + qemu_bh_delete(dataplane->bh); + if (dataplane->iothread) { + object_unref(OBJECT(dataplane->iothread)); + } + + g_free(dataplane); +} + +void xen_block_dataplane_stop(XenBlockDataPlane *dataplane) +{ + XenDevice *xendev; + + if (!dataplane) { + return; + } + + aio_context_acquire(dataplane->ctx); + blk_set_aio_context(dataplane->blk, qemu_get_aio_context()); + aio_context_release(dataplane->ctx); + + xendev = dataplane->xendev; + + if (dataplane->event_channel) { + Error *local_err = NULL; + + xen_device_unbind_event_channel(xendev, dataplane->event_channel, + &local_err); + dataplane->event_channel = NULL; + + if (local_err) { + error_report_err(local_err); + } + } + + if (dataplane->sring) { + Error *local_err = NULL; + + xen_device_unmap_grant_refs(xendev, dataplane->sring, + dataplane->nr_ring_ref, &local_err); + dataplane->sring = NULL; + + if (local_err) { + error_report_err(local_err); + } + } + + g_free(dataplane->ring_ref); + dataplane->ring_ref = NULL; +} + +void xen_block_dataplane_start(XenBlockDataPlane *dataplane, + const unsigned int ring_ref[], + unsigned int nr_ring_ref, + unsigned int event_channel, + unsigned int protocol, + Error **errp) +{ + XenDevice *xendev = dataplane->xendev; + Error *local_err = NULL; + unsigned int ring_size; + unsigned int i; + + dataplane->nr_ring_ref = nr_ring_ref; + dataplane->ring_ref = g_new(unsigned int, nr_ring_ref); + + for (i = 0; i < nr_ring_ref; i++) { + dataplane->ring_ref[i] = ring_ref[i]; + } + + dataplane->protocol = protocol; + + ring_size = XC_PAGE_SIZE * dataplane->nr_ring_ref; + switch (dataplane->protocol) { + case BLKIF_PROTOCOL_NATIVE: + { + dataplane->max_requests = __CONST_RING_SIZE(blkif, ring_size); + break; + } + case BLKIF_PROTOCOL_X86_32: + { + dataplane->max_requests = __CONST_RING_SIZE(blkif_x86_32, ring_size); + break; + } + case BLKIF_PROTOCOL_X86_64: + { + dataplane->max_requests = __CONST_RING_SIZE(blkif_x86_64, ring_size); + break; + } + default: + error_setg(errp, "unknown protocol %u", dataplane->protocol); + return; + } + + xen_device_set_max_grant_refs(xendev, dataplane->nr_ring_ref, + &local_err); + if (local_err) { + error_propagate(errp, local_err); + goto stop; + } + + dataplane->sring = xen_device_map_grant_refs(xendev, + dataplane->ring_ref, + dataplane->nr_ring_ref, + PROT_READ | PROT_WRITE, + &local_err); + if (local_err) { + error_propagate(errp, local_err); + goto stop; + } + + switch (dataplane->protocol) { + case BLKIF_PROTOCOL_NATIVE: + { + blkif_sring_t *sring_native = dataplane->sring; + + BACK_RING_INIT(&dataplane->rings.native, sring_native, ring_size); + break; + } + case BLKIF_PROTOCOL_X86_32: + { + blkif_x86_32_sring_t *sring_x86_32 = dataplane->sring; + + BACK_RING_INIT(&dataplane->rings.x86_32_part, sring_x86_32, + ring_size); + break; + } + case BLKIF_PROTOCOL_X86_64: + { + blkif_x86_64_sring_t *sring_x86_64 = dataplane->sring; + + BACK_RING_INIT(&dataplane->rings.x86_64_part, sring_x86_64, + ring_size); + break; + } + } + + dataplane->event_channel = + xen_device_bind_event_channel(xendev, event_channel, + xen_block_dataplane_event, dataplane, + &local_err); + if (local_err) { + error_propagate(errp, local_err); + goto stop; + } + + aio_context_acquire(dataplane->ctx); + blk_set_aio_context(dataplane->blk, dataplane->ctx); + aio_context_release(dataplane->ctx); + return; + +stop: + xen_block_dataplane_stop(dataplane); +} diff --git a/hw/block/dataplane/xen-block.h b/hw/block/dataplane/xen-block.h new file mode 100644 index 0000000000..d6fa6d26dd --- /dev/null +++ b/hw/block/dataplane/xen-block.h @@ -0,0 +1,29 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef HW_BLOCK_DATAPLANE_XEN_BLOCK_H +#define HW_BLOCK_DATAPLANE_XEN_BLOCK_H + +#include "hw/block/block.h" +#include "hw/xen/xen-bus.h" +#include "sysemu/iothread.h" + +typedef struct XenBlockDataPlane XenBlockDataPlane; + +XenBlockDataPlane *xen_block_dataplane_create(XenDevice *xendev, + BlockConf *conf, + IOThread *iothread); +void xen_block_dataplane_destroy(XenBlockDataPlane *dataplane); +void xen_block_dataplane_start(XenBlockDataPlane *dataplane, + const unsigned int ring_ref[], + unsigned int nr_ring_ref, + unsigned int event_channel, + unsigned int protocol, + Error **errp); +void xen_block_dataplane_stop(XenBlockDataPlane *dataplane); + +#endif /* HW_BLOCK_DATAPLANE_XEN_BLOCK_H */ diff --git a/hw/block/trace-events b/hw/block/trace-events index 335c092450..55e5a5500c 100644 --- a/hw/block/trace-events +++ b/hw/block/trace-events @@ -127,3 +127,17 @@ xen_disk_init(char *name) "%s" xen_disk_connect(char *name) "%s" xen_disk_disconnect(char *name) "%s" xen_disk_free(char *name) "%s" + +# hw/block/xen-block.c +xen_block_realize(const char *type, uint32_t disk, uint32_t partition) "%s d%up%u" +xen_block_connect(const char *type, uint32_t disk, uint32_t partition) "%s d%up%u" +xen_block_disconnect(const char *type, uint32_t disk, uint32_t partition) "%s d%up%u" +xen_block_unrealize(const char *type, uint32_t disk, uint32_t partition) "%s d%up%u" +xen_disk_realize(void) "" +xen_disk_unrealize(void) "" +xen_cdrom_realize(void) "" +xen_cdrom_unrealize(void) "" +xen_block_blockdev_add(char *str) "%s" +xen_block_blockdev_del(const char *node_name) "%s" +xen_block_device_create(unsigned int number) "%u" +xen_block_device_destroy(unsigned int number) "%u" diff --git a/hw/block/xen-block.c b/hw/block/xen-block.c new file mode 100644 index 0000000000..be28b63442 --- /dev/null +++ b/hw/block/xen-block.c @@ -0,0 +1,963 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" +#include "qemu/cutils.h" +#include "qemu/option.h" +#include "qapi/error.h" +#include "qapi/qapi-commands-block-core.h" +#include "qapi/qapi-commands-misc.h" +#include "qapi/qapi-visit-block-core.h" +#include "qapi/qobject-input-visitor.h" +#include "qapi/visitor.h" +#include "qapi/qmp/qdict.h" +#include "qapi/qmp/qstring.h" +#include "hw/hw.h" +#include "hw/xen/xen_common.h" +#include "hw/block/xen_blkif.h" +#include "hw/xen/xen-block.h" +#include "hw/xen/xen-backend.h" +#include "sysemu/blockdev.h" +#include "sysemu/block-backend.h" +#include "sysemu/iothread.h" +#include "dataplane/xen-block.h" +#include "trace.h" + +static char *xen_block_get_name(XenDevice *xendev, Error **errp) +{ + XenBlockDevice *blockdev = XEN_BLOCK_DEVICE(xendev); + XenBlockVdev *vdev = &blockdev->props.vdev; + + return g_strdup_printf("%lu", vdev->number); +} + +static void xen_block_disconnect(XenDevice *xendev, Error **errp) +{ + XenBlockDevice *blockdev = XEN_BLOCK_DEVICE(xendev); + const char *type = object_get_typename(OBJECT(blockdev)); + XenBlockVdev *vdev = &blockdev->props.vdev; + + trace_xen_block_disconnect(type, vdev->disk, vdev->partition); + + xen_block_dataplane_stop(blockdev->dataplane); +} + +static void xen_block_connect(XenDevice *xendev, Error **errp) +{ + XenBlockDevice *blockdev = XEN_BLOCK_DEVICE(xendev); + const char *type = object_get_typename(OBJECT(blockdev)); + XenBlockVdev *vdev = &blockdev->props.vdev; + unsigned int order, nr_ring_ref, *ring_ref, event_channel, protocol; + char *str; + + trace_xen_block_connect(type, vdev->disk, vdev->partition); + + if (xen_device_frontend_scanf(xendev, "ring-page-order", "%u", + &order) != 1) { + nr_ring_ref = 1; + ring_ref = g_new(unsigned int, nr_ring_ref); + + if (xen_device_frontend_scanf(xendev, "ring-ref", "%u", + &ring_ref[0]) != 1) { + error_setg(errp, "failed to read ring-ref"); + g_free(ring_ref); + return; + } + } else if (order <= blockdev->props.max_ring_page_order) { + unsigned int i; + + nr_ring_ref = 1 << order; + ring_ref = g_new(unsigned int, nr_ring_ref); + + for (i = 0; i < nr_ring_ref; i++) { + const char *key = g_strdup_printf("ring-ref%u", i); + + if (xen_device_frontend_scanf(xendev, key, "%u", + &ring_ref[i]) != 1) { + error_setg(errp, "failed to read %s", key); + g_free((gpointer)key); + g_free(ring_ref); + return; + } + + g_free((gpointer)key); + } + } else { + error_setg(errp, "invalid ring-page-order (%d)", order); + return; + } + + if (xen_device_frontend_scanf(xendev, "event-channel", "%u", + &event_channel) != 1) { + error_setg(errp, "failed to read event-channel"); + g_free(ring_ref); + return; + } + + if (xen_device_frontend_scanf(xendev, "protocol", "%ms", + &str) != 1) { + protocol = BLKIF_PROTOCOL_NATIVE; + } else { + if (strcmp(str, XEN_IO_PROTO_ABI_X86_32) == 0) { + protocol = BLKIF_PROTOCOL_X86_32; + } else if (strcmp(str, XEN_IO_PROTO_ABI_X86_64) == 0) { + protocol = BLKIF_PROTOCOL_X86_64; + } else { + protocol = BLKIF_PROTOCOL_NATIVE; + } + + free(str); + } + + xen_block_dataplane_start(blockdev->dataplane, ring_ref, nr_ring_ref, + event_channel, protocol, errp); + + g_free(ring_ref); +} + +static void xen_block_unrealize(XenDevice *xendev, Error **errp) +{ + XenBlockDevice *blockdev = XEN_BLOCK_DEVICE(xendev); + XenBlockDeviceClass *blockdev_class = + XEN_BLOCK_DEVICE_GET_CLASS(xendev); + const char *type = object_get_typename(OBJECT(blockdev)); + XenBlockVdev *vdev = &blockdev->props.vdev; + + if (vdev->type == XEN_BLOCK_VDEV_TYPE_INVALID) { + return; + } + + trace_xen_block_unrealize(type, vdev->disk, vdev->partition); + + /* Disconnect from the frontend in case this has not already happened */ + xen_block_disconnect(xendev, NULL); + + xen_block_dataplane_destroy(blockdev->dataplane); + blockdev->dataplane = NULL; + + if (blockdev_class->unrealize) { + blockdev_class->unrealize(blockdev, errp); + } +} + +static void xen_block_realize(XenDevice *xendev, Error **errp) +{ + XenBlockDevice *blockdev = XEN_BLOCK_DEVICE(xendev); + XenBlockDeviceClass *blockdev_class = + XEN_BLOCK_DEVICE_GET_CLASS(xendev); + const char *type = object_get_typename(OBJECT(blockdev)); + XenBlockVdev *vdev = &blockdev->props.vdev; + BlockConf *conf = &blockdev->props.conf; + Error *local_err = NULL; + + if (vdev->type == XEN_BLOCK_VDEV_TYPE_INVALID) { + error_setg(errp, "vdev property not set"); + return; + } + + trace_xen_block_realize(type, vdev->disk, vdev->partition); + + if (blockdev_class->realize) { + blockdev_class->realize(blockdev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + return; + } + } + + /* + * The blkif protocol does not deal with removable media, so it must + * always be present, even for CDRom devices. + */ + assert(conf->blk); + if (!blk_is_inserted(conf->blk)) { + error_setg(errp, "device needs media, but drive is empty"); + return; + } + + if (!blkconf_apply_backend_options(conf, blockdev->info & VDISK_READONLY, + false, errp)) { + return; + } + + if (!(blockdev->info & VDISK_CDROM) && + !blkconf_geometry(conf, NULL, 65535, 255, 255, errp)) { + return; + } + + blkconf_blocksizes(conf); + + if (conf->logical_block_size > conf->physical_block_size) { + error_setg( + errp, "logical_block_size > physical_block_size not supported"); + return; + } + + blk_set_guest_block_size(conf->blk, conf->logical_block_size); + + if (conf->discard_granularity > 0) { + xen_device_backend_printf(xendev, "feature-discard", "%u", 1); + } + + xen_device_backend_printf(xendev, "feature-flush-cache", "%u", 1); + xen_device_backend_printf(xendev, "max-ring-page-order", "%u", + blockdev->props.max_ring_page_order); + xen_device_backend_printf(xendev, "info", "%u", blockdev->info); + + xen_device_frontend_printf(xendev, "virtual-device", "%lu", + vdev->number); + xen_device_frontend_printf(xendev, "device-type", "%s", + blockdev->device_type); + + xen_device_backend_printf(xendev, "sector-size", "%u", + conf->logical_block_size); + xen_device_backend_printf(xendev, "sectors", "%lu", + blk_getlength(conf->blk) / + conf->logical_block_size); + + blockdev->dataplane = + xen_block_dataplane_create(xendev, conf, blockdev->props.iothread); +} + +static void xen_block_frontend_changed(XenDevice *xendev, + enum xenbus_state frontend_state, + Error **errp) +{ + enum xenbus_state backend_state = xen_device_backend_get_state(xendev); + Error *local_err = NULL; + + switch (frontend_state) { + case XenbusStateInitialised: + case XenbusStateConnected: + if (backend_state == XenbusStateConnected) { + break; + } + + xen_block_disconnect(xendev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + break; + } + + xen_block_connect(xendev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + break; + } + + xen_device_backend_set_state(xendev, XenbusStateConnected); + break; + + case XenbusStateClosing: + xen_device_backend_set_state(xendev, XenbusStateClosing); + break; + + case XenbusStateClosed: + xen_block_disconnect(xendev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + break; + } + + xen_device_backend_set_state(xendev, XenbusStateClosed); + break; + + default: + break; + } +} + +static char *disk_to_vbd_name(unsigned int disk) +{ + char *name, *prefix = (disk >= 26) ? + disk_to_vbd_name((disk / 26) - 1) : g_strdup(""); + + name = g_strdup_printf("%s%c", prefix, 'a' + disk % 26); + g_free(prefix); + + return name; +} + +static void xen_block_get_vdev(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + DeviceState *dev = DEVICE(obj); + Property *prop = opaque; + XenBlockVdev *vdev = qdev_get_prop_ptr(dev, prop); + char *str; + + switch (vdev->type) { + case XEN_BLOCK_VDEV_TYPE_DP: + str = g_strdup_printf("d%lup%lu", vdev->disk, vdev->partition); + break; + + case XEN_BLOCK_VDEV_TYPE_XVD: + case XEN_BLOCK_VDEV_TYPE_HD: + case XEN_BLOCK_VDEV_TYPE_SD: { + char *name = disk_to_vbd_name(vdev->disk); + + str = g_strdup_printf("%s%s%lu", + (vdev->type == XEN_BLOCK_VDEV_TYPE_XVD) ? + "xvd" : + (vdev->type == XEN_BLOCK_VDEV_TYPE_HD) ? + "hd" : + "sd", + name, vdev->partition); + g_free(name); + break; + } + default: + error_setg(errp, "invalid vdev type"); + return; + } + + visit_type_str(v, name, &str, errp); + g_free(str); +} + +static unsigned int vbd_name_to_disk(const char *name, const char **endp) +{ + unsigned int disk = 0; + + while (*name != '\0') { + if (!g_ascii_isalpha(*name) || !g_ascii_islower(*name)) { + break; + } + + disk *= 26; + disk += *name++ - 'a' + 1; + } + *endp = name; + + return disk - 1; +} + +static void xen_block_set_vdev(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + DeviceState *dev = DEVICE(obj); + Property *prop = opaque; + XenBlockVdev *vdev = qdev_get_prop_ptr(dev, prop); + Error *local_err = NULL; + char *str, *p; + const char *end; + + if (dev->realized) { + qdev_prop_set_after_realize(dev, name, errp); + return; + } + + visit_type_str(v, name, &str, &local_err); + if (local_err) { + error_propagate(errp, local_err); + return; + } + + p = strchr(str, 'd'); + if (!p) { + goto invalid; + } + + *p++ = '\0'; + if (*str == '\0') { + vdev->type = XEN_BLOCK_VDEV_TYPE_DP; + } else if (strcmp(str, "xv") == 0) { + vdev->type = XEN_BLOCK_VDEV_TYPE_XVD; + } else if (strcmp(str, "h") == 0) { + vdev->type = XEN_BLOCK_VDEV_TYPE_HD; + } else if (strcmp(str, "s") == 0) { + vdev->type = XEN_BLOCK_VDEV_TYPE_SD; + } else { + goto invalid; + } + + if (vdev->type == XEN_BLOCK_VDEV_TYPE_DP) { + if (qemu_strtoul(p, &end, 10, &vdev->disk)) { + goto invalid; + } + + if (*end == 'p') { + p = (char *) ++end; + if (*end == '\0') { + goto invalid; + } + } + } else { + vdev->disk = vbd_name_to_disk(p, &end); + } + + if (*end != '\0') { + p = (char *)end; + + if (qemu_strtoul(p, &end, 10, &vdev->partition)) { + goto invalid; + } + + if (*end != '\0') { + goto invalid; + } + } else { + vdev->partition = 0; + } + + switch (vdev->type) { + case XEN_BLOCK_VDEV_TYPE_DP: + case XEN_BLOCK_VDEV_TYPE_XVD: + if (vdev->disk < (1 << 4) && vdev->partition < (1 << 4)) { + vdev->number = (202 << 8) | (vdev->disk << 4) | + vdev->partition; + } else if (vdev->disk < (1 << 20) && vdev->partition < (1 << 8)) { + vdev->number = (1 << 28) | (vdev->disk << 8) | + vdev->partition; + } else { + goto invalid; + } + break; + + case XEN_BLOCK_VDEV_TYPE_HD: + if ((vdev->disk == 0 || vdev->disk == 1) && + vdev->partition < (1 << 6)) { + vdev->number = (3 << 8) | (vdev->disk << 6) | vdev->partition; + } else if ((vdev->disk == 2 || vdev->disk == 3) && + vdev->partition < (1 << 6)) { + vdev->number = (22 << 8) | ((vdev->disk - 2) << 6) | + vdev->partition; + } else { + goto invalid; + } + break; + + case XEN_BLOCK_VDEV_TYPE_SD: + if (vdev->disk < (1 << 4) && vdev->partition < (1 << 4)) { + vdev->number = (8 << 8) | (vdev->disk << 4) | vdev->partition; + } else { + goto invalid; + } + break; + + default: + goto invalid; + } + + g_free(str); + return; + +invalid: + error_setg(errp, "invalid virtual disk specifier"); + + vdev->type = XEN_BLOCK_VDEV_TYPE_INVALID; + g_free(str); +} + +/* + * This property deals with 'vdev' names adhering to the Xen VBD naming + * scheme described in: + * + * https://xenbits.xen.org/docs/unstable/man/xen-vbd-interface.7.html + */ +const PropertyInfo xen_block_prop_vdev = { + .name = "str", + .description = "Virtual Disk specifier: d*p*/xvd*/hd*/sd*", + .get = xen_block_get_vdev, + .set = xen_block_set_vdev, +}; + +static Property xen_block_props[] = { + DEFINE_PROP("vdev", XenBlockDevice, props.vdev, + xen_block_prop_vdev, XenBlockVdev), + DEFINE_BLOCK_PROPERTIES(XenBlockDevice, props.conf), + DEFINE_PROP_UINT32("max-ring-page-order", XenBlockDevice, + props.max_ring_page_order, 4), + DEFINE_PROP_LINK("iothread", XenBlockDevice, props.iothread, + TYPE_IOTHREAD, IOThread *), + DEFINE_PROP_END_OF_LIST() +}; + +static void xen_block_class_init(ObjectClass *class, void *data) +{ + DeviceClass *dev_class = DEVICE_CLASS(class); + XenDeviceClass *xendev_class = XEN_DEVICE_CLASS(class); + + xendev_class->backend = "qdisk"; + xendev_class->device = "vbd"; + xendev_class->get_name = xen_block_get_name; + xendev_class->realize = xen_block_realize; + xendev_class->frontend_changed = xen_block_frontend_changed; + xendev_class->unrealize = xen_block_unrealize; + + dev_class->props = xen_block_props; +} + +static const TypeInfo xen_block_type_info = { + .name = TYPE_XEN_BLOCK_DEVICE, + .parent = TYPE_XEN_DEVICE, + .instance_size = sizeof(XenBlockDevice), + .abstract = true, + .class_size = sizeof(XenBlockDeviceClass), + .class_init = xen_block_class_init, +}; + +static void xen_disk_unrealize(XenBlockDevice *blockdev, Error **errp) +{ + trace_xen_disk_unrealize(); +} + +static void xen_disk_realize(XenBlockDevice *blockdev, Error **errp) +{ + BlockConf *conf = &blockdev->props.conf; + + trace_xen_disk_realize(); + + blockdev->device_type = "disk"; + + if (!conf->blk) { + error_setg(errp, "drive property not set"); + return; + } + + blockdev->info = blk_is_read_only(conf->blk) ? VDISK_READONLY : 0; +} + +static void xen_disk_class_init(ObjectClass *class, void *data) +{ + DeviceClass *dev_class = DEVICE_CLASS(class); + XenBlockDeviceClass *blockdev_class = XEN_BLOCK_DEVICE_CLASS(class); + + blockdev_class->realize = xen_disk_realize; + blockdev_class->unrealize = xen_disk_unrealize; + + dev_class->desc = "Xen Disk Device"; +} + +static const TypeInfo xen_disk_type_info = { + .name = TYPE_XEN_DISK_DEVICE, + .parent = TYPE_XEN_BLOCK_DEVICE, + .instance_size = sizeof(XenDiskDevice), + .class_init = xen_disk_class_init, +}; + +static void xen_cdrom_unrealize(XenBlockDevice *blockdev, Error **errp) +{ + trace_xen_cdrom_unrealize(); +} + +static void xen_cdrom_realize(XenBlockDevice *blockdev, Error **errp) +{ + BlockConf *conf = &blockdev->props.conf; + + trace_xen_cdrom_realize(); + + blockdev->device_type = "cdrom"; + + if (!conf->blk) { + int rc; + + /* Set up an empty drive */ + conf->blk = blk_new(0, BLK_PERM_ALL); + + rc = blk_attach_dev(conf->blk, DEVICE(blockdev)); + if (!rc) { + error_setg_errno(errp, -rc, "failed to create drive"); + return; + } + } + + blockdev->info = VDISK_READONLY | VDISK_CDROM; +} + +static void xen_cdrom_class_init(ObjectClass *class, void *data) +{ + DeviceClass *dev_class = DEVICE_CLASS(class); + XenBlockDeviceClass *blockdev_class = XEN_BLOCK_DEVICE_CLASS(class); + + blockdev_class->realize = xen_cdrom_realize; + blockdev_class->unrealize = xen_cdrom_unrealize; + + dev_class->desc = "Xen CD-ROM Device"; +} + +static const TypeInfo xen_cdrom_type_info = { + .name = TYPE_XEN_CDROM_DEVICE, + .parent = TYPE_XEN_BLOCK_DEVICE, + .instance_size = sizeof(XenCDRomDevice), + .class_init = xen_cdrom_class_init, +}; + +static void xen_block_register_types(void) +{ + type_register_static(&xen_block_type_info); + type_register_static(&xen_disk_type_info); + type_register_static(&xen_cdrom_type_info); +} + +type_init(xen_block_register_types) + +static void xen_block_blockdev_del(const char *node_name, Error **errp) +{ + trace_xen_block_blockdev_del(node_name); + + qmp_blockdev_del(node_name, errp); +} + +static char *xen_block_blockdev_add(const char *id, QDict *qdict, + Error **errp) +{ + const char *driver = qdict_get_try_str(qdict, "driver"); + BlockdevOptions *options = NULL; + Error *local_err = NULL; + char *node_name; + Visitor *v; + + if (!driver) { + error_setg(errp, "no 'driver' parameter"); + return NULL; + } + + node_name = g_strdup_printf("%s-%s", id, driver); + qdict_put_str(qdict, "node-name", node_name); + + trace_xen_block_blockdev_add(node_name); + + v = qobject_input_visitor_new(QOBJECT(qdict)); + visit_type_BlockdevOptions(v, NULL, &options, &local_err); + visit_free(v); + + if (local_err) { + error_propagate(errp, local_err); + goto fail; + } + + qmp_blockdev_add(options, &local_err); + + if (local_err) { + error_propagate(errp, local_err); + goto fail; + } + + qapi_free_BlockdevOptions(options); + + return node_name; + +fail: + if (options) { + qapi_free_BlockdevOptions(options); + } + g_free(node_name); + + return NULL; +} + +static void xen_block_drive_destroy(XenBlockDrive *drive, Error **errp) +{ + char *node_name = drive->node_name; + + if (node_name) { + Error *local_err = NULL; + + xen_block_blockdev_del(node_name, &local_err); + if (local_err) { + error_propagate(errp, local_err); + return; + } + g_free(node_name); + drive->node_name = NULL; + } + g_free(drive->id); + g_free(drive); +} + +static XenBlockDrive *xen_block_drive_create(const char *id, + const char *device_type, + QDict *opts, Error **errp) +{ + const char *params = qdict_get_try_str(opts, "params"); + const char *mode = qdict_get_try_str(opts, "mode"); + const char *direct_io_safe = qdict_get_try_str(opts, "direct-io-safe"); + const char *discard_enable = qdict_get_try_str(opts, "discard-enable"); + char *driver = NULL; + char *filename = NULL; + XenBlockDrive *drive = NULL; + Error *local_err = NULL; + QDict *file_layer; + QDict *driver_layer; + + if (params) { + char **v = g_strsplit(params, ":", 2); + + if (v[1] == NULL) { + filename = g_strdup(v[0]); + driver = g_strdup("raw"); + } else { + if (strcmp(v[0], "aio") == 0) { + driver = g_strdup("raw"); + } else if (strcmp(v[0], "vhd") == 0) { + driver = g_strdup("vpc"); + } else { + driver = g_strdup(v[0]); + } + filename = g_strdup(v[1]); + } + + g_strfreev(v); + } + + if (!filename) { + error_setg(errp, "no filename"); + goto done; + } + assert(driver); + + drive = g_new0(XenBlockDrive, 1); + drive->id = g_strdup(id); + + file_layer = qdict_new(); + + qdict_put_str(file_layer, "driver", "file"); + qdict_put_str(file_layer, "filename", filename); + + if (mode && *mode != 'w') { + qdict_put_bool(file_layer, "read-only", true); + } + + if (direct_io_safe) { + unsigned long value; + + if (!qemu_strtoul(direct_io_safe, NULL, 2, &value) && !!value) { + QDict *cache_qdict = qdict_new(); + + qdict_put_bool(cache_qdict, "direct", true); + qdict_put_obj(file_layer, "cache", QOBJECT(cache_qdict)); + + qdict_put_str(file_layer, "aio", "native"); + } + } + + if (discard_enable) { + unsigned long value; + + if (!qemu_strtoul(discard_enable, NULL, 2, &value) && !!value) { + qdict_put_str(file_layer, "discard", "unmap"); + } + } + + /* + * It is necessary to turn file locking off as an emulated device + * may have already opened the same image file. + */ + qdict_put_str(file_layer, "locking", "off"); + + driver_layer = qdict_new(); + + qdict_put_str(driver_layer, "driver", driver); + qdict_put_obj(driver_layer, "file", QOBJECT(file_layer)); + + g_assert(!drive->node_name); + drive->node_name = xen_block_blockdev_add(drive->id, driver_layer, + &local_err); + +done: + g_free(driver); + g_free(filename); + + if (local_err) { + error_propagate(errp, local_err); + xen_block_drive_destroy(drive, NULL); + return NULL; + } + + return drive; +} + +static const char *xen_block_drive_get_node_name(XenBlockDrive *drive) +{ + return drive->node_name ? drive->node_name : ""; +} + +static void xen_block_iothread_destroy(XenBlockIOThread *iothread, + Error **errp) +{ + qmp_object_del(iothread->id, errp); + + g_free(iothread->id); + g_free(iothread); +} + +static XenBlockIOThread *xen_block_iothread_create(const char *id, + Error **errp) +{ + XenBlockIOThread *iothread = g_new(XenBlockIOThread, 1); + Error *local_err = NULL; + + iothread->id = g_strdup(id); + + qmp_object_add(TYPE_IOTHREAD, id, false, NULL, &local_err); + if (local_err) { + error_propagate(errp, local_err); + + g_free(iothread->id); + g_free(iothread); + return NULL; + } + + return iothread; +} + +static void xen_block_device_create(XenBackendInstance *backend, + QDict *opts, Error **errp) +{ + XenBus *xenbus = xen_backend_get_bus(backend); + const char *name = xen_backend_get_name(backend); + unsigned long number; + const char *vdev, *device_type; + XenBlockDrive *drive = NULL; + XenBlockIOThread *iothread = NULL; + XenDevice *xendev = NULL; + Error *local_err = NULL; + const char *type; + XenBlockDevice *blockdev; + + if (qemu_strtoul(name, NULL, 10, &number)) { + error_setg(errp, "failed to parse name '%s'", name); + goto fail; + } + + trace_xen_block_device_create(number); + + vdev = qdict_get_try_str(opts, "dev"); + if (!vdev) { + error_setg(errp, "no dev parameter"); + goto fail; + } + + device_type = qdict_get_try_str(opts, "device-type"); + if (!device_type) { + error_setg(errp, "no device-type parameter"); + goto fail; + } + + if (!strcmp(device_type, "disk")) { + type = TYPE_XEN_DISK_DEVICE; + } else if (!strcmp(device_type, "cdrom")) { + type = TYPE_XEN_CDROM_DEVICE; + } else { + error_setg(errp, "invalid device-type parameter '%s'", device_type); + goto fail; + } + + drive = xen_block_drive_create(vdev, device_type, opts, &local_err); + if (!drive) { + error_propagate_prepend(errp, local_err, "failed to create drive: "); + goto fail; + } + + iothread = xen_block_iothread_create(vdev, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to create iothread: "); + goto fail; + } + + xendev = XEN_DEVICE(qdev_create(BUS(xenbus), type)); + blockdev = XEN_BLOCK_DEVICE(xendev); + + object_property_set_str(OBJECT(xendev), vdev, "vdev", &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, "failed to set 'vdev': "); + goto fail; + } + + object_property_set_str(OBJECT(xendev), + xen_block_drive_get_node_name(drive), "drive", + &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, "failed to set 'drive': "); + goto fail; + } + + object_property_set_str(OBJECT(xendev), iothread->id, "iothread", + &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to set 'iothread': "); + goto fail; + } + + blockdev->iothread = iothread; + blockdev->drive = drive; + + object_property_set_bool(OBJECT(xendev), true, "realized", &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "realization of device %s failed: ", + type); + goto fail; + } + + xen_backend_set_device(backend, xendev); + return; + +fail: + if (xendev) { + object_unparent(OBJECT(xendev)); + } + + if (iothread) { + xen_block_iothread_destroy(iothread, NULL); + } + + if (drive) { + xen_block_drive_destroy(drive, NULL); + } +} + +static void xen_block_device_destroy(XenBackendInstance *backend, + Error **errp) +{ + XenDevice *xendev = xen_backend_get_device(backend); + XenBlockDevice *blockdev = XEN_BLOCK_DEVICE(xendev); + XenBlockVdev *vdev = &blockdev->props.vdev; + XenBlockDrive *drive = blockdev->drive; + XenBlockIOThread *iothread = blockdev->iothread; + + trace_xen_block_device_destroy(vdev->number); + + object_unparent(OBJECT(xendev)); + + if (iothread) { + Error *local_err = NULL; + + xen_block_iothread_destroy(iothread, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to destroy iothread: "); + return; + } + } + + if (drive) { + Error *local_err = NULL; + + xen_block_drive_destroy(drive, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to destroy drive: "); + } + } +} + +static const XenBackendInfo xen_block_backend_info = { + .type = "qdisk", + .create = xen_block_device_create, + .destroy = xen_block_device_destroy, +}; + +static void xen_block_register_backend(void) +{ + xen_backend_register(&xen_block_backend_info); +} + +xen_backend_init(xen_block_register_backend); diff --git a/hw/block/xen_disk.c b/hw/block/xen_disk.c deleted file mode 100644 index 2a254b99d0..0000000000 --- a/hw/block/xen_disk.c +++ /dev/null @@ -1,1011 +0,0 @@ -/* - * xen paravirt block device backend - * - * (c) Gerd Hoffmann - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; under version 2 of the License. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, see . - * - * Contributions after 2012-01-13 are licensed under the terms of the - * GNU GPL, version 2 or (at your option) any later version. - */ - -#include "qemu/osdep.h" -#include "qemu/units.h" -#include -#include - -#include "hw/hw.h" -#include "hw/xen/xen_backend.h" -#include "xen_blkif.h" -#include "sysemu/blockdev.h" -#include "sysemu/iothread.h" -#include "sysemu/block-backend.h" -#include "qapi/error.h" -#include "qapi/qmp/qdict.h" -#include "qapi/qmp/qstring.h" -#include "trace.h" - -/* ------------------------------------------------------------- */ - -#define BLOCK_SIZE 512 -#define IOCB_COUNT (BLKIF_MAX_SEGMENTS_PER_REQUEST + 2) - -struct ioreq { - blkif_request_t req; - int16_t status; - - /* parsed request */ - off_t start; - QEMUIOVector v; - void *buf; - size_t size; - int presync; - - /* aio status */ - int aio_inflight; - int aio_errors; - - struct XenBlkDev *blkdev; - QLIST_ENTRY(ioreq) list; - BlockAcctCookie acct; -}; - -#define MAX_RING_PAGE_ORDER 4 - -struct XenBlkDev { - struct XenDevice xendev; /* must be first */ - char *params; - char *mode; - char *type; - char *dev; - char *devtype; - bool directiosafe; - const char *fileproto; - const char *filename; - unsigned int ring_ref[1 << MAX_RING_PAGE_ORDER]; - unsigned int nr_ring_ref; - void *sring; - int64_t file_blk; - int64_t file_size; - int protocol; - blkif_back_rings_t rings; - int more_work; - - /* request lists */ - QLIST_HEAD(, ioreq) inflight; - QLIST_HEAD(, ioreq) finished; - QLIST_HEAD(, ioreq) freelist; - int requests_total; - int requests_inflight; - int requests_finished; - unsigned int max_requests; - - gboolean feature_discard; - - /* qemu block driver */ - DriveInfo *dinfo; - BlockBackend *blk; - QEMUBH *bh; - - IOThread *iothread; - AioContext *ctx; -}; - -/* ------------------------------------------------------------- */ - -static void ioreq_reset(struct ioreq *ioreq) -{ - memset(&ioreq->req, 0, sizeof(ioreq->req)); - ioreq->status = 0; - ioreq->start = 0; - ioreq->buf = NULL; - ioreq->size = 0; - ioreq->presync = 0; - - ioreq->aio_inflight = 0; - ioreq->aio_errors = 0; - - ioreq->blkdev = NULL; - memset(&ioreq->list, 0, sizeof(ioreq->list)); - memset(&ioreq->acct, 0, sizeof(ioreq->acct)); - - qemu_iovec_reset(&ioreq->v); -} - -static struct ioreq *ioreq_start(struct XenBlkDev *blkdev) -{ - struct ioreq *ioreq = NULL; - - if (QLIST_EMPTY(&blkdev->freelist)) { - if (blkdev->requests_total >= blkdev->max_requests) { - goto out; - } - /* allocate new struct */ - ioreq = g_malloc0(sizeof(*ioreq)); - ioreq->blkdev = blkdev; - blkdev->requests_total++; - qemu_iovec_init(&ioreq->v, 1); - } else { - /* get one from freelist */ - ioreq = QLIST_FIRST(&blkdev->freelist); - QLIST_REMOVE(ioreq, list); - } - QLIST_INSERT_HEAD(&blkdev->inflight, ioreq, list); - blkdev->requests_inflight++; - -out: - return ioreq; -} - -static void ioreq_finish(struct ioreq *ioreq) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - - QLIST_REMOVE(ioreq, list); - QLIST_INSERT_HEAD(&blkdev->finished, ioreq, list); - blkdev->requests_inflight--; - blkdev->requests_finished++; -} - -static void ioreq_release(struct ioreq *ioreq, bool finish) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - - QLIST_REMOVE(ioreq, list); - ioreq_reset(ioreq); - ioreq->blkdev = blkdev; - QLIST_INSERT_HEAD(&blkdev->freelist, ioreq, list); - if (finish) { - blkdev->requests_finished--; - } else { - blkdev->requests_inflight--; - } -} - -/* - * translate request into iovec + start offset - * do sanity checks along the way - */ -static int ioreq_parse(struct ioreq *ioreq) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - struct XenDevice *xendev = &blkdev->xendev; - size_t len; - int i; - - xen_pv_printf(xendev, 3, - "op %d, nr %d, handle %d, id %" PRId64 ", sector %" PRId64 "\n", - ioreq->req.operation, ioreq->req.nr_segments, - ioreq->req.handle, ioreq->req.id, ioreq->req.sector_number); - switch (ioreq->req.operation) { - case BLKIF_OP_READ: - break; - case BLKIF_OP_FLUSH_DISKCACHE: - ioreq->presync = 1; - if (!ioreq->req.nr_segments) { - return 0; - } - /* fall through */ - case BLKIF_OP_WRITE: - break; - case BLKIF_OP_DISCARD: - return 0; - default: - xen_pv_printf(xendev, 0, "error: unknown operation (%d)\n", - ioreq->req.operation); - goto err; - }; - - if (ioreq->req.operation != BLKIF_OP_READ && blkdev->mode[0] != 'w') { - xen_pv_printf(xendev, 0, "error: write req for ro device\n"); - goto err; - } - - ioreq->start = ioreq->req.sector_number * blkdev->file_blk; - for (i = 0; i < ioreq->req.nr_segments; i++) { - if (i == BLKIF_MAX_SEGMENTS_PER_REQUEST) { - xen_pv_printf(xendev, 0, "error: nr_segments too big\n"); - goto err; - } - if (ioreq->req.seg[i].first_sect > ioreq->req.seg[i].last_sect) { - xen_pv_printf(xendev, 0, "error: first > last sector\n"); - goto err; - } - if (ioreq->req.seg[i].last_sect * BLOCK_SIZE >= XC_PAGE_SIZE) { - xen_pv_printf(xendev, 0, "error: page crossing\n"); - goto err; - } - - len = (ioreq->req.seg[i].last_sect - ioreq->req.seg[i].first_sect + 1) * blkdev->file_blk; - ioreq->size += len; - } - if (ioreq->start + ioreq->size > blkdev->file_size) { - xen_pv_printf(xendev, 0, "error: access beyond end of file\n"); - goto err; - } - return 0; - -err: - ioreq->status = BLKIF_RSP_ERROR; - return -1; -} - -static int ioreq_grant_copy(struct ioreq *ioreq) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - struct XenDevice *xendev = &blkdev->xendev; - XenGrantCopySegment segs[BLKIF_MAX_SEGMENTS_PER_REQUEST]; - int i, count, rc; - int64_t file_blk = blkdev->file_blk; - bool to_domain = (ioreq->req.operation == BLKIF_OP_READ); - void *virt = ioreq->buf; - - if (ioreq->req.nr_segments == 0) { - return 0; - } - - count = ioreq->req.nr_segments; - - for (i = 0; i < count; i++) { - if (to_domain) { - segs[i].dest.foreign.ref = ioreq->req.seg[i].gref; - segs[i].dest.foreign.offset = ioreq->req.seg[i].first_sect * file_blk; - segs[i].source.virt = virt; - } else { - segs[i].source.foreign.ref = ioreq->req.seg[i].gref; - segs[i].source.foreign.offset = ioreq->req.seg[i].first_sect * file_blk; - segs[i].dest.virt = virt; - } - segs[i].len = (ioreq->req.seg[i].last_sect - - ioreq->req.seg[i].first_sect + 1) * file_blk; - virt += segs[i].len; - } - - rc = xen_be_copy_grant_refs(xendev, to_domain, segs, count); - - if (rc) { - xen_pv_printf(xendev, 0, - "failed to copy data %d\n", rc); - ioreq->aio_errors++; - return -1; - } - - return rc; -} - -static int ioreq_runio_qemu_aio(struct ioreq *ioreq); - -static void qemu_aio_complete(void *opaque, int ret) -{ - struct ioreq *ioreq = opaque; - struct XenBlkDev *blkdev = ioreq->blkdev; - struct XenDevice *xendev = &blkdev->xendev; - - aio_context_acquire(blkdev->ctx); - - if (ret != 0) { - xen_pv_printf(xendev, 0, "%s I/O error\n", - ioreq->req.operation == BLKIF_OP_READ ? "read" : "write"); - ioreq->aio_errors++; - } - - ioreq->aio_inflight--; - if (ioreq->presync) { - ioreq->presync = 0; - ioreq_runio_qemu_aio(ioreq); - goto done; - } - if (ioreq->aio_inflight > 0) { - goto done; - } - - switch (ioreq->req.operation) { - case BLKIF_OP_READ: - /* in case of failure ioreq->aio_errors is increased */ - if (ret == 0) { - ioreq_grant_copy(ioreq); - } - qemu_vfree(ioreq->buf); - break; - case BLKIF_OP_WRITE: - case BLKIF_OP_FLUSH_DISKCACHE: - if (!ioreq->req.nr_segments) { - break; - } - qemu_vfree(ioreq->buf); - break; - default: - break; - } - - ioreq->status = ioreq->aio_errors ? BLKIF_RSP_ERROR : BLKIF_RSP_OKAY; - ioreq_finish(ioreq); - - switch (ioreq->req.operation) { - case BLKIF_OP_WRITE: - case BLKIF_OP_FLUSH_DISKCACHE: - if (!ioreq->req.nr_segments) { - break; - } - case BLKIF_OP_READ: - if (ioreq->status == BLKIF_RSP_OKAY) { - block_acct_done(blk_get_stats(blkdev->blk), &ioreq->acct); - } else { - block_acct_failed(blk_get_stats(blkdev->blk), &ioreq->acct); - } - break; - case BLKIF_OP_DISCARD: - default: - break; - } - qemu_bh_schedule(blkdev->bh); - -done: - aio_context_release(blkdev->ctx); -} - -static bool blk_split_discard(struct ioreq *ioreq, blkif_sector_t sector_number, - uint64_t nr_sectors) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - int64_t byte_offset; - int byte_chunk; - uint64_t byte_remaining, limit; - uint64_t sec_start = sector_number; - uint64_t sec_count = nr_sectors; - - /* Wrap around, or overflowing byte limit? */ - if (sec_start + sec_count < sec_count || - sec_start + sec_count > INT64_MAX >> BDRV_SECTOR_BITS) { - return false; - } - - limit = BDRV_REQUEST_MAX_SECTORS << BDRV_SECTOR_BITS; - byte_offset = sec_start << BDRV_SECTOR_BITS; - byte_remaining = sec_count << BDRV_SECTOR_BITS; - - do { - byte_chunk = byte_remaining > limit ? limit : byte_remaining; - ioreq->aio_inflight++; - blk_aio_pdiscard(blkdev->blk, byte_offset, byte_chunk, - qemu_aio_complete, ioreq); - byte_remaining -= byte_chunk; - byte_offset += byte_chunk; - } while (byte_remaining > 0); - - return true; -} - -static int ioreq_runio_qemu_aio(struct ioreq *ioreq) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - - ioreq->buf = qemu_memalign(XC_PAGE_SIZE, ioreq->size); - if (ioreq->req.nr_segments && - (ioreq->req.operation == BLKIF_OP_WRITE || - ioreq->req.operation == BLKIF_OP_FLUSH_DISKCACHE) && - ioreq_grant_copy(ioreq)) { - qemu_vfree(ioreq->buf); - goto err; - } - - ioreq->aio_inflight++; - if (ioreq->presync) { - blk_aio_flush(ioreq->blkdev->blk, qemu_aio_complete, ioreq); - return 0; - } - - switch (ioreq->req.operation) { - case BLKIF_OP_READ: - qemu_iovec_add(&ioreq->v, ioreq->buf, ioreq->size); - block_acct_start(blk_get_stats(blkdev->blk), &ioreq->acct, - ioreq->v.size, BLOCK_ACCT_READ); - ioreq->aio_inflight++; - blk_aio_preadv(blkdev->blk, ioreq->start, &ioreq->v, 0, - qemu_aio_complete, ioreq); - break; - case BLKIF_OP_WRITE: - case BLKIF_OP_FLUSH_DISKCACHE: - if (!ioreq->req.nr_segments) { - break; - } - - qemu_iovec_add(&ioreq->v, ioreq->buf, ioreq->size); - block_acct_start(blk_get_stats(blkdev->blk), &ioreq->acct, - ioreq->v.size, - ioreq->req.operation == BLKIF_OP_WRITE ? - BLOCK_ACCT_WRITE : BLOCK_ACCT_FLUSH); - ioreq->aio_inflight++; - blk_aio_pwritev(blkdev->blk, ioreq->start, &ioreq->v, 0, - qemu_aio_complete, ioreq); - break; - case BLKIF_OP_DISCARD: - { - struct blkif_request_discard *req = (void *)&ioreq->req; - if (!blk_split_discard(ioreq, req->sector_number, req->nr_sectors)) { - goto err; - } - break; - } - default: - /* unknown operation (shouldn't happen -- parse catches this) */ - goto err; - } - - qemu_aio_complete(ioreq, 0); - - return 0; - -err: - ioreq_finish(ioreq); - ioreq->status = BLKIF_RSP_ERROR; - return -1; -} - -static int blk_send_response_one(struct ioreq *ioreq) -{ - struct XenBlkDev *blkdev = ioreq->blkdev; - int send_notify = 0; - int have_requests = 0; - blkif_response_t *resp; - - /* Place on the response ring for the relevant domain. */ - switch (blkdev->protocol) { - case BLKIF_PROTOCOL_NATIVE: - resp = (blkif_response_t *) RING_GET_RESPONSE(&blkdev->rings.native, - blkdev->rings.native.rsp_prod_pvt); - break; - case BLKIF_PROTOCOL_X86_32: - resp = (blkif_response_t *) RING_GET_RESPONSE(&blkdev->rings.x86_32_part, - blkdev->rings.x86_32_part.rsp_prod_pvt); - break; - case BLKIF_PROTOCOL_X86_64: - resp = (blkif_response_t *) RING_GET_RESPONSE(&blkdev->rings.x86_64_part, - blkdev->rings.x86_64_part.rsp_prod_pvt); - break; - default: - return 0; - } - - resp->id = ioreq->req.id; - resp->operation = ioreq->req.operation; - resp->status = ioreq->status; - - blkdev->rings.common.rsp_prod_pvt++; - - RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blkdev->rings.common, send_notify); - if (blkdev->rings.common.rsp_prod_pvt == blkdev->rings.common.req_cons) { - /* - * Tail check for pending requests. Allows frontend to avoid - * notifications if requests are already in flight (lower - * overheads and promotes batching). - */ - RING_FINAL_CHECK_FOR_REQUESTS(&blkdev->rings.common, have_requests); - } else if (RING_HAS_UNCONSUMED_REQUESTS(&blkdev->rings.common)) { - have_requests = 1; - } - - if (have_requests) { - blkdev->more_work++; - } - return send_notify; -} - -/* walk finished list, send outstanding responses, free requests */ -static void blk_send_response_all(struct XenBlkDev *blkdev) -{ - struct ioreq *ioreq; - int send_notify = 0; - - while (!QLIST_EMPTY(&blkdev->finished)) { - ioreq = QLIST_FIRST(&blkdev->finished); - send_notify += blk_send_response_one(ioreq); - ioreq_release(ioreq, true); - } - if (send_notify) { - xen_pv_send_notify(&blkdev->xendev); - } -} - -static int blk_get_request(struct XenBlkDev *blkdev, struct ioreq *ioreq, RING_IDX rc) -{ - switch (blkdev->protocol) { - case BLKIF_PROTOCOL_NATIVE: - memcpy(&ioreq->req, RING_GET_REQUEST(&blkdev->rings.native, rc), - sizeof(ioreq->req)); - break; - case BLKIF_PROTOCOL_X86_32: - blkif_get_x86_32_req(&ioreq->req, - RING_GET_REQUEST(&blkdev->rings.x86_32_part, rc)); - break; - case BLKIF_PROTOCOL_X86_64: - blkif_get_x86_64_req(&ioreq->req, - RING_GET_REQUEST(&blkdev->rings.x86_64_part, rc)); - break; - } - /* Prevent the compiler from accessing the on-ring fields instead. */ - barrier(); - return 0; -} - -static void blk_handle_requests(struct XenBlkDev *blkdev) -{ - RING_IDX rc, rp; - struct ioreq *ioreq; - - blkdev->more_work = 0; - - rc = blkdev->rings.common.req_cons; - rp = blkdev->rings.common.sring->req_prod; - xen_rmb(); /* Ensure we see queued requests up to 'rp'. */ - - blk_send_response_all(blkdev); - while (rc != rp) { - /* pull request from ring */ - if (RING_REQUEST_CONS_OVERFLOW(&blkdev->rings.common, rc)) { - break; - } - ioreq = ioreq_start(blkdev); - if (ioreq == NULL) { - blkdev->more_work++; - break; - } - blk_get_request(blkdev, ioreq, rc); - blkdev->rings.common.req_cons = ++rc; - - /* parse them */ - if (ioreq_parse(ioreq) != 0) { - - switch (ioreq->req.operation) { - case BLKIF_OP_READ: - block_acct_invalid(blk_get_stats(blkdev->blk), - BLOCK_ACCT_READ); - break; - case BLKIF_OP_WRITE: - block_acct_invalid(blk_get_stats(blkdev->blk), - BLOCK_ACCT_WRITE); - break; - case BLKIF_OP_FLUSH_DISKCACHE: - block_acct_invalid(blk_get_stats(blkdev->blk), - BLOCK_ACCT_FLUSH); - default: - break; - }; - - if (blk_send_response_one(ioreq)) { - xen_pv_send_notify(&blkdev->xendev); - } - ioreq_release(ioreq, false); - continue; - } - - ioreq_runio_qemu_aio(ioreq); - } - - if (blkdev->more_work && blkdev->requests_inflight < blkdev->max_requests) { - qemu_bh_schedule(blkdev->bh); - } -} - -/* ------------------------------------------------------------- */ - -static void blk_bh(void *opaque) -{ - struct XenBlkDev *blkdev = opaque; - - aio_context_acquire(blkdev->ctx); - blk_handle_requests(blkdev); - aio_context_release(blkdev->ctx); -} - -static void blk_alloc(struct XenDevice *xendev) -{ - struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev); - Error *err = NULL; - - trace_xen_disk_alloc(xendev->name); - - QLIST_INIT(&blkdev->inflight); - QLIST_INIT(&blkdev->finished); - QLIST_INIT(&blkdev->freelist); - - blkdev->iothread = iothread_create(xendev->name, &err); - assert(!err); - - blkdev->ctx = iothread_get_aio_context(blkdev->iothread); - blkdev->bh = aio_bh_new(blkdev->ctx, blk_bh, blkdev); -} - -static void blk_parse_discard(struct XenBlkDev *blkdev) -{ - struct XenDevice *xendev = &blkdev->xendev; - int enable; - - blkdev->feature_discard = true; - - if (xenstore_read_be_int(xendev, "discard-enable", &enable) == 0) { - blkdev->feature_discard = !!enable; - } - - if (blkdev->feature_discard) { - xenstore_write_be_int(xendev, "feature-discard", 1); - } -} - -static int blk_init(struct XenDevice *xendev) -{ - struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev); - int info = 0; - char *directiosafe = NULL; - - trace_xen_disk_init(xendev->name); - - /* read xenstore entries */ - if (blkdev->params == NULL) { - char *h = NULL; - blkdev->params = xenstore_read_be_str(xendev, "params"); - if (blkdev->params != NULL) { - h = strchr(blkdev->params, ':'); - } - if (h != NULL) { - blkdev->fileproto = blkdev->params; - blkdev->filename = h+1; - *h = 0; - } else { - blkdev->fileproto = ""; - blkdev->filename = blkdev->params; - } - } - if (!strcmp("aio", blkdev->fileproto)) { - blkdev->fileproto = "raw"; - } - if (!strcmp("vhd", blkdev->fileproto)) { - blkdev->fileproto = "vpc"; - } - if (blkdev->mode == NULL) { - blkdev->mode = xenstore_read_be_str(xendev, "mode"); - } - if (blkdev->type == NULL) { - blkdev->type = xenstore_read_be_str(xendev, "type"); - } - if (blkdev->dev == NULL) { - blkdev->dev = xenstore_read_be_str(xendev, "dev"); - } - if (blkdev->devtype == NULL) { - blkdev->devtype = xenstore_read_be_str(xendev, "device-type"); - } - directiosafe = xenstore_read_be_str(xendev, "direct-io-safe"); - blkdev->directiosafe = (directiosafe && atoi(directiosafe)); - - /* do we have all we need? */ - if (blkdev->params == NULL || - blkdev->mode == NULL || - blkdev->type == NULL || - blkdev->dev == NULL) { - goto out_error; - } - - /* read-only ? */ - if (strcmp(blkdev->mode, "w")) { - info |= VDISK_READONLY; - } - - /* cdrom ? */ - if (blkdev->devtype && !strcmp(blkdev->devtype, "cdrom")) { - info |= VDISK_CDROM; - } - - blkdev->file_blk = BLOCK_SIZE; - - /* fill info - * blk_connect supplies sector-size and sectors - */ - xenstore_write_be_int(xendev, "feature-flush-cache", 1); - xenstore_write_be_int(xendev, "info", info); - - xenstore_write_be_int(xendev, "max-ring-page-order", - MAX_RING_PAGE_ORDER); - - blk_parse_discard(blkdev); - - g_free(directiosafe); - return 0; - -out_error: - g_free(blkdev->params); - blkdev->params = NULL; - g_free(blkdev->mode); - blkdev->mode = NULL; - g_free(blkdev->type); - blkdev->type = NULL; - g_free(blkdev->dev); - blkdev->dev = NULL; - g_free(blkdev->devtype); - blkdev->devtype = NULL; - g_free(directiosafe); - blkdev->directiosafe = false; - return -1; -} - -static int blk_connect(struct XenDevice *xendev) -{ - struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev); - int index, qflags; - bool readonly = true; - bool writethrough = true; - int order, ring_ref; - unsigned int ring_size, max_grants; - unsigned int i; - - trace_xen_disk_connect(xendev->name); - - /* read-only ? */ - if (blkdev->directiosafe) { - qflags = BDRV_O_NOCACHE | BDRV_O_NATIVE_AIO; - } else { - qflags = 0; - writethrough = false; - } - if (strcmp(blkdev->mode, "w") == 0) { - qflags |= BDRV_O_RDWR; - readonly = false; - } - if (blkdev->feature_discard) { - qflags |= BDRV_O_UNMAP; - } - - /* init qemu block driver */ - index = (xendev->dev - 202 * 256) / 16; - blkdev->dinfo = drive_get(IF_XEN, 0, index); - if (!blkdev->dinfo) { - Error *local_err = NULL; - QDict *options = NULL; - - if (strcmp(blkdev->fileproto, "")) { - options = qdict_new(); - qdict_put_str(options, "driver", blkdev->fileproto); - } - - /* setup via xenbus -> create new block driver instance */ - xen_pv_printf(xendev, 2, "create new bdrv (xenbus setup)\n"); - blkdev->blk = blk_new_open(blkdev->filename, NULL, options, - qflags, &local_err); - if (!blkdev->blk) { - xen_pv_printf(xendev, 0, "error: %s\n", - error_get_pretty(local_err)); - error_free(local_err); - return -1; - } - blk_set_enable_write_cache(blkdev->blk, !writethrough); - } else { - /* setup via qemu cmdline -> already setup for us */ - xen_pv_printf(xendev, 2, - "get configured bdrv (cmdline setup)\n"); - blkdev->blk = blk_by_legacy_dinfo(blkdev->dinfo); - if (blk_is_read_only(blkdev->blk) && !readonly) { - xen_pv_printf(xendev, 0, "Unexpected read-only drive"); - blkdev->blk = NULL; - return -1; - } - /* blkdev->blk is not create by us, we get a reference - * so we can blk_unref() unconditionally */ - blk_ref(blkdev->blk); - } - blk_attach_dev_legacy(blkdev->blk, blkdev); - blkdev->file_size = blk_getlength(blkdev->blk); - if (blkdev->file_size < 0) { - BlockDriverState *bs = blk_bs(blkdev->blk); - const char *drv_name = bs ? bdrv_get_format_name(bs) : NULL; - xen_pv_printf(xendev, 1, "blk_getlength: %d (%s) | drv %s\n", - (int)blkdev->file_size, strerror(-blkdev->file_size), - drv_name ?: "-"); - blkdev->file_size = 0; - } - - xen_pv_printf(xendev, 1, "type \"%s\", fileproto \"%s\", filename \"%s\"," - " size %" PRId64 " (%" PRId64 " MB)\n", - blkdev->type, blkdev->fileproto, blkdev->filename, - blkdev->file_size, blkdev->file_size / MiB); - - /* Fill in number of sector size and number of sectors */ - xenstore_write_be_int(xendev, "sector-size", blkdev->file_blk); - xenstore_write_be_int64(xendev, "sectors", - blkdev->file_size / blkdev->file_blk); - - if (xenstore_read_fe_int(xendev, "ring-page-order", - &order) == -1) { - blkdev->nr_ring_ref = 1; - - if (xenstore_read_fe_int(xendev, "ring-ref", - &ring_ref) == -1) { - return -1; - } - blkdev->ring_ref[0] = ring_ref; - - } else if (order >= 0 && order <= MAX_RING_PAGE_ORDER) { - blkdev->nr_ring_ref = 1 << order; - - for (i = 0; i < blkdev->nr_ring_ref; i++) { - char *key; - - key = g_strdup_printf("ring-ref%u", i); - if (!key) { - return -1; - } - - if (xenstore_read_fe_int(xendev, key, - &ring_ref) == -1) { - g_free(key); - return -1; - } - blkdev->ring_ref[i] = ring_ref; - - g_free(key); - } - } else { - xen_pv_printf(xendev, 0, "invalid ring-page-order: %d\n", - order); - return -1; - } - - if (xenstore_read_fe_int(xendev, "event-channel", - &xendev->remote_port) == -1) { - return -1; - } - - if (!xendev->protocol) { - blkdev->protocol = BLKIF_PROTOCOL_NATIVE; - } else if (strcmp(xendev->protocol, XEN_IO_PROTO_ABI_NATIVE) == 0) { - blkdev->protocol = BLKIF_PROTOCOL_NATIVE; - } else if (strcmp(xendev->protocol, XEN_IO_PROTO_ABI_X86_32) == 0) { - blkdev->protocol = BLKIF_PROTOCOL_X86_32; - } else if (strcmp(xendev->protocol, XEN_IO_PROTO_ABI_X86_64) == 0) { - blkdev->protocol = BLKIF_PROTOCOL_X86_64; - } else { - blkdev->protocol = BLKIF_PROTOCOL_NATIVE; - } - - ring_size = XC_PAGE_SIZE * blkdev->nr_ring_ref; - switch (blkdev->protocol) { - case BLKIF_PROTOCOL_NATIVE: - { - blkdev->max_requests = __CONST_RING_SIZE(blkif, ring_size); - break; - } - case BLKIF_PROTOCOL_X86_32: - { - blkdev->max_requests = __CONST_RING_SIZE(blkif_x86_32, ring_size); - break; - } - case BLKIF_PROTOCOL_X86_64: - { - blkdev->max_requests = __CONST_RING_SIZE(blkif_x86_64, ring_size); - break; - } - default: - return -1; - } - - /* Add on the number needed for the ring pages */ - max_grants = blkdev->nr_ring_ref; - - xen_be_set_max_grant_refs(xendev, max_grants); - blkdev->sring = xen_be_map_grant_refs(xendev, blkdev->ring_ref, - blkdev->nr_ring_ref, - PROT_READ | PROT_WRITE); - if (!blkdev->sring) { - return -1; - } - - switch (blkdev->protocol) { - case BLKIF_PROTOCOL_NATIVE: - { - blkif_sring_t *sring_native = blkdev->sring; - BACK_RING_INIT(&blkdev->rings.native, sring_native, ring_size); - break; - } - case BLKIF_PROTOCOL_X86_32: - { - blkif_x86_32_sring_t *sring_x86_32 = blkdev->sring; - - BACK_RING_INIT(&blkdev->rings.x86_32_part, sring_x86_32, ring_size); - break; - } - case BLKIF_PROTOCOL_X86_64: - { - blkif_x86_64_sring_t *sring_x86_64 = blkdev->sring; - - BACK_RING_INIT(&blkdev->rings.x86_64_part, sring_x86_64, ring_size); - break; - } - } - - blk_set_aio_context(blkdev->blk, blkdev->ctx); - - xen_be_bind_evtchn(xendev); - - xen_pv_printf(xendev, 1, "ok: proto %s, nr-ring-ref %u, " - "remote port %d, local port %d\n", - xendev->protocol, blkdev->nr_ring_ref, - xendev->remote_port, xendev->local_port); - return 0; -} - -static void blk_disconnect(struct XenDevice *xendev) -{ - struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev); - - trace_xen_disk_disconnect(xendev->name); - - aio_context_acquire(blkdev->ctx); - - if (blkdev->blk) { - blk_set_aio_context(blkdev->blk, qemu_get_aio_context()); - blk_detach_dev(blkdev->blk, blkdev); - blk_unref(blkdev->blk); - blkdev->blk = NULL; - } - xen_pv_unbind_evtchn(xendev); - - aio_context_release(blkdev->ctx); - - if (blkdev->sring) { - xen_be_unmap_grant_refs(xendev, blkdev->sring, - blkdev->nr_ring_ref); - blkdev->sring = NULL; - } -} - -static int blk_free(struct XenDevice *xendev) -{ - struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev); - struct ioreq *ioreq; - - trace_xen_disk_free(xendev->name); - - blk_disconnect(xendev); - - while (!QLIST_EMPTY(&blkdev->freelist)) { - ioreq = QLIST_FIRST(&blkdev->freelist); - QLIST_REMOVE(ioreq, list); - qemu_iovec_destroy(&ioreq->v); - g_free(ioreq); - } - - g_free(blkdev->params); - g_free(blkdev->mode); - g_free(blkdev->type); - g_free(blkdev->dev); - g_free(blkdev->devtype); - qemu_bh_delete(blkdev->bh); - iothread_destroy(blkdev->iothread); - return 0; -} - -static void blk_event(struct XenDevice *xendev) -{ - struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev); - - qemu_bh_schedule(blkdev->bh); -} - -struct XenDevOps xen_blkdev_ops = { - .flags = DEVOPS_FLAG_NEED_GNTDEV, - .size = sizeof(struct XenBlkDev), - .alloc = blk_alloc, - .init = blk_init, - .initialise = blk_connect, - .disconnect = blk_disconnect, - .event = blk_event, - .free = blk_free, -}; diff --git a/hw/char/xen_console.c b/hw/char/xen_console.c index b1a1e66d5a..dc6ff0e5b3 100644 --- a/hw/char/xen_console.c +++ b/hw/char/xen_console.c @@ -26,7 +26,7 @@ #include "qapi/error.h" #include "hw/hw.h" #include "chardev/char-fe.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include @@ -39,7 +39,7 @@ struct buffer { }; struct XenConsole { - struct XenDevice xendev; /* must be first */ + struct XenLegacyDevice xendev; /* must be first */ struct buffer buffer; char console[XEN_BUFSIZE]; int ring_ref; @@ -173,7 +173,7 @@ static void xencons_send(struct XenConsole *con) /* -------------------------------------------------------------------- */ -static int con_init(struct XenDevice *xendev) +static int con_init(struct XenLegacyDevice *xendev) { struct XenConsole *con = container_of(xendev, struct XenConsole, xendev); char *type, *dom, label[32]; @@ -222,7 +222,7 @@ out: return ret; } -static int con_initialise(struct XenDevice *xendev) +static int con_initialise(struct XenLegacyDevice *xendev) { struct XenConsole *con = container_of(xendev, struct XenConsole, xendev); int limit; @@ -259,7 +259,7 @@ static int con_initialise(struct XenDevice *xendev) return 0; } -static void con_disconnect(struct XenDevice *xendev) +static void con_disconnect(struct XenLegacyDevice *xendev) { struct XenConsole *con = container_of(xendev, struct XenConsole, xendev); @@ -276,7 +276,7 @@ static void con_disconnect(struct XenDevice *xendev) } } -static void con_event(struct XenDevice *xendev) +static void con_event(struct XenLegacyDevice *xendev) { struct XenConsole *con = container_of(xendev, struct XenConsole, xendev); diff --git a/hw/display/xenfb.c b/hw/display/xenfb.c index 0330dc6f61..6202f1150e 100644 --- a/hw/display/xenfb.c +++ b/hw/display/xenfb.c @@ -30,7 +30,7 @@ #include "hw/hw.h" #include "ui/input.h" #include "ui/console.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include #include @@ -46,7 +46,7 @@ /* -------------------------------------------------------------------- */ struct common { - struct XenDevice xendev; /* must be first */ + struct XenLegacyDevice xendev; /* must be first */ void *page; }; @@ -342,14 +342,14 @@ static QemuInputHandler xenfb_rel_mouse = { .sync = xenfb_mouse_sync, }; -static int input_init(struct XenDevice *xendev) +static int input_init(struct XenLegacyDevice *xendev) { xenstore_write_be_int(xendev, "feature-abs-pointer", 1); xenstore_write_be_int(xendev, "feature-raw-pointer", 1); return 0; } -static int input_initialise(struct XenDevice *xendev) +static int input_initialise(struct XenLegacyDevice *xendev) { struct XenInput *in = container_of(xendev, struct XenInput, c.xendev); int rc; @@ -361,7 +361,7 @@ static int input_initialise(struct XenDevice *xendev) return 0; } -static void input_connected(struct XenDevice *xendev) +static void input_connected(struct XenLegacyDevice *xendev) { struct XenInput *in = container_of(xendev, struct XenInput, c.xendev); @@ -395,7 +395,7 @@ static void input_connected(struct XenDevice *xendev) } } -static void input_disconnect(struct XenDevice *xendev) +static void input_disconnect(struct XenLegacyDevice *xendev) { struct XenInput *in = container_of(xendev, struct XenInput, c.xendev); @@ -410,7 +410,7 @@ static void input_disconnect(struct XenDevice *xendev) common_unbind(&in->c); } -static void input_event(struct XenDevice *xendev) +static void input_event(struct XenLegacyDevice *xendev) { struct XenInput *xenfb = container_of(xendev, struct XenInput, c.xendev); struct xenkbd_page *page = xenfb->c.page; @@ -867,7 +867,7 @@ static void xenfb_handle_events(struct XenFB *xenfb) page->out_cons = cons; } -static int fb_init(struct XenDevice *xendev) +static int fb_init(struct XenLegacyDevice *xendev) { #ifdef XENFB_TYPE_RESIZE xenstore_write_be_int(xendev, "feature-resize", 1); @@ -875,7 +875,7 @@ static int fb_init(struct XenDevice *xendev) return 0; } -static int fb_initialise(struct XenDevice *xendev) +static int fb_initialise(struct XenLegacyDevice *xendev) { struct XenFB *fb = container_of(xendev, struct XenFB, c.xendev); struct xenfb_page *fb_page; @@ -912,7 +912,7 @@ static int fb_initialise(struct XenDevice *xendev) return 0; } -static void fb_disconnect(struct XenDevice *xendev) +static void fb_disconnect(struct XenLegacyDevice *xendev) { struct XenFB *fb = container_of(xendev, struct XenFB, c.xendev); @@ -935,7 +935,8 @@ static void fb_disconnect(struct XenDevice *xendev) fb->bug_trigger = 0; } -static void fb_frontend_changed(struct XenDevice *xendev, const char *node) +static void fb_frontend_changed(struct XenLegacyDevice *xendev, + const char *node) { struct XenFB *fb = container_of(xendev, struct XenFB, c.xendev); @@ -953,7 +954,7 @@ static void fb_frontend_changed(struct XenDevice *xendev, const char *node) } } -static void fb_event(struct XenDevice *xendev) +static void fb_event(struct XenLegacyDevice *xendev) { struct XenFB *xenfb = container_of(xendev, struct XenFB, c.xendev); diff --git a/hw/i386/xen/xen-hvm.c b/hw/i386/xen/xen-hvm.c index 8c8562f359..2939122e7c 100644 --- a/hw/i386/xen/xen-hvm.c +++ b/hw/i386/xen/xen-hvm.c @@ -16,7 +16,8 @@ #include "hw/i386/pc.h" #include "hw/i386/apic-msidef.h" #include "hw/xen/xen_common.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" +#include "hw/xen/xen-bus.h" #include "qapi/error.h" #include "qapi/qapi-commands-misc.h" #include "qemu/error-report.h" @@ -1484,6 +1485,8 @@ void xen_hvm_init(PCMachineState *pcms, MemoryRegion **ram_memory) QLIST_INIT(&state->dev_list); device_listener_register(&state->device_listener); + xen_bus_init(); + /* Initialize backend core & drivers */ if (xen_be_init() != 0) { error_report("xen backend core setup failed"); diff --git a/hw/i386/xen/xen-mapcache.c b/hw/i386/xen/xen-mapcache.c index 02e823c5a2..349f72d00c 100644 --- a/hw/i386/xen/xen-mapcache.c +++ b/hw/i386/xen/xen-mapcache.c @@ -14,7 +14,7 @@ #include -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "qemu/bitmap.h" #include diff --git a/hw/i386/xen/xen_platform.c b/hw/i386/xen/xen_platform.c index deb7a0c374..16afb54fee 100644 --- a/hw/i386/xen/xen_platform.c +++ b/hw/i386/xen/xen_platform.c @@ -30,7 +30,7 @@ #include "hw/pci/pci.h" #include "hw/irq.h" #include "hw/xen/xen_common.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "trace.h" #include "exec/address-spaces.h" #include "sysemu/block-backend.h" diff --git a/hw/net/xen_nic.c b/hw/net/xen_nic.c index 46a8dbfc90..37cda8e4be 100644 --- a/hw/net/xen_nic.c +++ b/hw/net/xen_nic.c @@ -28,14 +28,14 @@ #include "net/net.h" #include "net/checksum.h" #include "net/util.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include /* ------------------------------------------------------------- */ struct XenNetDev { - struct XenDevice xendev; /* must be first */ + struct XenLegacyDevice xendev; /* must be first */ char *mac; int tx_work; int tx_ring_ref; @@ -276,7 +276,7 @@ static NetClientInfo net_xen_info = { .receive = net_rx_packet, }; -static int net_init(struct XenDevice *xendev) +static int net_init(struct XenLegacyDevice *xendev) { struct XenNetDev *netdev = container_of(xendev, struct XenNetDev, xendev); @@ -308,7 +308,7 @@ static int net_init(struct XenDevice *xendev) return 0; } -static int net_connect(struct XenDevice *xendev) +static int net_connect(struct XenLegacyDevice *xendev) { struct XenNetDev *netdev = container_of(xendev, struct XenNetDev, xendev); int rx_copy; @@ -363,7 +363,7 @@ static int net_connect(struct XenDevice *xendev) return 0; } -static void net_disconnect(struct XenDevice *xendev) +static void net_disconnect(struct XenLegacyDevice *xendev) { struct XenNetDev *netdev = container_of(xendev, struct XenNetDev, xendev); @@ -379,14 +379,14 @@ static void net_disconnect(struct XenDevice *xendev) } } -static void net_event(struct XenDevice *xendev) +static void net_event(struct XenLegacyDevice *xendev) { struct XenNetDev *netdev = container_of(xendev, struct XenNetDev, xendev); net_tx_packets(netdev); qemu_flush_queued_packets(qemu_get_queue(netdev->nic)); } -static int net_free(struct XenDevice *xendev) +static int net_free(struct XenLegacyDevice *xendev) { struct XenNetDev *netdev = container_of(xendev, struct XenNetDev, xendev); diff --git a/hw/usb/xen-usb.c b/hw/usb/xen-usb.c index 5758a105a2..b20d0cfadf 100644 --- a/hw/usb/xen-usb.c +++ b/hw/usb/xen-usb.c @@ -27,7 +27,7 @@ #include "qemu/option.h" #include "hw/sysbus.h" #include "hw/usb.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "monitor/qdev.h" #include "qapi/qmp/qdict.h" #include "qapi/qmp/qstring.h" @@ -99,7 +99,7 @@ struct usbback_hotplug { }; struct usbback_info { - struct XenDevice xendev; /* must be first */ + struct XenLegacyDevice xendev; /* must be first */ USBBus bus; void *urb_sring; void *conn_sring; @@ -142,7 +142,7 @@ static int usbback_gnttab_map(struct usbback_req *usbback_req) unsigned int nr_segs, i, prot; uint32_t ref[USBIF_MAX_SEGMENTS_PER_REQUEST]; struct usbback_info *usbif = usbback_req->usbif; - struct XenDevice *xendev = &usbif->xendev; + struct XenLegacyDevice *xendev = &usbif->xendev; struct usbif_request_segment *seg; void *addr; @@ -220,7 +220,7 @@ static int usbback_gnttab_map(struct usbback_req *usbback_req) static int usbback_init_packet(struct usbback_req *usbback_req) { - struct XenDevice *xendev = &usbback_req->usbif->xendev; + struct XenLegacyDevice *xendev = &usbback_req->usbif->xendev; USBPacket *packet = &usbback_req->packet; USBDevice *dev = usbback_req->stub->dev; USBEndpoint *ep; @@ -279,7 +279,7 @@ static void usbback_do_response(struct usbback_req *usbback_req, int32_t status, { struct usbback_info *usbif; struct usbif_urb_response *res; - struct XenDevice *xendev; + struct XenLegacyDevice *xendev; unsigned int notify; usbif = usbback_req->usbif; @@ -824,7 +824,7 @@ static void usbback_process_port(struct usbback_info *usbif, unsigned port) g_free(busid); } -static void usbback_disconnect(struct XenDevice *xendev) +static void usbback_disconnect(struct XenLegacyDevice *xendev) { struct usbback_info *usbif; unsigned int i; @@ -853,7 +853,7 @@ static void usbback_disconnect(struct XenDevice *xendev) TR_BUS(xendev, "finished\n"); } -static int usbback_connect(struct XenDevice *xendev) +static int usbback_connect(struct XenLegacyDevice *xendev) { struct usbback_info *usbif; struct usbif_urb_sring *urb_sring; @@ -913,7 +913,8 @@ static int usbback_connect(struct XenDevice *xendev) return 0; } -static void usbback_backend_changed(struct XenDevice *xendev, const char *node) +static void usbback_backend_changed(struct XenLegacyDevice *xendev, + const char *node) { struct usbback_info *usbif; unsigned int i; @@ -926,7 +927,7 @@ static void usbback_backend_changed(struct XenDevice *xendev, const char *node) } } -static int usbback_init(struct XenDevice *xendev) +static int usbback_init(struct XenLegacyDevice *xendev) { struct usbback_info *usbif; @@ -1005,7 +1006,7 @@ static USBPortOps xen_usb_port_ops = { static USBBusOps xen_usb_bus_ops = { }; -static void usbback_alloc(struct XenDevice *xendev) +static void usbback_alloc(struct XenLegacyDevice *xendev) { struct usbback_info *usbif; USBPort *p; @@ -1027,7 +1028,7 @@ static void usbback_alloc(struct XenDevice *xendev) usbif->bh = qemu_bh_new(usbback_bh, usbif); } -static int usbback_free(struct XenDevice *xendev) +static int usbback_free(struct XenLegacyDevice *xendev) { struct usbback_info *usbif; struct usbback_req *usbback_req; @@ -1066,7 +1067,7 @@ static int usbback_free(struct XenDevice *xendev) return 0; } -static void usbback_event(struct XenDevice *xendev) +static void usbback_event(struct XenLegacyDevice *xendev) { struct usbback_info *usbif; diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs index 9ea5c73423..84df60a928 100644 --- a/hw/xen/Makefile.objs +++ b/hw/xen/Makefile.objs @@ -1,5 +1,5 @@ # xen backend driver support -common-obj-$(CONFIG_XEN) += xen_backend.o xen_devconfig.o xen_pvdev.o xen-common.o +common-obj-$(CONFIG_XEN) += xen-legacy-backend.o xen_devconfig.o xen_pvdev.o xen-common.o xen-bus.o xen-bus-helper.o xen-backend.o obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_graphics.o xen_pt_msi.o diff --git a/hw/xen/trace-events b/hw/xen/trace-events index c7e7a3b523..f6944624b2 100644 --- a/hw/xen/trace-events +++ b/hw/xen/trace-events @@ -12,3 +12,29 @@ xen_unmap_portio_range(uint32_t id, uint64_t start_addr, uint64_t end_addr) "id: xen_map_pcidev(uint32_t id, uint8_t bus, uint8_t dev, uint8_t func) "id: %u bdf: %02x.%02x.%02x" xen_unmap_pcidev(uint32_t id, uint8_t bus, uint8_t dev, uint8_t func) "id: %u bdf: %02x.%02x.%02x" xen_domid_restrict(int err) "err: %u" + +# include/hw/xen/xen-bus.c +xen_bus_realize(void) "" +xen_bus_unrealize(void) "" +xen_bus_enumerate(void) "" +xen_bus_type_enumerate(const char *type) "type: %s" +xen_bus_backend_create(const char *type, const char *path) "type: %s path: %s" +xen_bus_add_watch(const char *node, const char *key, char *token) "node: %s key: %s token: %s" +xen_bus_remove_watch(const char *node, const char *key, char *token) "node: %s key: %s token: %s" +xen_bus_watch(const char *token) "token: %s" +xen_device_realize(const char *type, char *name) "type: %s name: %s" +xen_device_unrealize(const char *type, char *name) "type: %s name: %s" +xen_device_backend_state(const char *type, char *name, const char *state) "type: %s name: %s -> %s" +xen_device_backend_online(const char *type, char *name, bool online) "type: %s name: %s -> %u" +xen_device_backend_changed(const char *type, char *name) "type: %s name: %s" +xen_device_frontend_state(const char *type, char *name, const char *state) "type: %s name: %s -> %s" +xen_device_frontend_changed(const char *type, char *name) "type: %s name: %s" +xen_device_unplug(const char *type, char *name) "type: %s name: %s" + +# include/hw/xen/xen-bus-helper.c +xs_node_create(const char *node) "%s" +xs_node_destroy(const char *node) "%s" +xs_node_vprintf(char *path, char *value) "%s %s" +xs_node_vscanf(char *path, char *value) "%s %s" +xs_node_watch(char *path) "%s" +xs_node_unwatch(char *path) "%s" diff --git a/hw/xen/xen-backend.c b/hw/xen/xen-backend.c new file mode 100644 index 0000000000..da065f81b7 --- /dev/null +++ b/hw/xen/xen-backend.c @@ -0,0 +1,165 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" +#include "qemu/error-report.h" +#include "qapi/error.h" +#include "hw/xen/xen-backend.h" +#include "hw/xen/xen-bus.h" + +typedef struct XenBackendImpl { + const char *type; + XenBackendDeviceCreate create; + XenBackendDeviceDestroy destroy; +} XenBackendImpl; + +struct XenBackendInstance { + QLIST_ENTRY(XenBackendInstance) entry; + const XenBackendImpl *impl; + XenBus *xenbus; + char *name; + XenDevice *xendev; +}; + +static GHashTable *xen_backend_table_get(void) +{ + static GHashTable *table; + + if (table == NULL) { + table = g_hash_table_new(g_str_hash, g_str_equal); + } + + return table; +} + +static void xen_backend_table_add(XenBackendImpl *impl) +{ + g_hash_table_insert(xen_backend_table_get(), (void *)impl->type, impl); +} + +static const XenBackendImpl *xen_backend_table_lookup(const char *type) +{ + return g_hash_table_lookup(xen_backend_table_get(), type); +} + +void xen_backend_register(const XenBackendInfo *info) +{ + XenBackendImpl *impl = g_new0(XenBackendImpl, 1); + + g_assert(info->type); + + if (xen_backend_table_lookup(info->type)) { + error_report("attempt to register duplicate Xen backend type '%s'", + info->type); + abort(); + } + + if (!info->create) { + error_report("backend type '%s' has no creator", info->type); + abort(); + } + + impl->type = info->type; + impl->create = info->create; + impl->destroy = info->destroy; + + xen_backend_table_add(impl); +} + +static QLIST_HEAD(, XenBackendInstance) backend_list; + +static void xen_backend_list_add(XenBackendInstance *backend) +{ + QLIST_INSERT_HEAD(&backend_list, backend, entry); +} + +static XenBackendInstance *xen_backend_list_find(XenDevice *xendev) +{ + XenBackendInstance *backend; + + QLIST_FOREACH(backend, &backend_list, entry) { + if (backend->xendev == xendev) { + return backend; + } + } + + return NULL; +} + +static void xen_backend_list_remove(XenBackendInstance *backend) +{ + QLIST_REMOVE(backend, entry); +} + +void xen_backend_device_create(XenBus *xenbus, const char *type, + const char *name, QDict *opts, Error **errp) +{ + const XenBackendImpl *impl = xen_backend_table_lookup(type); + XenBackendInstance *backend; + Error *local_error = NULL; + + if (!impl) { + return; + } + + backend = g_new0(XenBackendInstance, 1); + backend->xenbus = xenbus; + backend->name = g_strdup(name); + + impl->create(backend, opts, &local_error); + if (local_error) { + error_propagate(errp, local_error); + g_free(backend->name); + g_free(backend); + return; + } + + backend->impl = impl; + xen_backend_list_add(backend); +} + +XenBus *xen_backend_get_bus(XenBackendInstance *backend) +{ + return backend->xenbus; +} + +const char *xen_backend_get_name(XenBackendInstance *backend) +{ + return backend->name; +} + +void xen_backend_set_device(XenBackendInstance *backend, + XenDevice *xendev) +{ + g_assert(!backend->xendev); + backend->xendev = xendev; +} + +XenDevice *xen_backend_get_device(XenBackendInstance *backend) +{ + return backend->xendev; +} + + +bool xen_backend_try_device_destroy(XenDevice *xendev, Error **errp) +{ + XenBackendInstance *backend = xen_backend_list_find(xendev); + const XenBackendImpl *impl; + + if (!backend) { + return false; + } + + impl = backend->impl; + impl->destroy(backend, errp); + + xen_backend_list_remove(backend); + g_free(backend->name); + g_free(backend); + + return true; +} diff --git a/hw/xen/xen-bus-helper.c b/hw/xen/xen-bus-helper.c new file mode 100644 index 0000000000..5f7a4b2612 --- /dev/null +++ b/hw/xen/xen-bus-helper.c @@ -0,0 +1,184 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" +#include "hw/hw.h" +#include "hw/sysbus.h" +#include "hw/xen/xen.h" +#include "hw/xen/xen-bus.h" +#include "hw/xen/xen-bus-helper.h" +#include "qapi/error.h" + +#include + +struct xs_state { + enum xenbus_state statenum; + const char *statestr; +}; +#define XS_STATE(state) { state, #state } + +static struct xs_state xs_state[] = { + XS_STATE(XenbusStateUnknown), + XS_STATE(XenbusStateInitialising), + XS_STATE(XenbusStateInitWait), + XS_STATE(XenbusStateInitialised), + XS_STATE(XenbusStateConnected), + XS_STATE(XenbusStateClosing), + XS_STATE(XenbusStateClosed), + XS_STATE(XenbusStateReconfiguring), + XS_STATE(XenbusStateReconfigured), +}; + +#undef XS_STATE + +const char *xs_strstate(enum xenbus_state state) +{ + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(xs_state); i++) { + if (xs_state[i].statenum == state) { + return xs_state[i].statestr; + } + } + + return "INVALID"; +} + +void xs_node_create(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, struct xs_permissions perms[], + unsigned int nr_perms, Error **errp) +{ + trace_xs_node_create(node); + + if (!xs_write(xsh, tid, node, "", 0)) { + error_setg_errno(errp, errno, "failed to create node '%s'", node); + return; + } + + if (!xs_set_permissions(xsh, tid, node, perms, nr_perms)) { + error_setg_errno(errp, errno, "failed to set node '%s' permissions", + node); + } +} + +void xs_node_destroy(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, Error **errp) +{ + trace_xs_node_destroy(node); + + if (!xs_rm(xsh, tid, node)) { + error_setg_errno(errp, errno, "failed to destroy node '%s'", node); + } +} + +void xs_node_vprintf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, va_list ap) +{ + char *path, *value; + int len; + + path = (strlen(node) != 0) ? g_strdup_printf("%s/%s", node, key) : + g_strdup(key); + len = g_vasprintf(&value, fmt, ap); + + trace_xs_node_vprintf(path, value); + + if (!xs_write(xsh, tid, path, value, len)) { + error_setg_errno(errp, errno, "failed to write '%s' to '%s'", + value, path); + } + + g_free(value); + g_free(path); +} + +void xs_node_printf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + xs_node_vprintf(xsh, tid, node, key, errp, fmt, ap); + va_end(ap); +} + +int xs_node_vscanf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, va_list ap) +{ + char *path, *value; + int rc; + + path = (strlen(node) != 0) ? g_strdup_printf("%s/%s", node, key) : + g_strdup(key); + value = xs_read(xsh, tid, path, NULL); + + trace_xs_node_vscanf(path, value); + + if (value) { + rc = vsscanf(value, fmt, ap); + } else { + error_setg_errno(errp, errno, "failed to read from '%s'", + path); + rc = EOF; + } + + free(value); + g_free(path); + + return rc; +} + +int xs_node_scanf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, ...) +{ + va_list ap; + int rc; + + va_start(ap, fmt); + rc = xs_node_vscanf(xsh, tid, node, key, errp, fmt, ap); + va_end(ap); + + return rc; +} + +void xs_node_watch(struct xs_handle *xsh, const char *node, const char *key, + char *token, Error **errp) +{ + char *path; + + path = (strlen(node) != 0) ? g_strdup_printf("%s/%s", node, key) : + g_strdup(key); + + trace_xs_node_watch(path); + + if (!xs_watch(xsh, path, token)) { + error_setg_errno(errp, errno, "failed to watch node '%s'", path); + } + + g_free(path); +} + +void xs_node_unwatch(struct xs_handle *xsh, const char *node, + const char *key, const char *token, Error **errp) +{ + char *path; + + path = (strlen(node) != 0) ? g_strdup_printf("%s/%s", node, key) : + g_strdup(key); + + trace_xs_node_unwatch(path); + + if (!xs_unwatch(xsh, path, token)) { + error_setg_errno(errp, errno, "failed to unwatch node '%s'", path); + } + + g_free(path); +} diff --git a/hw/xen/xen-bus.c b/hw/xen/xen-bus.c new file mode 100644 index 0000000000..3aeccec69c --- /dev/null +++ b/hw/xen/xen-bus.c @@ -0,0 +1,1199 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" +#include "qemu/main-loop.h" +#include "qemu/uuid.h" +#include "hw/hw.h" +#include "hw/sysbus.h" +#include "hw/xen/xen.h" +#include "hw/xen/xen-backend.h" +#include "hw/xen/xen-bus.h" +#include "hw/xen/xen-bus-helper.h" +#include "monitor/monitor.h" +#include "qapi/error.h" +#include "qapi/qmp/qdict.h" +#include "sysemu/sysemu.h" +#include "trace.h" + +static char *xen_device_get_backend_path(XenDevice *xendev) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + XenDeviceClass *xendev_class = XEN_DEVICE_GET_CLASS(xendev); + const char *type = object_get_typename(OBJECT(xendev)); + const char *backend = xendev_class->backend; + + if (!backend) { + backend = type; + } + + return g_strdup_printf("/local/domain/%u/backend/%s/%u/%s", + xenbus->backend_id, backend, xendev->frontend_id, + xendev->name); +} + +static char *xen_device_get_frontend_path(XenDevice *xendev) +{ + XenDeviceClass *xendev_class = XEN_DEVICE_GET_CLASS(xendev); + const char *type = object_get_typename(OBJECT(xendev)); + const char *device = xendev_class->device; + + if (!device) { + device = type; + } + + return g_strdup_printf("/local/domain/%u/device/%s/%s", + xendev->frontend_id, device, xendev->name); +} + +static void xen_device_unplug(XenDevice *xendev, Error **errp) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + const char *type = object_get_typename(OBJECT(xendev)); + Error *local_err = NULL; + xs_transaction_t tid; + + trace_xen_device_unplug(type, xendev->name); + + /* Mimic the way the Xen toolstack does an unplug */ +again: + tid = xs_transaction_start(xenbus->xsh); + if (tid == XBT_NULL) { + error_setg_errno(errp, errno, "failed xs_transaction_start"); + return; + } + + xs_node_printf(xenbus->xsh, tid, xendev->backend_path, "online", + &local_err, "%u", 0); + if (local_err) { + goto abort; + } + + xs_node_printf(xenbus->xsh, tid, xendev->backend_path, "state", + &local_err, "%u", XenbusStateClosing); + if (local_err) { + goto abort; + } + + if (!xs_transaction_end(xenbus->xsh, tid, false)) { + if (errno == EAGAIN) { + goto again; + } + + error_setg_errno(errp, errno, "failed xs_transaction_end"); + } + + return; + +abort: + /* + * We only abort if there is already a failure so ignore any error + * from ending the transaction. + */ + xs_transaction_end(xenbus->xsh, tid, true); + error_propagate(errp, local_err); +} + +static void xen_bus_print_dev(Monitor *mon, DeviceState *dev, int indent) +{ + XenDevice *xendev = XEN_DEVICE(dev); + + monitor_printf(mon, "%*sname = '%s' frontend_id = %u\n", + indent, "", xendev->name, xendev->frontend_id); +} + +static char *xen_bus_get_dev_path(DeviceState *dev) +{ + return xen_device_get_backend_path(XEN_DEVICE(dev)); +} + +struct XenWatch { + char *node, *key; + char *token; + XenWatchHandler handler; + void *opaque; + Notifier notifier; +}; + +static void watch_notify(Notifier *n, void *data) +{ + XenWatch *watch = container_of(n, XenWatch, notifier); + const char *token = data; + + if (!strcmp(watch->token, token)) { + watch->handler(watch->opaque); + } +} + +static XenWatch *new_watch(const char *node, const char *key, + XenWatchHandler handler, void *opaque) +{ + XenWatch *watch = g_new0(XenWatch, 1); + QemuUUID uuid; + + qemu_uuid_generate(&uuid); + + watch->token = qemu_uuid_unparse_strdup(&uuid); + watch->node = g_strdup(node); + watch->key = g_strdup(key); + watch->handler = handler; + watch->opaque = opaque; + watch->notifier.notify = watch_notify; + + return watch; +} + +static void free_watch(XenWatch *watch) +{ + g_free(watch->token); + g_free(watch->key); + g_free(watch->node); + + g_free(watch); +} + +static XenWatch *xen_bus_add_watch(XenBus *xenbus, const char *node, + const char *key, XenWatchHandler handler, + void *opaque, Error **errp) +{ + XenWatch *watch = new_watch(node, key, handler, opaque); + Error *local_err = NULL; + + trace_xen_bus_add_watch(watch->node, watch->key, watch->token); + + notifier_list_add(&xenbus->watch_notifiers, &watch->notifier); + + xs_node_watch(xenbus->xsh, node, key, watch->token, &local_err); + if (local_err) { + error_propagate(errp, local_err); + + notifier_remove(&watch->notifier); + free_watch(watch); + + return NULL; + } + + return watch; +} + +static void xen_bus_remove_watch(XenBus *xenbus, XenWatch *watch, + Error **errp) +{ + trace_xen_bus_remove_watch(watch->node, watch->key, watch->token); + + xs_node_unwatch(xenbus->xsh, watch->node, watch->key, watch->token, + errp); + + notifier_remove(&watch->notifier); + free_watch(watch); +} + +static void xen_bus_backend_create(XenBus *xenbus, const char *type, + const char *name, char *path, + Error **errp) +{ + xs_transaction_t tid; + char **key; + QDict *opts; + unsigned int i, n; + Error *local_err = NULL; + + trace_xen_bus_backend_create(type, path); + +again: + tid = xs_transaction_start(xenbus->xsh); + if (tid == XBT_NULL) { + error_setg(errp, "failed xs_transaction_start"); + return; + } + + key = xs_directory(xenbus->xsh, tid, path, &n); + if (!key) { + if (!xs_transaction_end(xenbus->xsh, tid, true)) { + error_setg_errno(errp, errno, "failed xs_transaction_end"); + } + return; + } + + opts = qdict_new(); + for (i = 0; i < n; i++) { + char *val; + + /* + * Assume anything found in the xenstore backend area, other than + * the keys created for a generic XenDevice, are parameters + * to be used to configure the backend. + */ + if (!strcmp(key[i], "state") || + !strcmp(key[i], "online") || + !strcmp(key[i], "frontend") || + !strcmp(key[i], "frontend-id") || + !strcmp(key[i], "hotplug-status")) + continue; + + if (xs_node_scanf(xenbus->xsh, tid, path, key[i], NULL, "%ms", + &val) == 1) { + qdict_put_str(opts, key[i], val); + free(val); + } + } + + free(key); + + if (!xs_transaction_end(xenbus->xsh, tid, false)) { + qobject_unref(opts); + + if (errno == EAGAIN) { + goto again; + } + + error_setg_errno(errp, errno, "failed xs_transaction_end"); + return; + } + + xen_backend_device_create(xenbus, type, name, opts, &local_err); + qobject_unref(opts); + + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to create '%s' device '%s': ", + type, name); + } +} + +static void xen_bus_type_enumerate(XenBus *xenbus, const char *type) +{ + char *domain_path = g_strdup_printf("backend/%s/%u", type, xen_domid); + char **backend; + unsigned int i, n; + + trace_xen_bus_type_enumerate(type); + + backend = xs_directory(xenbus->xsh, XBT_NULL, domain_path, &n); + if (!backend) { + goto out; + } + + for (i = 0; i < n; i++) { + char *backend_path = g_strdup_printf("%s/%s", domain_path, + backend[i]); + enum xenbus_state backend_state; + + if (xs_node_scanf(xenbus->xsh, XBT_NULL, backend_path, "state", + NULL, "%u", &backend_state) != 1) + backend_state = XenbusStateUnknown; + + if (backend_state == XenbusStateInitialising) { + Error *local_err = NULL; + + xen_bus_backend_create(xenbus, type, backend[i], backend_path, + &local_err); + if (local_err) { + error_report_err(local_err); + } + } + + g_free(backend_path); + } + + free(backend); + +out: + g_free(domain_path); +} + +static void xen_bus_enumerate(void *opaque) +{ + XenBus *xenbus = opaque; + char **type; + unsigned int i, n; + + trace_xen_bus_enumerate(); + + type = xs_directory(xenbus->xsh, XBT_NULL, "backend", &n); + if (!type) { + return; + } + + for (i = 0; i < n; i++) { + xen_bus_type_enumerate(xenbus, type[i]); + } + + free(type); +} + +static void xen_bus_unrealize(BusState *bus, Error **errp) +{ + XenBus *xenbus = XEN_BUS(bus); + + trace_xen_bus_unrealize(); + + if (xenbus->backend_watch) { + xen_bus_remove_watch(xenbus, xenbus->backend_watch, NULL); + xenbus->backend_watch = NULL; + } + + if (!xenbus->xsh) { + return; + } + + qemu_set_fd_handler(xs_fileno(xenbus->xsh), NULL, NULL, NULL); + + xs_close(xenbus->xsh); +} + +static void xen_bus_watch(void *opaque) +{ + XenBus *xenbus = opaque; + char **v; + const char *token; + + g_assert(xenbus->xsh); + + v = xs_check_watch(xenbus->xsh); + if (!v) { + return; + } + + token = v[XS_WATCH_TOKEN]; + + trace_xen_bus_watch(token); + + notifier_list_notify(&xenbus->watch_notifiers, (void *)token); + + free(v); +} + +static void xen_bus_realize(BusState *bus, Error **errp) +{ + XenBus *xenbus = XEN_BUS(bus); + unsigned int domid; + Error *local_err = NULL; + + trace_xen_bus_realize(); + + xenbus->xsh = xs_open(0); + if (!xenbus->xsh) { + error_setg_errno(errp, errno, "failed xs_open"); + goto fail; + } + + if (xs_node_scanf(xenbus->xsh, XBT_NULL, "", /* domain root node */ + "domid", NULL, "%u", &domid) == 1) { + xenbus->backend_id = domid; + } else { + xenbus->backend_id = 0; /* Assume lack of node means dom0 */ + } + + notifier_list_init(&xenbus->watch_notifiers); + qemu_set_fd_handler(xs_fileno(xenbus->xsh), xen_bus_watch, NULL, + xenbus); + + module_call_init(MODULE_INIT_XEN_BACKEND); + + xenbus->backend_watch = + xen_bus_add_watch(xenbus, "", /* domain root node */ + "backend", xen_bus_enumerate, xenbus, &local_err); + if (local_err) { + /* This need not be treated as a hard error so don't propagate */ + error_reportf_err(local_err, + "failed to set up enumeration watch: "); + } + + return; + +fail: + xen_bus_unrealize(bus, &error_abort); +} + +static void xen_bus_unplug_request(HotplugHandler *hotplug, + DeviceState *dev, + Error **errp) +{ + XenDevice *xendev = XEN_DEVICE(dev); + + xen_device_unplug(xendev, errp); +} + +static void xen_bus_class_init(ObjectClass *class, void *data) +{ + BusClass *bus_class = BUS_CLASS(class); + HotplugHandlerClass *hotplug_class = HOTPLUG_HANDLER_CLASS(class); + + bus_class->print_dev = xen_bus_print_dev; + bus_class->get_dev_path = xen_bus_get_dev_path; + bus_class->realize = xen_bus_realize; + bus_class->unrealize = xen_bus_unrealize; + + hotplug_class->unplug_request = xen_bus_unplug_request; +} + +static const TypeInfo xen_bus_type_info = { + .name = TYPE_XEN_BUS, + .parent = TYPE_BUS, + .instance_size = sizeof(XenBus), + .class_size = sizeof(XenBusClass), + .class_init = xen_bus_class_init, + .interfaces = (InterfaceInfo[]) { + { TYPE_HOTPLUG_HANDLER }, + { } + }, +}; + +void xen_device_backend_printf(XenDevice *xendev, const char *key, + const char *fmt, ...) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + Error *local_err = NULL; + va_list ap; + + g_assert(xenbus->xsh); + + va_start(ap, fmt); + xs_node_vprintf(xenbus->xsh, XBT_NULL, xendev->backend_path, key, + &local_err, fmt, ap); + va_end(ap); + + if (local_err) { + error_report_err(local_err); + } +} + +static int xen_device_backend_scanf(XenDevice *xendev, const char *key, + const char *fmt, ...) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + va_list ap; + int rc; + + g_assert(xenbus->xsh); + + va_start(ap, fmt); + rc = xs_node_vscanf(xenbus->xsh, XBT_NULL, xendev->backend_path, key, + NULL, fmt, ap); + va_end(ap); + + return rc; +} + +void xen_device_backend_set_state(XenDevice *xendev, + enum xenbus_state state) +{ + const char *type = object_get_typename(OBJECT(xendev)); + + if (xendev->backend_state == state) { + return; + } + + trace_xen_device_backend_state(type, xendev->name, + xs_strstate(state)); + + xendev->backend_state = state; + xen_device_backend_printf(xendev, "state", "%u", state); +} + +enum xenbus_state xen_device_backend_get_state(XenDevice *xendev) +{ + return xendev->backend_state; +} + +static void xen_device_backend_set_online(XenDevice *xendev, bool online) +{ + const char *type = object_get_typename(OBJECT(xendev)); + + if (xendev->backend_online == online) { + return; + } + + trace_xen_device_backend_online(type, xendev->name, online); + + xendev->backend_online = online; + xen_device_backend_printf(xendev, "online", "%u", online); +} + +static void xen_device_backend_changed(void *opaque) +{ + XenDevice *xendev = opaque; + const char *type = object_get_typename(OBJECT(xendev)); + enum xenbus_state state; + unsigned int online; + + trace_xen_device_backend_changed(type, xendev->name); + + if (xen_device_backend_scanf(xendev, "state", "%u", &state) != 1) { + state = XenbusStateUnknown; + } + + xen_device_backend_set_state(xendev, state); + + if (xen_device_backend_scanf(xendev, "online", "%u", &online) != 1) { + online = 0; + } + + xen_device_backend_set_online(xendev, !!online); + + /* + * If the toolstack (or unplug request callback) has set the backend + * state to Closing, but there is no active frontend (i.e. the + * state is not Connected) then set the backend state to Closed. + */ + if (xendev->backend_state == XenbusStateClosing && + xendev->frontend_state != XenbusStateConnected) { + xen_device_backend_set_state(xendev, XenbusStateClosed); + } + + /* + * If a backend is still 'online' then its state should be cycled + * back round to InitWait in order for a new frontend instance to + * connect. This may happen when, for example, a frontend driver is + * re-installed or updated. + * If a backend is not 'online' then the device should be destroyed. + */ + if (xendev->backend_online && + xendev->backend_state == XenbusStateClosed) { + xen_device_backend_set_state(xendev, XenbusStateInitWait); + } else if (!xendev->backend_online && + (xendev->backend_state == XenbusStateClosed || + xendev->backend_state == XenbusStateInitialising || + xendev->backend_state == XenbusStateInitWait || + xendev->backend_state == XenbusStateUnknown)) { + Error *local_err = NULL; + + if (!xen_backend_try_device_destroy(xendev, &local_err)) { + object_unparent(OBJECT(xendev)); + } + + if (local_err) { + error_report_err(local_err); + } + } +} + +static void xen_device_backend_create(XenDevice *xendev, Error **errp) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + struct xs_permissions perms[2]; + Error *local_err = NULL; + + xendev->backend_path = xen_device_get_backend_path(xendev); + + perms[0].id = xenbus->backend_id; + perms[0].perms = XS_PERM_NONE; + perms[1].id = xendev->frontend_id; + perms[1].perms = XS_PERM_READ; + + g_assert(xenbus->xsh); + + xs_node_create(xenbus->xsh, XBT_NULL, xendev->backend_path, perms, + ARRAY_SIZE(perms), &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to create backend: "); + return; + } + + xendev->backend_state_watch = + xen_bus_add_watch(xenbus, xendev->backend_path, + "state", xen_device_backend_changed, + xendev, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to watch backend state: "); + return; + } + + xendev->backend_online_watch = + xen_bus_add_watch(xenbus, xendev->backend_path, + "online", xen_device_backend_changed, + xendev, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to watch backend online: "); + return; + } +} + +static void xen_device_backend_destroy(XenDevice *xendev) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + Error *local_err = NULL; + + if (xendev->backend_online_watch) { + xen_bus_remove_watch(xenbus, xendev->backend_online_watch, NULL); + xendev->backend_online_watch = NULL; + } + + if (xendev->backend_state_watch) { + xen_bus_remove_watch(xenbus, xendev->backend_state_watch, NULL); + xendev->backend_state_watch = NULL; + } + + if (!xendev->backend_path) { + return; + } + + g_assert(xenbus->xsh); + + xs_node_destroy(xenbus->xsh, XBT_NULL, xendev->backend_path, + &local_err); + g_free(xendev->backend_path); + xendev->backend_path = NULL; + + if (local_err) { + error_report_err(local_err); + } +} + +void xen_device_frontend_printf(XenDevice *xendev, const char *key, + const char *fmt, ...) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + Error *local_err = NULL; + va_list ap; + + g_assert(xenbus->xsh); + + va_start(ap, fmt); + xs_node_vprintf(xenbus->xsh, XBT_NULL, xendev->frontend_path, key, + &local_err, fmt, ap); + va_end(ap); + + if (local_err) { + error_report_err(local_err); + } +} + +int xen_device_frontend_scanf(XenDevice *xendev, const char *key, + const char *fmt, ...) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + va_list ap; + int rc; + + g_assert(xenbus->xsh); + + va_start(ap, fmt); + rc = xs_node_vscanf(xenbus->xsh, XBT_NULL, xendev->frontend_path, key, + NULL, fmt, ap); + va_end(ap); + + return rc; +} + +static void xen_device_frontend_set_state(XenDevice *xendev, + enum xenbus_state state) +{ + const char *type = object_get_typename(OBJECT(xendev)); + + if (xendev->frontend_state == state) { + return; + } + + trace_xen_device_frontend_state(type, xendev->name, + xs_strstate(state)); + + xendev->frontend_state = state; + xen_device_frontend_printf(xendev, "state", "%u", state); +} + +static void xen_device_frontend_changed(void *opaque) +{ + XenDevice *xendev = opaque; + XenDeviceClass *xendev_class = XEN_DEVICE_GET_CLASS(xendev); + const char *type = object_get_typename(OBJECT(xendev)); + enum xenbus_state state; + + trace_xen_device_frontend_changed(type, xendev->name); + + if (xen_device_frontend_scanf(xendev, "state", "%u", &state) != 1) { + state = XenbusStateUnknown; + } + + xen_device_frontend_set_state(xendev, state); + + if (xendev_class->frontend_changed) { + Error *local_err = NULL; + + xendev_class->frontend_changed(xendev, state, &local_err); + + if (local_err) { + error_reportf_err(local_err, "frontend change error: "); + } + } +} + +static void xen_device_frontend_create(XenDevice *xendev, Error **errp) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + struct xs_permissions perms[2]; + Error *local_err = NULL; + + xendev->frontend_path = xen_device_get_frontend_path(xendev); + + perms[0].id = xendev->frontend_id; + perms[0].perms = XS_PERM_NONE; + perms[1].id = xenbus->backend_id; + perms[1].perms = XS_PERM_READ | XS_PERM_WRITE; + + g_assert(xenbus->xsh); + + xs_node_create(xenbus->xsh, XBT_NULL, xendev->frontend_path, perms, + ARRAY_SIZE(perms), &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to create frontend: "); + return; + } + + xendev->frontend_state_watch = + xen_bus_add_watch(xenbus, xendev->frontend_path, "state", + xen_device_frontend_changed, xendev, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to watch frontend state: "); + } +} + +static void xen_device_frontend_destroy(XenDevice *xendev) +{ + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + Error *local_err = NULL; + + if (xendev->frontend_state_watch) { + xen_bus_remove_watch(xenbus, xendev->frontend_state_watch, NULL); + xendev->frontend_state_watch = NULL; + } + + if (!xendev->frontend_path) { + return; + } + + g_assert(xenbus->xsh); + + xs_node_destroy(xenbus->xsh, XBT_NULL, xendev->frontend_path, + &local_err); + g_free(xendev->frontend_path); + xendev->frontend_path = NULL; + + if (local_err) { + error_report_err(local_err); + } +} + +void xen_device_set_max_grant_refs(XenDevice *xendev, unsigned int nr_refs, + Error **errp) +{ + if (xengnttab_set_max_grants(xendev->xgth, nr_refs)) { + error_setg_errno(errp, errno, "xengnttab_set_max_grants failed"); + } +} + +void *xen_device_map_grant_refs(XenDevice *xendev, uint32_t *refs, + unsigned int nr_refs, int prot, + Error **errp) +{ + void *map = xengnttab_map_domain_grant_refs(xendev->xgth, nr_refs, + xendev->frontend_id, refs, + prot); + + if (!map) { + error_setg_errno(errp, errno, + "xengnttab_map_domain_grant_refs failed"); + } + + return map; +} + +void xen_device_unmap_grant_refs(XenDevice *xendev, void *map, + unsigned int nr_refs, Error **errp) +{ + if (xengnttab_unmap(xendev->xgth, map, nr_refs)) { + error_setg_errno(errp, errno, "xengnttab_unmap failed"); + } +} + +static void compat_copy_grant_refs(XenDevice *xendev, bool to_domain, + XenDeviceGrantCopySegment segs[], + unsigned int nr_segs, Error **errp) +{ + uint32_t *refs = g_new(uint32_t, nr_segs); + int prot = to_domain ? PROT_WRITE : PROT_READ; + void *map; + unsigned int i; + + for (i = 0; i < nr_segs; i++) { + XenDeviceGrantCopySegment *seg = &segs[i]; + + refs[i] = to_domain ? seg->dest.foreign.ref : + seg->source.foreign.ref; + } + + map = xengnttab_map_domain_grant_refs(xendev->xgth, nr_segs, + xendev->frontend_id, refs, + prot); + if (!map) { + error_setg_errno(errp, errno, + "xengnttab_map_domain_grant_refs failed"); + goto done; + } + + for (i = 0; i < nr_segs; i++) { + XenDeviceGrantCopySegment *seg = &segs[i]; + void *page = map + (i * XC_PAGE_SIZE); + + if (to_domain) { + memcpy(page + seg->dest.foreign.offset, seg->source.virt, + seg->len); + } else { + memcpy(seg->dest.virt, page + seg->source.foreign.offset, + seg->len); + } + } + + if (xengnttab_unmap(xendev->xgth, map, nr_segs)) { + error_setg_errno(errp, errno, "xengnttab_unmap failed"); + } + +done: + g_free(refs); +} + +void xen_device_copy_grant_refs(XenDevice *xendev, bool to_domain, + XenDeviceGrantCopySegment segs[], + unsigned int nr_segs, Error **errp) +{ + xengnttab_grant_copy_segment_t *xengnttab_segs; + unsigned int i; + + if (!xendev->feature_grant_copy) { + compat_copy_grant_refs(xendev, to_domain, segs, nr_segs, errp); + return; + } + + xengnttab_segs = g_new0(xengnttab_grant_copy_segment_t, nr_segs); + + for (i = 0; i < nr_segs; i++) { + XenDeviceGrantCopySegment *seg = &segs[i]; + xengnttab_grant_copy_segment_t *xengnttab_seg = &xengnttab_segs[i]; + + if (to_domain) { + xengnttab_seg->flags = GNTCOPY_dest_gref; + xengnttab_seg->dest.foreign.domid = xendev->frontend_id; + xengnttab_seg->dest.foreign.ref = seg->dest.foreign.ref; + xengnttab_seg->dest.foreign.offset = seg->dest.foreign.offset; + xengnttab_seg->source.virt = seg->source.virt; + } else { + xengnttab_seg->flags = GNTCOPY_source_gref; + xengnttab_seg->source.foreign.domid = xendev->frontend_id; + xengnttab_seg->source.foreign.ref = seg->source.foreign.ref; + xengnttab_seg->source.foreign.offset = + seg->source.foreign.offset; + xengnttab_seg->dest.virt = seg->dest.virt; + } + + xengnttab_seg->len = seg->len; + } + + if (xengnttab_grant_copy(xendev->xgth, nr_segs, xengnttab_segs)) { + error_setg_errno(errp, errno, "xengnttab_grant_copy failed"); + goto done; + } + + for (i = 0; i < nr_segs; i++) { + xengnttab_grant_copy_segment_t *xengnttab_seg = &xengnttab_segs[i]; + + if (xengnttab_seg->status != GNTST_okay) { + error_setg(errp, "xengnttab_grant_copy seg[%u] failed", i); + break; + } + } + +done: + g_free(xengnttab_segs); +} + +struct XenEventChannel { + evtchn_port_t local_port; + XenEventHandler handler; + void *opaque; + Notifier notifier; +}; + +static void event_notify(Notifier *n, void *data) +{ + XenEventChannel *channel = container_of(n, XenEventChannel, notifier); + unsigned long port = (unsigned long)data; + + if (port == channel->local_port) { + channel->handler(channel->opaque); + } +} + +XenEventChannel *xen_device_bind_event_channel(XenDevice *xendev, + unsigned int port, + XenEventHandler handler, + void *opaque, Error **errp) +{ + XenEventChannel *channel = g_new0(XenEventChannel, 1); + xenevtchn_port_or_error_t local_port; + + local_port = xenevtchn_bind_interdomain(xendev->xeh, + xendev->frontend_id, + port); + if (local_port < 0) { + error_setg_errno(errp, errno, "xenevtchn_bind_interdomain failed"); + + g_free(channel); + return NULL; + } + + channel->local_port = local_port; + channel->handler = handler; + channel->opaque = opaque; + channel->notifier.notify = event_notify; + + notifier_list_add(&xendev->event_notifiers, &channel->notifier); + + return channel; +} + +void xen_device_notify_event_channel(XenDevice *xendev, + XenEventChannel *channel, + Error **errp) +{ + if (!channel) { + error_setg(errp, "bad channel"); + return; + } + + if (xenevtchn_notify(xendev->xeh, channel->local_port) < 0) { + error_setg_errno(errp, errno, "xenevtchn_notify failed"); + } +} + +void xen_device_unbind_event_channel(XenDevice *xendev, + XenEventChannel *channel, + Error **errp) +{ + if (!channel) { + error_setg(errp, "bad channel"); + return; + } + + notifier_remove(&channel->notifier); + + if (xenevtchn_unbind(xendev->xeh, channel->local_port) < 0) { + error_setg_errno(errp, errno, "xenevtchn_unbind failed"); + } + + g_free(channel); +} + +static void xen_device_unrealize(DeviceState *dev, Error **errp) +{ + XenDevice *xendev = XEN_DEVICE(dev); + XenDeviceClass *xendev_class = XEN_DEVICE_GET_CLASS(xendev); + const char *type = object_get_typename(OBJECT(xendev)); + + if (!xendev->name) { + return; + } + + trace_xen_device_unrealize(type, xendev->name); + + if (xendev->exit.notify) { + qemu_remove_exit_notifier(&xendev->exit); + xendev->exit.notify = NULL; + } + + if (xendev_class->unrealize) { + xendev_class->unrealize(xendev, errp); + } + + xen_device_frontend_destroy(xendev); + xen_device_backend_destroy(xendev); + + if (xendev->xeh) { + qemu_set_fd_handler(xenevtchn_fd(xendev->xeh), NULL, NULL, NULL); + xenevtchn_close(xendev->xeh); + xendev->xeh = NULL; + } + + if (xendev->xgth) { + xengnttab_close(xendev->xgth); + xendev->xgth = NULL; + } + + g_free(xendev->name); + xendev->name = NULL; +} + +static void xen_device_exit(Notifier *n, void *data) +{ + XenDevice *xendev = container_of(n, XenDevice, exit); + + xen_device_unrealize(DEVICE(xendev), &error_abort); +} + +static void xen_device_event(void *opaque) +{ + XenDevice *xendev = opaque; + unsigned long port = xenevtchn_pending(xendev->xeh); + + notifier_list_notify(&xendev->event_notifiers, (void *)port); + + xenevtchn_unmask(xendev->xeh, port); +} + +static void xen_device_realize(DeviceState *dev, Error **errp) +{ + XenDevice *xendev = XEN_DEVICE(dev); + XenDeviceClass *xendev_class = XEN_DEVICE_GET_CLASS(xendev); + XenBus *xenbus = XEN_BUS(qdev_get_parent_bus(DEVICE(xendev))); + const char *type = object_get_typename(OBJECT(xendev)); + Error *local_err = NULL; + + if (xendev->frontend_id == DOMID_INVALID) { + xendev->frontend_id = xen_domid; + } + + if (xendev->frontend_id >= DOMID_FIRST_RESERVED) { + error_setg(errp, "invalid frontend-id"); + goto unrealize; + } + + if (!xendev_class->get_name) { + error_setg(errp, "get_name method not implemented"); + goto unrealize; + } + + xendev->name = xendev_class->get_name(xendev, &local_err); + if (local_err) { + error_propagate_prepend(errp, local_err, + "failed to get device name: "); + goto unrealize; + } + + trace_xen_device_realize(type, xendev->name); + + xendev->xgth = xengnttab_open(NULL, 0); + if (!xendev->xgth) { + error_setg_errno(errp, errno, "failed xengnttab_open"); + goto unrealize; + } + + xendev->feature_grant_copy = + (xengnttab_grant_copy(xendev->xgth, 0, NULL) == 0); + + xendev->xeh = xenevtchn_open(NULL, 0); + if (!xendev->xeh) { + error_setg_errno(errp, errno, "failed xenevtchn_open"); + goto unrealize; + } + + notifier_list_init(&xendev->event_notifiers); + qemu_set_fd_handler(xenevtchn_fd(xendev->xeh), xen_device_event, NULL, + xendev); + + xen_device_backend_create(xendev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + goto unrealize; + } + + xen_device_frontend_create(xendev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + goto unrealize; + } + + if (xendev_class->realize) { + xendev_class->realize(xendev, &local_err); + if (local_err) { + error_propagate(errp, local_err); + goto unrealize; + } + } + + xen_device_backend_printf(xendev, "frontend", "%s", + xendev->frontend_path); + xen_device_backend_printf(xendev, "frontend-id", "%u", + xendev->frontend_id); + xen_device_backend_printf(xendev, "hotplug-status", "connected"); + + xen_device_backend_set_online(xendev, true); + xen_device_backend_set_state(xendev, XenbusStateInitWait); + + xen_device_frontend_printf(xendev, "backend", "%s", + xendev->backend_path); + xen_device_frontend_printf(xendev, "backend-id", "%u", + xenbus->backend_id); + + xen_device_frontend_set_state(xendev, XenbusStateInitialising); + + xendev->exit.notify = xen_device_exit; + qemu_add_exit_notifier(&xendev->exit); + return; + +unrealize: + xen_device_unrealize(dev, &error_abort); +} + +static Property xen_device_props[] = { + DEFINE_PROP_UINT16("frontend-id", XenDevice, frontend_id, + DOMID_INVALID), + DEFINE_PROP_END_OF_LIST() +}; + +static void xen_device_class_init(ObjectClass *class, void *data) +{ + DeviceClass *dev_class = DEVICE_CLASS(class); + + dev_class->realize = xen_device_realize; + dev_class->unrealize = xen_device_unrealize; + dev_class->props = xen_device_props; + dev_class->bus_type = TYPE_XEN_BUS; +} + +static const TypeInfo xen_device_type_info = { + .name = TYPE_XEN_DEVICE, + .parent = TYPE_DEVICE, + .instance_size = sizeof(XenDevice), + .abstract = true, + .class_size = sizeof(XenDeviceClass), + .class_init = xen_device_class_init, +}; + +typedef struct XenBridge { + SysBusDevice busdev; +} XenBridge; + +#define TYPE_XEN_BRIDGE "xen-bridge" + +static const TypeInfo xen_bridge_type_info = { + .name = TYPE_XEN_BRIDGE, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(XenBridge), +}; + +static void xen_register_types(void) +{ + type_register_static(&xen_bridge_type_info); + type_register_static(&xen_bus_type_info); + type_register_static(&xen_device_type_info); +} + +type_init(xen_register_types) + +void xen_bus_init(void) +{ + DeviceState *dev = qdev_create(NULL, TYPE_XEN_BRIDGE); + BusState *bus = qbus_create(TYPE_XEN_BUS, dev, NULL); + + qdev_init_nofail(dev); + qbus_set_bus_hotplug_handler(bus, &error_abort); +} diff --git a/hw/xen/xen-common.c b/hw/xen/xen-common.c index 18a9045556..0e9e58f04d 100644 --- a/hw/xen/xen-common.c +++ b/hw/xen/xen-common.c @@ -10,7 +10,7 @@ #include "qemu/osdep.h" #include "qemu/error-report.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "chardev/char.h" #include "sysemu/accel.h" #include "migration/misc.h" diff --git a/hw/xen/xen_backend.c b/hw/xen/xen-legacy-backend.c similarity index 89% rename from hw/xen/xen_backend.c rename to hw/xen/xen-legacy-backend.c index 0bc6b1de60..36fd1e9b09 100644 --- a/hw/xen/xen_backend.c +++ b/hw/xen/xen-legacy-backend.c @@ -30,7 +30,7 @@ #include "hw/boards.h" #include "qemu/log.h" #include "qapi/error.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "hw/xen/xen_pvdev.h" #include "monitor/qdev.h" @@ -42,49 +42,54 @@ BusState *xen_sysbus; /* ------------------------------------------------------------- */ /* public */ -struct xs_handle *xenstore = NULL; +struct xs_handle *xenstore; const char *xen_protocol; /* private */ static bool xen_feature_grant_copy; static int debug; -int xenstore_write_be_str(struct XenDevice *xendev, const char *node, const char *val) +int xenstore_write_be_str(struct XenLegacyDevice *xendev, const char *node, + const char *val) { return xenstore_write_str(xendev->be, node, val); } -int xenstore_write_be_int(struct XenDevice *xendev, const char *node, int ival) +int xenstore_write_be_int(struct XenLegacyDevice *xendev, const char *node, + int ival) { return xenstore_write_int(xendev->be, node, ival); } -int xenstore_write_be_int64(struct XenDevice *xendev, const char *node, int64_t ival) +int xenstore_write_be_int64(struct XenLegacyDevice *xendev, const char *node, + int64_t ival) { return xenstore_write_int64(xendev->be, node, ival); } -char *xenstore_read_be_str(struct XenDevice *xendev, const char *node) +char *xenstore_read_be_str(struct XenLegacyDevice *xendev, const char *node) { return xenstore_read_str(xendev->be, node); } -int xenstore_read_be_int(struct XenDevice *xendev, const char *node, int *ival) +int xenstore_read_be_int(struct XenLegacyDevice *xendev, const char *node, + int *ival) { return xenstore_read_int(xendev->be, node, ival); } -char *xenstore_read_fe_str(struct XenDevice *xendev, const char *node) +char *xenstore_read_fe_str(struct XenLegacyDevice *xendev, const char *node) { return xenstore_read_str(xendev->fe, node); } -int xenstore_read_fe_int(struct XenDevice *xendev, const char *node, int *ival) +int xenstore_read_fe_int(struct XenLegacyDevice *xendev, const char *node, + int *ival) { return xenstore_read_int(xendev->fe, node, ival); } -int xenstore_read_fe_uint64(struct XenDevice *xendev, const char *node, +int xenstore_read_fe_uint64(struct XenLegacyDevice *xendev, const char *node, uint64_t *uval) { return xenstore_read_uint64(xendev->fe, node, uval); @@ -92,7 +97,7 @@ int xenstore_read_fe_uint64(struct XenDevice *xendev, const char *node, /* ------------------------------------------------------------- */ -int xen_be_set_state(struct XenDevice *xendev, enum xenbus_state state) +int xen_be_set_state(struct XenLegacyDevice *xendev, enum xenbus_state state) { int rc; @@ -106,7 +111,7 @@ int xen_be_set_state(struct XenDevice *xendev, enum xenbus_state state) return 0; } -void xen_be_set_max_grant_refs(struct XenDevice *xendev, +void xen_be_set_max_grant_refs(struct XenLegacyDevice *xendev, unsigned int nr_refs) { assert(xendev->ops->flags & DEVOPS_FLAG_NEED_GNTDEV); @@ -117,7 +122,7 @@ void xen_be_set_max_grant_refs(struct XenDevice *xendev, } } -void *xen_be_map_grant_refs(struct XenDevice *xendev, uint32_t *refs, +void *xen_be_map_grant_refs(struct XenLegacyDevice *xendev, uint32_t *refs, unsigned int nr_refs, int prot) { void *ptr; @@ -135,7 +140,7 @@ void *xen_be_map_grant_refs(struct XenDevice *xendev, uint32_t *refs, return ptr; } -void xen_be_unmap_grant_refs(struct XenDevice *xendev, void *ptr, +void xen_be_unmap_grant_refs(struct XenLegacyDevice *xendev, void *ptr, unsigned int nr_refs) { assert(xendev->ops->flags & DEVOPS_FLAG_NEED_GNTDEV); @@ -146,7 +151,7 @@ void xen_be_unmap_grant_refs(struct XenDevice *xendev, void *ptr, } } -static int compat_copy_grant_refs(struct XenDevice *xendev, +static int compat_copy_grant_refs(struct XenLegacyDevice *xendev, bool to_domain, XenGrantCopySegment segs[], unsigned int nr_segs) @@ -195,7 +200,7 @@ static int compat_copy_grant_refs(struct XenDevice *xendev, return 0; } -int xen_be_copy_grant_refs(struct XenDevice *xendev, +int xen_be_copy_grant_refs(struct XenLegacyDevice *xendev, bool to_domain, XenGrantCopySegment segs[], unsigned int nr_segs) @@ -259,10 +264,11 @@ int xen_be_copy_grant_refs(struct XenDevice *xendev, /* * get xen backend device, allocate a new one if it doesn't exist. */ -static struct XenDevice *xen_be_get_xendev(const char *type, int dom, int dev, - struct XenDevOps *ops) +static struct XenLegacyDevice *xen_be_get_xendev(const char *type, int dom, + int dev, + struct XenDevOps *ops) { - struct XenDevice *xendev; + struct XenLegacyDevice *xendev; xendev = xen_pv_find_xendev(type, dom, dev); if (xendev) { @@ -314,7 +320,8 @@ static struct XenDevice *xen_be_get_xendev(const char *type, int dom, int dev, * Node specifies the changed field. node = NULL means * update all fields (used for initialization). */ -static void xen_be_backend_changed(struct XenDevice *xendev, const char *node) +static void xen_be_backend_changed(struct XenLegacyDevice *xendev, + const char *node) { if (node == NULL || strcmp(node, "online") == 0) { if (xenstore_read_be_int(xendev, "online", &xendev->online) == -1) { @@ -330,7 +337,8 @@ static void xen_be_backend_changed(struct XenDevice *xendev, const char *node) } } -static void xen_be_frontend_changed(struct XenDevice *xendev, const char *node) +static void xen_be_frontend_changed(struct XenLegacyDevice *xendev, + const char *node) { int fe_state; @@ -373,7 +381,7 @@ static void xen_be_frontend_changed(struct XenDevice *xendev, const char *node) * only affects the xendev->be_state variable as xenbus should * already be put into that state by xend. */ -static int xen_be_try_setup(struct XenDevice *xendev) +static int xen_be_try_setup(struct XenLegacyDevice *xendev) { char token[XEN_BUFSIZE]; int be_state; @@ -417,7 +425,7 @@ static int xen_be_try_setup(struct XenDevice *xendev) * * Goes to InitWait on success. */ -static int xen_be_try_init(struct XenDevice *xendev) +static int xen_be_try_init(struct XenLegacyDevice *xendev) { int rc = 0; @@ -446,7 +454,7 @@ static int xen_be_try_init(struct XenDevice *xendev) * * Goes to Connected on success. */ -static int xen_be_try_initialise(struct XenDevice *xendev) +static int xen_be_try_initialise(struct XenLegacyDevice *xendev) { int rc = 0; @@ -487,7 +495,7 @@ static int xen_be_try_initialise(struct XenDevice *xendev) * frontend being Connected. Note that this may be called more * than once since the backend state is not modified. */ -static void xen_be_try_connected(struct XenDevice *xendev) +static void xen_be_try_connected(struct XenLegacyDevice *xendev) { if (!xendev->ops->connected) { return; @@ -510,7 +518,8 @@ static void xen_be_try_connected(struct XenDevice *xendev) * * Goes to Closed when done. */ -static void xen_be_disconnect(struct XenDevice *xendev, enum xenbus_state state) +static void xen_be_disconnect(struct XenLegacyDevice *xendev, + enum xenbus_state state) { if (xendev->be_state != XenbusStateClosing && xendev->be_state != XenbusStateClosed && @@ -529,7 +538,7 @@ static void xen_be_disconnect(struct XenDevice *xendev, enum xenbus_state state) /* * Try to reset xendev, for reconnection by another frontend instance. */ -static int xen_be_try_reset(struct XenDevice *xendev) +static int xen_be_try_reset(struct XenLegacyDevice *xendev) { if (xendev->fe_state != XenbusStateInitialising) { return -1; @@ -543,7 +552,7 @@ static int xen_be_try_reset(struct XenDevice *xendev) /* * state change dispatcher function */ -void xen_be_check_state(struct XenDevice *xendev) +void xen_be_check_state(struct XenLegacyDevice *xendev) { int rc = 0; @@ -587,7 +596,7 @@ void xen_be_check_state(struct XenDevice *xendev) static int xenstore_scan(const char *type, int dom, struct XenDevOps *ops) { - struct XenDevice *xendev; + struct XenLegacyDevice *xendev; char path[XEN_BUFSIZE], token[XEN_BUFSIZE]; char **dev = NULL; unsigned int cdev, j; @@ -620,7 +629,7 @@ static int xenstore_scan(const char *type, int dom, struct XenDevOps *ops) void xenstore_update_be(char *watch, char *type, int dom, struct XenDevOps *ops) { - struct XenDevice *xendev; + struct XenLegacyDevice *xendev; char path[XEN_BUFSIZE], *bepath; unsigned int len, dev; @@ -628,9 +637,9 @@ void xenstore_update_be(char *watch, char *type, int dom, if (strncmp(path, watch, len) != 0) { return; } - if (sscanf(watch+len, "/%u/%255s", &dev, path) != 2) { + if (sscanf(watch + len, "/%u/%255s", &dev, path) != 2) { strcpy(path, ""); - if (sscanf(watch+len, "/%u", &dev) != 1) { + if (sscanf(watch + len, "/%u", &dev) != 1) { dev = -1; } } @@ -651,7 +660,7 @@ void xenstore_update_be(char *watch, char *type, int dom, } } -void xenstore_update_fe(char *watch, struct XenDevice *xendev) +void xenstore_update_fe(char *watch, struct XenLegacyDevice *xendev) { char *node; unsigned int len; @@ -744,7 +753,6 @@ void xen_be_register_common(void) xen_be_register("console", &xen_console_ops); xen_be_register("vkbd", &xen_kbdmouse_ops); - xen_be_register("qdisk", &xen_blkdev_ops); #ifdef CONFIG_VIRTFS xen_be_register("9pfs", &xen_9pfs_ops); #endif @@ -753,7 +761,7 @@ void xen_be_register_common(void) #endif } -int xen_be_bind_evtchn(struct XenDevice *xendev) +int xen_be_bind_evtchn(struct XenLegacyDevice *xendev) { if (xendev->local_port != -1) { return 0; @@ -789,7 +797,7 @@ static const TypeInfo xendev_type_info = { .name = TYPE_XENBACKEND, .parent = TYPE_XENSYSDEV, .class_init = xendev_class_init, - .instance_size = sizeof(struct XenDevice), + .instance_size = sizeof(struct XenLegacyDevice), }; static void xen_sysbus_class_init(ObjectClass *klass, void *data) diff --git a/hw/xen/xen_devconfig.c b/hw/xen/xen_devconfig.c index 3500d88a3e..315dbc9c51 100644 --- a/hw/xen/xen_devconfig.c +++ b/hw/xen/xen_devconfig.c @@ -1,5 +1,5 @@ #include "qemu/osdep.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "qemu/option.h" #include "sysemu/blockdev.h" diff --git a/hw/xen/xen_pt.c b/hw/xen/xen_pt.c index f1f3a3727c..5539d56c3a 100644 --- a/hw/xen/xen_pt.c +++ b/hw/xen/xen_pt.c @@ -59,7 +59,7 @@ #include "hw/pci/pci.h" #include "hw/xen/xen.h" #include "hw/i386/pc.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "xen_pt.h" #include "qemu/range.h" #include "exec/address-spaces.h" @@ -847,6 +847,12 @@ static void xen_pt_realize(PCIDevice *d, Error **errp) } machine_irq = s->real_device.irq; + if (machine_irq == 0) { + XEN_PT_LOG(d, "machine irq is 0\n"); + cmd |= PCI_COMMAND_INTX_DISABLE; + goto out; + } + rc = xc_physdev_map_pirq(xen_xc, xen_domid, machine_irq, &pirq); if (rc < 0) { error_setg_errno(errp, errno, "Mapping machine irq %u to" diff --git a/hw/xen/xen_pt_config_init.c b/hw/xen/xen_pt_config_init.c index 47f9010c75..31ec5add1d 100644 --- a/hw/xen/xen_pt_config_init.c +++ b/hw/xen/xen_pt_config_init.c @@ -15,7 +15,7 @@ #include "qemu/osdep.h" #include "qapi/error.h" #include "qemu/timer.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "xen_pt.h" #define XEN_PT_MERGE_VALUE(value, data, val_mask) \ @@ -300,7 +300,9 @@ static int xen_pt_irqpin_reg_init(XenPCIPassthroughState *s, XenPTRegInfo *reg, uint32_t real_offset, uint32_t *data) { - *data = xen_pt_pci_read_intx(s); + if (s->real_device.irq) { + *data = xen_pt_pci_read_intx(s); + } return 0; } diff --git a/hw/xen/xen_pt_graphics.c b/hw/xen/xen_pt_graphics.c index 135c8df1e7..b69732729b 100644 --- a/hw/xen/xen_pt_graphics.c +++ b/hw/xen/xen_pt_graphics.c @@ -5,7 +5,7 @@ #include "qapi/error.h" #include "xen_pt.h" #include "xen-host-pci-device.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" static unsigned long igd_guest_opregion; static unsigned long igd_host_opregion; @@ -185,8 +185,19 @@ void xen_pt_setup_vga(XenPCIPassthroughState *s, XenHostPCIDevice *dev, return; } + if (bios_size < sizeof(struct rom_header)) { + error_setg(errp, "VGA: VBIOS image corrupt (too small)"); + return; + } + /* Currently we fixed this address as a primary. */ rom = (struct rom_header *)bios; + + if (rom->pcioffset + sizeof(struct pci_data) > bios_size) { + error_setg(errp, "VGA: VBIOS image corrupt (bad pcioffset field)"); + return; + } + pd = (void *)(bios + (unsigned char)rom->pcioffset); /* We may need to fixup Device Identification. */ @@ -194,6 +205,11 @@ void xen_pt_setup_vga(XenPCIPassthroughState *s, XenHostPCIDevice *dev, pd->device = s->real_device.device_id; len = rom->size * 512; + if (len > bios_size) { + error_setg(errp, "VGA: VBIOS image corrupt (bad size field)"); + return; + } + /* Then adjust the bios checksum */ for (c = (char *)bios; c < ((char *)bios + len); c++) { checksum += *c; diff --git a/hw/xen/xen_pt_msi.c b/hw/xen/xen_pt_msi.c index cc514f9157..fb4b887b92 100644 --- a/hw/xen/xen_pt_msi.c +++ b/hw/xen/xen_pt_msi.c @@ -11,7 +11,7 @@ #include "qemu/osdep.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "xen_pt.h" #include "hw/i386/apic-msidef.h" diff --git a/hw/xen/xen_pvdev.c b/hw/xen/xen_pvdev.c index f026556f62..6ef09cbf9d 100644 --- a/hw/xen/xen_pvdev.c +++ b/hw/xen/xen_pvdev.c @@ -20,7 +20,7 @@ #include "qemu/osdep.h" #include "qemu/log.h" #include "hw/qdev-core.h" -#include "hw/xen/xen_backend.h" +#include "hw/xen/xen-legacy-backend.h" #include "hw/xen/xen_pvdev.h" /* private */ @@ -34,7 +34,7 @@ struct xs_dirs { static QTAILQ_HEAD(, xs_dirs) xs_cleanup = QTAILQ_HEAD_INITIALIZER(xs_cleanup); -static QTAILQ_HEAD(, XenDevice) xendevs = +static QTAILQ_HEAD(, XenLegacyDevice) xendevs = QTAILQ_HEAD_INITIALIZER(xendevs); /* ------------------------------------------------------------- */ @@ -195,7 +195,7 @@ const char *xenbus_strstate(enum xenbus_state state) * 2 == noisy debug messages (logfile only). * 3 == will flood your log (logfile only). */ -void xen_pv_printf(struct XenDevice *xendev, int msg_level, +void xen_pv_printf(struct XenLegacyDevice *xendev, int msg_level, const char *fmt, ...) { va_list args; @@ -230,7 +230,7 @@ void xen_pv_printf(struct XenDevice *xendev, int msg_level, void xen_pv_evtchn_event(void *opaque) { - struct XenDevice *xendev = opaque; + struct XenLegacyDevice *xendev = opaque; evtchn_port_t port; port = xenevtchn_pending(xendev->evtchndev); @@ -247,7 +247,7 @@ void xen_pv_evtchn_event(void *opaque) } } -void xen_pv_unbind_evtchn(struct XenDevice *xendev) +void xen_pv_unbind_evtchn(struct XenLegacyDevice *xendev) { if (xendev->local_port == -1) { return; @@ -258,16 +258,16 @@ void xen_pv_unbind_evtchn(struct XenDevice *xendev) xendev->local_port = -1; } -int xen_pv_send_notify(struct XenDevice *xendev) +int xen_pv_send_notify(struct XenLegacyDevice *xendev) { return xenevtchn_notify(xendev->evtchndev, xendev->local_port); } /* ------------------------------------------------------------- */ -struct XenDevice *xen_pv_find_xendev(const char *type, int dom, int dev) +struct XenLegacyDevice *xen_pv_find_xendev(const char *type, int dom, int dev) { - struct XenDevice *xendev; + struct XenLegacyDevice *xendev; QTAILQ_FOREACH(xendev, &xendevs, next) { if (xendev->dom != dom) { @@ -287,7 +287,7 @@ struct XenDevice *xen_pv_find_xendev(const char *type, int dom, int dev) /* * release xen backend device. */ -void xen_pv_del_xendev(struct XenDevice *xendev) +void xen_pv_del_xendev(struct XenLegacyDevice *xendev) { if (xendev->ops->free) { xendev->ops->free(xendev); @@ -312,7 +312,7 @@ void xen_pv_del_xendev(struct XenDevice *xendev) qdev_unplug(&xendev->qdev, NULL); } -void xen_pv_insert_xendev(struct XenDevice *xendev) +void xen_pv_insert_xendev(struct XenLegacyDevice *xendev) { QTAILQ_INSERT_TAIL(&xendevs, xendev, next); } diff --git a/hw/xenpv/Makefile.objs b/hw/xenpv/Makefile.objs index bbf5873fd1..8bfa4586ab 100644 --- a/hw/xenpv/Makefile.objs +++ b/hw/xenpv/Makefile.objs @@ -1,4 +1,2 @@ # Xen PV machine support obj-$(CONFIG_XEN) += xen_machine_pv.o -# Xen PV machine builder support -obj-$(CONFIG_XEN_PV_DOMAIN_BUILD) += xen_domainbuild.o diff --git a/hw/xenpv/xen_domainbuild.c b/hw/xenpv/xen_domainbuild.c deleted file mode 100644 index 2859280a6a..0000000000 --- a/hw/xenpv/xen_domainbuild.c +++ /dev/null @@ -1,299 +0,0 @@ -#include "qemu/osdep.h" -#include "qemu/units.h" -#include "hw/xen/xen_backend.h" -#include "xen_domainbuild.h" -#include "qemu/timer.h" -#include "qemu/log.h" - -#include - -static int xenstore_domain_mkdir(char *path) -{ - struct xs_permissions perms_ro[] = {{ - .id = 0, /* set owner: dom0 */ - },{ - .id = xen_domid, - .perms = XS_PERM_READ, - }}; - struct xs_permissions perms_rw[] = {{ - .id = 0, /* set owner: dom0 */ - },{ - .id = xen_domid, - .perms = XS_PERM_READ | XS_PERM_WRITE, - }}; - const char *writable[] = { "device", "control", "error", NULL }; - char subpath[256]; - int i; - - if (!xs_mkdir(xenstore, 0, path)) { - fprintf(stderr, "%s: xs_mkdir %s: failed\n", __func__, path); - return -1; - } - if (!xs_set_permissions(xenstore, 0, path, perms_ro, 2)) { - fprintf(stderr, "%s: xs_set_permissions failed\n", __func__); - return -1; - } - - for (i = 0; writable[i]; i++) { - snprintf(subpath, sizeof(subpath), "%s/%s", path, writable[i]); - if (!xs_mkdir(xenstore, 0, subpath)) { - fprintf(stderr, "%s: xs_mkdir %s: failed\n", __func__, subpath); - return -1; - } - if (!xs_set_permissions(xenstore, 0, subpath, perms_rw, 2)) { - fprintf(stderr, "%s: xs_set_permissions failed\n", __func__); - return -1; - } - } - return 0; -} - -int xenstore_domain_init1(const char *kernel, const char *ramdisk, - const char *cmdline) -{ - char *dom, uuid_string[42], vm[256], path[256]; - int i; - - qemu_uuid_unparse(&qemu_uuid, uuid_string); - dom = xs_get_domain_path(xenstore, xen_domid); - snprintf(vm, sizeof(vm), "/vm/%s", uuid_string); - - xenstore_domain_mkdir(dom); - - xenstore_write_str(vm, "image/ostype", "linux"); - if (kernel) - xenstore_write_str(vm, "image/kernel", kernel); - if (ramdisk) - xenstore_write_str(vm, "image/ramdisk", ramdisk); - if (cmdline) - xenstore_write_str(vm, "image/cmdline", cmdline); - - /* name + id */ - xenstore_write_str(vm, "name", qemu_name ? qemu_name : "no-name"); - xenstore_write_str(vm, "uuid", uuid_string); - xenstore_write_str(dom, "name", qemu_name ? qemu_name : "no-name"); - xenstore_write_int(dom, "domid", xen_domid); - xenstore_write_str(dom, "vm", vm); - - /* memory */ - xenstore_write_int(dom, "memory/target", ram_size / KiB); - xenstore_write_int(vm, "memory", ram_size / MiB); - xenstore_write_int(vm, "maxmem", ram_size / MiB); - - /* cpus */ - for (i = 0; i < smp_cpus; i++) { - snprintf(path, sizeof(path), "cpu/%d/availability",i); - xenstore_write_str(dom, path, "online"); - } - xenstore_write_int(vm, "vcpu_avail", smp_cpus); - xenstore_write_int(vm, "vcpus", smp_cpus); - - /* vnc password */ - xenstore_write_str(vm, "vncpassword", "" /* FIXME */); - - free(dom); - return 0; -} - -int xenstore_domain_init2(int xenstore_port, int xenstore_mfn, - int console_port, int console_mfn) -{ - char *dom; - - dom = xs_get_domain_path(xenstore, xen_domid); - - /* signal new domain */ - xs_introduce_domain(xenstore, - xen_domid, - xenstore_mfn, - xenstore_port); - - /* xenstore */ - xenstore_write_int(dom, "store/ring-ref", xenstore_mfn); - xenstore_write_int(dom, "store/port", xenstore_port); - - /* console */ - xenstore_write_str(dom, "console/type", "ioemu"); - xenstore_write_int(dom, "console/limit", 128 * KiB); - xenstore_write_int(dom, "console/ring-ref", console_mfn); - xenstore_write_int(dom, "console/port", console_port); - xen_config_dev_console(0); - - free(dom); - return 0; -} - -/* ------------------------------------------------------------- */ - -static QEMUTimer *xen_poll; - -/* check domain state once per second */ -static void xen_domain_poll(void *opaque) -{ - struct xc_dominfo info; - int rc; - - rc = xc_domain_getinfo(xen_xc, xen_domid, 1, &info); - if ((rc != 1) || (info.domid != xen_domid)) { - qemu_log("xen: domain %d is gone\n", xen_domid); - goto quit; - } - if (info.dying) { - qemu_log("xen: domain %d is dying (%s%s)\n", xen_domid, - info.crashed ? "crashed" : "", - info.shutdown ? "shutdown" : ""); - goto quit; - } - - timer_mod(xen_poll, qemu_clock_get_ms(QEMU_CLOCK_REALTIME) + 1000); - return; - -quit: - qemu_system_shutdown_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN); -} - -static int xen_domain_watcher(void) -{ - int qemu_running = 1; - int fd[2], i, n, rc; - char byte; - - if (pipe(fd) != 0) { - qemu_log("%s: Huh? pipe error: %s\n", __func__, strerror(errno)); - return -1; - } - if (fork() != 0) - return 0; /* not child */ - - /* close all file handles, except stdio/out/err, - * our watch pipe and the xen interface handle */ - n = getdtablesize(); - for (i = 3; i < n; i++) { - if (i == fd[0]) - continue; - close(i); - } - - /* - * Reopen xc interface, since the original is unsafe after fork - * and was closed above. - */ - xen_xc = xc_interface_open(0, 0, 0); - - /* ignore term signals */ - signal(SIGINT, SIG_IGN); - signal(SIGTERM, SIG_IGN); - - /* wait for qemu exiting */ - while (qemu_running) { - rc = read(fd[0], &byte, 1); - switch (rc) { - case -1: - if (errno == EINTR) - continue; - qemu_log("%s: Huh? read error: %s\n", __func__, strerror(errno)); - qemu_running = 0; - break; - case 0: - /* EOF -> qemu exited */ - qemu_running = 0; - break; - default: - qemu_log("%s: Huh? data on the watch pipe?\n", __func__); - break; - } - } - - /* cleanup */ - qemu_log("%s: destroy domain %d\n", __func__, xen_domid); - xc_domain_destroy(xen_xc, xen_domid); - _exit(0); -} - -/* normal cleanup */ -static void xen_domain_cleanup(void) -{ - char *dom; - - dom = xs_get_domain_path(xenstore, xen_domid); - if (dom) { - xs_rm(xenstore, 0, dom); - free(dom); - } - xs_release_domain(xenstore, xen_domid); -} - -int xen_domain_build_pv(const char *kernel, const char *ramdisk, - const char *cmdline) -{ - uint32_t ssidref = 0; - uint32_t flags = 0; - xen_domain_handle_t uuid; - unsigned int xenstore_port = 0, console_port = 0; - unsigned long xenstore_mfn = 0, console_mfn = 0; - int rc; - - memcpy(uuid, &qemu_uuid, sizeof(uuid)); - rc = xen_domain_create(xen_xc, ssidref, uuid, flags, &xen_domid); - if (rc < 0) { - fprintf(stderr, "xen: xc_domain_create() failed\n"); - goto err; - } - qemu_log("xen: created domain %d\n", xen_domid); - atexit(xen_domain_cleanup); - if (xen_domain_watcher() == -1) { - goto err; - } - - xenstore_domain_init1(kernel, ramdisk, cmdline); - - rc = xc_domain_max_vcpus(xen_xc, xen_domid, smp_cpus); - if (rc < 0) { - fprintf(stderr, "xen: xc_domain_max_vcpus() failed\n"); - goto err; - } - -#if 0 - rc = xc_domain_setcpuweight(xen_xc, xen_domid, 256); - if (rc < 0) { - fprintf(stderr, "xen: xc_domain_setcpuweight() failed\n"); - goto err; - } -#endif - - rc = xc_domain_setmaxmem(xen_xc, xen_domid, ram_size / KiB); - if (rc < 0) { - fprintf(stderr, "xen: xc_domain_setmaxmem() failed\n"); - goto err; - } - - xenstore_port = xc_evtchn_alloc_unbound(xen_xc, xen_domid, 0); - console_port = xc_evtchn_alloc_unbound(xen_xc, xen_domid, 0); - - rc = xc_linux_build(xen_xc, xen_domid, ram_size / MiB, - kernel, ramdisk, cmdline, - 0, flags, - xenstore_port, &xenstore_mfn, - console_port, &console_mfn); - if (rc < 0) { - fprintf(stderr, "xen: xc_linux_build() failed\n"); - goto err; - } - - xenstore_domain_init2(xenstore_port, xenstore_mfn, - console_port, console_mfn); - - qemu_log("xen: unpausing domain %d\n", xen_domid); - rc = xc_domain_unpause(xen_xc, xen_domid); - if (rc < 0) { - fprintf(stderr, "xen: xc_domain_unpause() failed\n"); - goto err; - } - - xen_poll = timer_new_ms(QEMU_CLOCK_REALTIME, xen_domain_poll, NULL); - timer_mod(xen_poll, qemu_clock_get_ms(QEMU_CLOCK_REALTIME) + 1000); - return 0; - -err: - return -1; -} diff --git a/hw/xenpv/xen_domainbuild.h b/hw/xenpv/xen_domainbuild.h deleted file mode 100644 index 652d9b410f..0000000000 --- a/hw/xenpv/xen_domainbuild.h +++ /dev/null @@ -1,13 +0,0 @@ -#ifndef QEMU_HW_XEN_DOMAINBUILD_H -#define QEMU_HW_XEN_DOMAINBUILD_H - -#include "hw/xen/xen_common.h" - -int xenstore_domain_init1(const char *kernel, const char *ramdisk, - const char *cmdline); -int xenstore_domain_init2(int xenstore_port, int xenstore_mfn, - int console_port, int console_mfn); -int xen_domain_build_pv(const char *kernel, const char *ramdisk, - const char *cmdline); - -#endif /* QEMU_HW_XEN_DOMAINBUILD_H */ diff --git a/hw/xenpv/xen_machine_pv.c b/hw/xenpv/xen_machine_pv.c index 44d67b87c4..dcaf2a01a3 100644 --- a/hw/xenpv/xen_machine_pv.c +++ b/hw/xenpv/xen_machine_pv.c @@ -26,8 +26,8 @@ #include "qemu/error-report.h" #include "hw/hw.h" #include "hw/boards.h" -#include "hw/xen/xen_backend.h" -#include "xen_domainbuild.h" +#include "hw/xen/xen-legacy-backend.h" +#include "hw/xen/xen-bus.h" #include "sysemu/block-backend.h" static void xen_init_pv(MachineState *machine) @@ -43,21 +43,8 @@ static void xen_init_pv(MachineState *machine) switch (xen_mode) { case XEN_ATTACH: - /* nothing to do, xend handles everything */ + /* nothing to do, libxl handles everything */ break; -#ifdef CONFIG_XEN_PV_DOMAIN_BUILD - case XEN_CREATE: { - const char *kernel_filename = machine->kernel_filename; - const char *kernel_cmdline = machine->kernel_cmdline; - const char *initrd_filename = machine->initrd_filename; - if (xen_domain_build_pv(kernel_filename, initrd_filename, - kernel_cmdline) < 0) { - error_report("xen pv domain creation failed"); - exit(1); - } - break; - } -#endif case XEN_EMULATE: error_report("xen emulation not implemented (yet)"); exit(1); @@ -93,6 +80,8 @@ static void xen_init_pv(MachineState *machine) xen_config_dev_nic(nd_table + i); } + xen_bus_init(); + /* config cleanup hook */ atexit(xen_config_cleanup); } diff --git a/include/hw/xen/xen-backend.h b/include/hw/xen/xen-backend.h new file mode 100644 index 0000000000..010d712638 --- /dev/null +++ b/include/hw/xen/xen-backend.h @@ -0,0 +1,39 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef HW_XEN_BACKEND_H +#define HW_XEN_BACKEND_H + +#include "hw/xen/xen-bus.h" + +typedef struct XenBackendInstance XenBackendInstance; + +typedef void (*XenBackendDeviceCreate)(XenBackendInstance *backend, + QDict *opts, Error **errp); +typedef void (*XenBackendDeviceDestroy)(XenBackendInstance *backend, + Error **errp); + +typedef struct XenBackendInfo { + const char *type; + XenBackendDeviceCreate create; + XenBackendDeviceDestroy destroy; +} XenBackendInfo; + +XenBus *xen_backend_get_bus(XenBackendInstance *backend); +const char *xen_backend_get_name(XenBackendInstance *backend); + +void xen_backend_set_device(XenBackendInstance *backend, + XenDevice *xendevice); +XenDevice *xen_backend_get_device(XenBackendInstance *backend); + +void xen_backend_register(const XenBackendInfo *info); + +void xen_backend_device_create(XenBus *xenbus, const char *type, + const char *name, QDict *opts, Error **errp); +bool xen_backend_try_device_destroy(XenDevice *xendev, Error **errp); + +#endif /* HW_XEN_BACKEND_H */ diff --git a/include/hw/xen/xen-block.h b/include/hw/xen/xen-block.h new file mode 100644 index 0000000000..11d351b4b3 --- /dev/null +++ b/include/hw/xen/xen-block.h @@ -0,0 +1,94 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef HW_XEN_BLOCK_H +#define HW_XEN_BLOCK_H + +#include "hw/xen/xen-bus.h" +#include "hw/block/block.h" +#include "hw/block/dataplane/xen-block.h" +#include "sysemu/iothread.h" + +typedef enum XenBlockVdevType { + XEN_BLOCK_VDEV_TYPE_INVALID, + XEN_BLOCK_VDEV_TYPE_DP, + XEN_BLOCK_VDEV_TYPE_XVD, + XEN_BLOCK_VDEV_TYPE_HD, + XEN_BLOCK_VDEV_TYPE_SD, + XEN_BLOCK_VDEV_TYPE__MAX +} XenBlockVdevType; + +typedef struct XenBlockVdev { + XenBlockVdevType type; + unsigned long disk; + unsigned long partition; + unsigned long number; +} XenBlockVdev; + + +typedef struct XenBlockProperties { + XenBlockVdev vdev; + BlockConf conf; + unsigned int max_ring_page_order; + IOThread *iothread; +} XenBlockProperties; + +typedef struct XenBlockDrive { + char *id; + char *node_name; +} XenBlockDrive; + +typedef struct XenBlockIOThread { + char *id; +} XenBlockIOThread; + +typedef struct XenBlockDevice { + XenDevice xendev; + XenBlockProperties props; + const char *device_type; + unsigned int info; + XenBlockDataPlane *dataplane; + XenBlockDrive *drive; + XenBlockIOThread *iothread; +} XenBlockDevice; + +typedef void (*XenBlockDeviceRealize)(XenBlockDevice *blockdev, Error **errp); +typedef void (*XenBlockDeviceUnrealize)(XenBlockDevice *blockdev, Error **errp); + +typedef struct XenBlockDeviceClass { + /*< private >*/ + XenDeviceClass parent_class; + /*< public >*/ + XenBlockDeviceRealize realize; + XenBlockDeviceUnrealize unrealize; +} XenBlockDeviceClass; + +#define TYPE_XEN_BLOCK_DEVICE "xen-block" +#define XEN_BLOCK_DEVICE(obj) \ + OBJECT_CHECK(XenBlockDevice, (obj), TYPE_XEN_BLOCK_DEVICE) +#define XEN_BLOCK_DEVICE_CLASS(class) \ + OBJECT_CLASS_CHECK(XenBlockDeviceClass, (class), TYPE_XEN_BLOCK_DEVICE) +#define XEN_BLOCK_DEVICE_GET_CLASS(obj) \ + OBJECT_GET_CLASS(XenBlockDeviceClass, (obj), TYPE_XEN_BLOCK_DEVICE) + +typedef struct XenDiskDevice { + XenBlockDevice blockdev; +} XenDiskDevice; + +#define TYPE_XEN_DISK_DEVICE "xen-disk" +#define XEN_DISK_DEVICE(obj) \ + OBJECT_CHECK(XenDiskDevice, (obj), TYPE_XEN_DISK_DEVICE) + +typedef struct XenCDRomDevice { + XenBlockDevice blockdev; +} XenCDRomDevice; + +#define TYPE_XEN_CDROM_DEVICE "xen-cdrom" +#define XEN_CDROM_DEVICE(obj) \ + OBJECT_CHECK(XenCDRomDevice, (obj), TYPE_XEN_CDROM_DEVICE) + +#endif /* HW_XEN_BLOCK_H */ diff --git a/include/hw/xen/xen-bus-helper.h b/include/hw/xen/xen-bus-helper.h new file mode 100644 index 0000000000..4c0f747445 --- /dev/null +++ b/include/hw/xen/xen-bus-helper.h @@ -0,0 +1,45 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef HW_XEN_BUS_HELPER_H +#define HW_XEN_BUS_HELPER_H + +#include "hw/xen/xen_common.h" + +const char *xs_strstate(enum xenbus_state state); + +void xs_node_create(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, struct xs_permissions perms[], + unsigned int nr_perms, Error **errp); +void xs_node_destroy(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, Error **errp); + +/* Write to node/key unless node is empty, in which case write to key */ +void xs_node_vprintf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, va_list ap) + GCC_FMT_ATTR(6, 0); +void xs_node_printf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, ...) + GCC_FMT_ATTR(6, 7); + +/* Read from node/key unless node is empty, in which case read from key */ +int xs_node_vscanf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, va_list ap); +int xs_node_scanf(struct xs_handle *xsh, xs_transaction_t tid, + const char *node, const char *key, Error **errp, + const char *fmt, ...); + +/* Watch node/key unless node is empty, in which case watch key */ +void xs_node_watch(struct xs_handle *xsh, const char *node, const char *key, + char *token, Error **errp); +void xs_node_unwatch(struct xs_handle *xsh, const char *node, const char *key, + const char *token, Error **errp); + +#endif /* HW_XEN_BUS_HELPER_H */ diff --git a/include/hw/xen/xen-bus.h b/include/hw/xen/xen-bus.h new file mode 100644 index 0000000000..3183f10e3c --- /dev/null +++ b/include/hw/xen/xen-bus.h @@ -0,0 +1,137 @@ +/* + * Copyright (c) 2018 Citrix Systems Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef HW_XEN_BUS_H +#define HW_XEN_BUS_H + +#include "hw/xen/xen_common.h" +#include "hw/sysbus.h" +#include "qemu/notify.h" + +typedef void (*XenWatchHandler)(void *opaque); + +typedef struct XenWatch XenWatch; + +typedef struct XenDevice { + DeviceState qdev; + domid_t frontend_id; + char *name; + char *backend_path, *frontend_path; + enum xenbus_state backend_state, frontend_state; + Notifier exit; + XenWatch *backend_state_watch, *frontend_state_watch; + bool backend_online; + XenWatch *backend_online_watch; + xengnttab_handle *xgth; + bool feature_grant_copy; + xenevtchn_handle *xeh; + NotifierList event_notifiers; +} XenDevice; + +typedef char *(*XenDeviceGetName)(XenDevice *xendev, Error **errp); +typedef void (*XenDeviceRealize)(XenDevice *xendev, Error **errp); +typedef void (*XenDeviceFrontendChanged)(XenDevice *xendev, + enum xenbus_state frontend_state, + Error **errp); +typedef void (*XenDeviceUnrealize)(XenDevice *xendev, Error **errp); + +typedef struct XenDeviceClass { + /*< private >*/ + DeviceClass parent_class; + /*< public >*/ + const char *backend; + const char *device; + XenDeviceGetName get_name; + XenDeviceRealize realize; + XenDeviceFrontendChanged frontend_changed; + XenDeviceUnrealize unrealize; +} XenDeviceClass; + +#define TYPE_XEN_DEVICE "xen-device" +#define XEN_DEVICE(obj) \ + OBJECT_CHECK(XenDevice, (obj), TYPE_XEN_DEVICE) +#define XEN_DEVICE_CLASS(class) \ + OBJECT_CLASS_CHECK(XenDeviceClass, (class), TYPE_XEN_DEVICE) +#define XEN_DEVICE_GET_CLASS(obj) \ + OBJECT_GET_CLASS(XenDeviceClass, (obj), TYPE_XEN_DEVICE) + +typedef struct XenBus { + BusState qbus; + domid_t backend_id; + struct xs_handle *xsh; + NotifierList watch_notifiers; + XenWatch *backend_watch; +} XenBus; + +typedef struct XenBusClass { + /*< private >*/ + BusClass parent_class; +} XenBusClass; + +#define TYPE_XEN_BUS "xen-bus" +#define XEN_BUS(obj) \ + OBJECT_CHECK(XenBus, (obj), TYPE_XEN_BUS) +#define XEN_BUS_CLASS(class) \ + OBJECT_CLASS_CHECK(XenBusClass, (class), TYPE_XEN_BUS) +#define XEN_BUS_GET_CLASS(obj) \ + OBJECT_GET_CLASS(XenBusClass, (obj), TYPE_XEN_BUS) + +void xen_bus_init(void); + +void xen_device_backend_set_state(XenDevice *xendev, + enum xenbus_state state); +enum xenbus_state xen_device_backend_get_state(XenDevice *xendev); + +void xen_device_backend_printf(XenDevice *xendev, const char *key, + const char *fmt, ...) + GCC_FMT_ATTR(3, 4); +void xen_device_frontend_printf(XenDevice *xendev, const char *key, + const char *fmt, ...) + GCC_FMT_ATTR(3, 4); + +int xen_device_frontend_scanf(XenDevice *xendev, const char *key, + const char *fmt, ...); + +void xen_device_set_max_grant_refs(XenDevice *xendev, unsigned int nr_refs, + Error **errp); +void *xen_device_map_grant_refs(XenDevice *xendev, uint32_t *refs, + unsigned int nr_refs, int prot, + Error **errp); +void xen_device_unmap_grant_refs(XenDevice *xendev, void *map, + unsigned int nr_refs, Error **errp); + +typedef struct XenDeviceGrantCopySegment { + union { + void *virt; + struct { + uint32_t ref; + off_t offset; + } foreign; + } source, dest; + size_t len; +} XenDeviceGrantCopySegment; + +void xen_device_copy_grant_refs(XenDevice *xendev, bool to_domain, + XenDeviceGrantCopySegment segs[], + unsigned int nr_segs, Error **errp); + +typedef struct XenEventChannel XenEventChannel; + +typedef void (*XenEventHandler)(void *opaque); + +XenEventChannel *xen_device_bind_event_channel(XenDevice *xendev, + unsigned int port, + XenEventHandler handler, + void *opaque, Error **errp); +void xen_device_notify_event_channel(XenDevice *xendev, + XenEventChannel *channel, + Error **errp); +void xen_device_unbind_event_channel(XenDevice *xendev, + XenEventChannel *channel, + Error **errp); + +#endif /* HW_XEN_BUS_H */ diff --git a/include/hw/xen/xen_backend.h b/include/hw/xen/xen-legacy-backend.h similarity index 61% rename from include/hw/xen/xen_backend.h rename to include/hw/xen/xen-legacy-backend.h index 9c17fdd85d..20cb47b5bf 100644 --- a/include/hw/xen/xen_backend.h +++ b/include/hw/xen/xen-legacy-backend.h @@ -11,7 +11,7 @@ #define TYPE_XENBACKEND "xen-backend" #define XENBACKEND_DEVICE(obj) \ - OBJECT_CHECK(XenDevice, (obj), TYPE_XENBACKEND) + OBJECT_CHECK(XenLegacyDevice, (obj), TYPE_XENBACKEND) /* variables */ extern struct xs_handle *xenstore; @@ -20,32 +20,37 @@ extern DeviceState *xen_sysdev; extern BusState *xen_sysbus; int xenstore_mkdir(char *path, int p); -int xenstore_write_be_str(struct XenDevice *xendev, const char *node, const char *val); -int xenstore_write_be_int(struct XenDevice *xendev, const char *node, int ival); -int xenstore_write_be_int64(struct XenDevice *xendev, const char *node, int64_t ival); -char *xenstore_read_be_str(struct XenDevice *xendev, const char *node); -int xenstore_read_be_int(struct XenDevice *xendev, const char *node, int *ival); -void xenstore_update_fe(char *watch, struct XenDevice *xendev); +int xenstore_write_be_str(struct XenLegacyDevice *xendev, const char *node, + const char *val); +int xenstore_write_be_int(struct XenLegacyDevice *xendev, const char *node, + int ival); +int xenstore_write_be_int64(struct XenLegacyDevice *xendev, const char *node, + int64_t ival); +char *xenstore_read_be_str(struct XenLegacyDevice *xendev, const char *node); +int xenstore_read_be_int(struct XenLegacyDevice *xendev, const char *node, + int *ival); +void xenstore_update_fe(char *watch, struct XenLegacyDevice *xendev); void xenstore_update_be(char *watch, char *type, int dom, struct XenDevOps *ops); -char *xenstore_read_fe_str(struct XenDevice *xendev, const char *node); -int xenstore_read_fe_int(struct XenDevice *xendev, const char *node, int *ival); -int xenstore_read_fe_uint64(struct XenDevice *xendev, const char *node, +char *xenstore_read_fe_str(struct XenLegacyDevice *xendev, const char *node); +int xenstore_read_fe_int(struct XenLegacyDevice *xendev, const char *node, + int *ival); +int xenstore_read_fe_uint64(struct XenLegacyDevice *xendev, const char *node, uint64_t *uval); -void xen_be_check_state(struct XenDevice *xendev); +void xen_be_check_state(struct XenLegacyDevice *xendev); /* xen backend driver bits */ int xen_be_init(void); void xen_be_register_common(void); int xen_be_register(const char *type, struct XenDevOps *ops); -int xen_be_set_state(struct XenDevice *xendev, enum xenbus_state state); -int xen_be_bind_evtchn(struct XenDevice *xendev); -void xen_be_set_max_grant_refs(struct XenDevice *xendev, +int xen_be_set_state(struct XenLegacyDevice *xendev, enum xenbus_state state); +int xen_be_bind_evtchn(struct XenLegacyDevice *xendev); +void xen_be_set_max_grant_refs(struct XenLegacyDevice *xendev, unsigned int nr_refs); -void *xen_be_map_grant_refs(struct XenDevice *xendev, uint32_t *refs, +void *xen_be_map_grant_refs(struct XenLegacyDevice *xendev, uint32_t *refs, unsigned int nr_refs, int prot); -void xen_be_unmap_grant_refs(struct XenDevice *xendev, void *ptr, +void xen_be_unmap_grant_refs(struct XenLegacyDevice *xendev, void *ptr, unsigned int nr_refs); typedef struct XenGrantCopySegment { @@ -59,17 +64,17 @@ typedef struct XenGrantCopySegment { size_t len; } XenGrantCopySegment; -int xen_be_copy_grant_refs(struct XenDevice *xendev, +int xen_be_copy_grant_refs(struct XenLegacyDevice *xendev, bool to_domain, XenGrantCopySegment segs[], unsigned int nr_segs); -static inline void *xen_be_map_grant_ref(struct XenDevice *xendev, +static inline void *xen_be_map_grant_ref(struct XenLegacyDevice *xendev, uint32_t ref, int prot) { return xen_be_map_grant_refs(xendev, &ref, 1, prot); } -static inline void xen_be_unmap_grant_ref(struct XenDevice *xendev, +static inline void xen_be_unmap_grant_ref(struct XenLegacyDevice *xendev, void *ptr) { return xen_be_unmap_grant_refs(xendev, ptr, 1); diff --git a/include/hw/xen/xen.h b/include/hw/xen/xen.h index 7efcdaa8fe..ba039c146d 100644 --- a/include/hw/xen/xen.h +++ b/include/hw/xen/xen.h @@ -15,8 +15,7 @@ /* xen-machine.c */ enum xen_mode { XEN_EMULATE = 0, // xen emulation, using xenner (default) - XEN_CREATE, // create xen domain - XEN_ATTACH // attach to xen domain created by xend + XEN_ATTACH // attach to xen domain created by libxl }; extern uint32_t xen_domid; diff --git a/include/hw/xen/xen_common.h b/include/hw/xen/xen_common.h index 93f631e5bf..9a8155e172 100644 --- a/include/hw/xen/xen_common.h +++ b/include/hw/xen/xen_common.h @@ -32,6 +32,7 @@ extern xc_interface *xen_xc; typedef xc_interface xenforeignmemory_handle; typedef xc_evtchn xenevtchn_handle; typedef xc_gnttab xengnttab_handle; +typedef evtchn_port_or_error_t xenevtchn_port_or_error_t; #define xenevtchn_open(l, f) xc_evtchn_open(l, f); #define xenevtchn_close(h) xc_evtchn_close(h) @@ -661,24 +662,6 @@ static inline int xen_set_ioreq_server_state(domid_t dom, #endif -#ifdef CONFIG_XEN_PV_DOMAIN_BUILD -#if CONFIG_XEN_CTRL_INTERFACE_VERSION < 40700 -static inline int xen_domain_create(xc_interface *xc, uint32_t ssidref, - xen_domain_handle_t handle, uint32_t flags, - uint32_t *pdomid) -{ - return xc_domain_create(xc, ssidref, handle, flags, pdomid); -} -#else -static inline int xen_domain_create(xc_interface *xc, uint32_t ssidref, - xen_domain_handle_t handle, uint32_t flags, - uint32_t *pdomid) -{ - return xc_domain_create(xc, ssidref, handle, flags, pdomid, NULL); -} -#endif -#endif - /* Xen before 4.8 */ #if CONFIG_XEN_CTRL_INTERFACE_VERSION < 40800 diff --git a/include/hw/xen/xen_pvdev.h b/include/hw/xen/xen_pvdev.h index d473e9b34d..83e5174d90 100644 --- a/include/hw/xen/xen_pvdev.h +++ b/include/hw/xen/xen_pvdev.h @@ -6,7 +6,7 @@ #define XEN_BUFSIZE 1024 -struct XenDevice; +struct XenLegacyDevice; /* driver uses grant tables -> open gntdev device (xendev->gnttabdev) */ #define DEVOPS_FLAG_NEED_GNTDEV 1 @@ -16,19 +16,21 @@ struct XenDevice; struct XenDevOps { size_t size; uint32_t flags; - void (*alloc)(struct XenDevice *xendev); - int (*init)(struct XenDevice *xendev); - int (*initialise)(struct XenDevice *xendev); - void (*connected)(struct XenDevice *xendev); - void (*event)(struct XenDevice *xendev); - void (*disconnect)(struct XenDevice *xendev); - int (*free)(struct XenDevice *xendev); - void (*backend_changed)(struct XenDevice *xendev, const char *node); - void (*frontend_changed)(struct XenDevice *xendev, const char *node); + void (*alloc)(struct XenLegacyDevice *xendev); + int (*init)(struct XenLegacyDevice *xendev); + int (*initialise)(struct XenLegacyDevice *xendev); + void (*connected)(struct XenLegacyDevice *xendev); + void (*event)(struct XenLegacyDevice *xendev); + void (*disconnect)(struct XenLegacyDevice *xendev); + int (*free)(struct XenLegacyDevice *xendev); + void (*backend_changed)(struct XenLegacyDevice *xendev, + const char *node); + void (*frontend_changed)(struct XenLegacyDevice *xendev, + const char *node); int (*backend_register)(void); }; -struct XenDevice { +struct XenLegacyDevice { DeviceState qdev; const char *type; int dom; @@ -49,7 +51,7 @@ struct XenDevice { xengnttab_handle *gnttabdev; struct XenDevOps *ops; - QTAILQ_ENTRY(XenDevice) next; + QTAILQ_ENTRY(XenLegacyDevice) next; }; /* ------------------------------------------------------------- */ @@ -66,14 +68,14 @@ void xenstore_update(void *unused); const char *xenbus_strstate(enum xenbus_state state); void xen_pv_evtchn_event(void *opaque); -void xen_pv_insert_xendev(struct XenDevice *xendev); -void xen_pv_del_xendev(struct XenDevice *xendev); -struct XenDevice *xen_pv_find_xendev(const char *type, int dom, int dev); +void xen_pv_insert_xendev(struct XenLegacyDevice *xendev); +void xen_pv_del_xendev(struct XenLegacyDevice *xendev); +struct XenLegacyDevice *xen_pv_find_xendev(const char *type, int dom, int dev); -void xen_pv_unbind_evtchn(struct XenDevice *xendev); -int xen_pv_send_notify(struct XenDevice *xendev); +void xen_pv_unbind_evtchn(struct XenLegacyDevice *xendev); +int xen_pv_send_notify(struct XenLegacyDevice *xendev); -void xen_pv_printf(struct XenDevice *xendev, int msg_level, +void xen_pv_printf(struct XenLegacyDevice *xendev, int msg_level, const char *fmt, ...) GCC_FMT_ATTR(3, 4); #endif /* QEMU_HW_XEN_PVDEV_H */ diff --git a/include/qemu/module.h b/include/qemu/module.h index 54300ab6e5..55dd2beea8 100644 --- a/include/qemu/module.h +++ b/include/qemu/module.h @@ -44,6 +44,7 @@ typedef enum { MODULE_INIT_OPTS, MODULE_INIT_QOM, MODULE_INIT_TRACE, + MODULE_INIT_XEN_BACKEND, MODULE_INIT_MAX } module_init_type; @@ -51,6 +52,8 @@ typedef enum { #define opts_init(function) module_init(function, MODULE_INIT_OPTS) #define type_init(function) module_init(function, MODULE_INIT_QOM) #define trace_init(function) module_init(function, MODULE_INIT_TRACE) +#define xen_backend_init(function) module_init(function, \ + MODULE_INIT_XEN_BACKEND) #define block_module_load_one(lib) module_load_one("block-", lib) #define ui_module_load_one(lib) module_load_one("ui-", lib) diff --git a/qemu-options.hx b/qemu-options.hx index d4f3564b78..521511ec13 100644 --- a/qemu-options.hx +++ b/qemu-options.hx @@ -3394,13 +3394,9 @@ ETEXI DEF("xen-domid", HAS_ARG, QEMU_OPTION_xen_domid, "-xen-domid id specify xen guest domain id\n", QEMU_ARCH_ALL) -DEF("xen-create", 0, QEMU_OPTION_xen_create, - "-xen-create create domain using xen hypercalls, bypassing xend\n" - " warning: should not be used when xend is in use\n", - QEMU_ARCH_ALL) DEF("xen-attach", 0, QEMU_OPTION_xen_attach, "-xen-attach attach to existing xen domain\n" - " xend will use this when starting QEMU\n", + " libxl will use this when starting QEMU\n", QEMU_ARCH_ALL) DEF("xen-domid-restrict", 0, QEMU_OPTION_xen_domid_restrict, "-xen-domid-restrict restrict set of available xen operations\n" @@ -3411,14 +3407,10 @@ STEXI @item -xen-domid @var{id} @findex -xen-domid Specify xen guest domain @var{id} (XEN only). -@item -xen-create -@findex -xen-create -Create domain using xen hypercalls, bypassing xend. -Warning: should not be used when xend is in use (XEN only). @item -xen-attach @findex -xen-attach Attach to existing xen domain. -xend will use this when starting QEMU (XEN only). +libxl will use this when starting QEMU (XEN only). @findex -xen-domid-restrict Restrict set of available xen operations to specified domain id (XEN only). ETEXI diff --git a/vl.c b/vl.c index 9b8ea3f654..bc9fbec654 100644 --- a/vl.c +++ b/vl.c @@ -3856,13 +3856,6 @@ int main(int argc, char **argv, char **envp) } xen_domid = atoi(optarg); break; - case QEMU_OPTION_xen_create: - if (!(xen_available())) { - error_report("Option not supported for this target"); - exit(1); - } - xen_mode = XEN_CREATE; - break; case QEMU_OPTION_xen_attach: if (!(xen_available())) { error_report("Option not supported for this target");