diff options
Diffstat (limited to '')
-rw-r--r-- | src/pulsecore/pstream.c | 1290 |
1 files changed, 1290 insertions, 0 deletions
diff --git a/src/pulsecore/pstream.c b/src/pulsecore/pstream.c new file mode 100644 index 0000000..eb70508 --- /dev/null +++ b/src/pulsecore/pstream.c @@ -0,0 +1,1290 @@ +/*** + This file is part of PulseAudio. + + Copyright 2004-2006 Lennart Poettering + Copyright 2006 Pierre Ossman <ossman@cendio.se> for Cendio AB + + PulseAudio is free software; you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as + published by the Free Software Foundation; either version 2.1 of the + License, or (at your option) any later version. + + PulseAudio 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 + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with PulseAudio; if not, see <http://www.gnu.org/licenses/>. +***/ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> + +#ifdef HAVE_NETINET_IN_H +#include <netinet/in.h> +#endif + +#include <pulse/xmalloc.h> + +#include <pulsecore/idxset.h> +#include <pulsecore/socket.h> +#include <pulsecore/queue.h> +#include <pulsecore/log.h> +#include <pulsecore/creds.h> +#include <pulsecore/refcnt.h> +#include <pulsecore/flist.h> +#include <pulsecore/macro.h> + +#include "pstream.h" + +/* We piggyback information if audio data blocks are stored in SHM on the seek mode */ +#define PA_FLAG_SHMDATA 0x80000000LU +#define PA_FLAG_SHMDATA_MEMFD_BLOCK 0x20000000LU +#define PA_FLAG_SHMRELEASE 0x40000000LU +#define PA_FLAG_SHMREVOKE 0xC0000000LU +#define PA_FLAG_SHMMASK 0xFF000000LU +#define PA_FLAG_SEEKMASK 0x000000FFLU +#define PA_FLAG_SHMWRITABLE 0x00800000LU + +/* The sequence descriptor header consists of 5 32bit integers: */ +enum { + PA_PSTREAM_DESCRIPTOR_LENGTH, + PA_PSTREAM_DESCRIPTOR_CHANNEL, + PA_PSTREAM_DESCRIPTOR_OFFSET_HI, + PA_PSTREAM_DESCRIPTOR_OFFSET_LO, + PA_PSTREAM_DESCRIPTOR_FLAGS, + PA_PSTREAM_DESCRIPTOR_MAX +}; + +/* If we have an SHM block, this info follows the descriptor */ +enum { + PA_PSTREAM_SHM_BLOCKID, + PA_PSTREAM_SHM_SHMID, + PA_PSTREAM_SHM_INDEX, + PA_PSTREAM_SHM_LENGTH, + PA_PSTREAM_SHM_MAX +}; + +typedef uint32_t pa_pstream_descriptor[PA_PSTREAM_DESCRIPTOR_MAX]; + +#define PA_PSTREAM_DESCRIPTOR_SIZE (PA_PSTREAM_DESCRIPTOR_MAX*sizeof(uint32_t)) + +#define MINIBUF_SIZE (256) + +/* To allow uploading a single sample in one frame, this value should be the + * same size (16 MB) as PA_SCACHE_ENTRY_SIZE_MAX from pulsecore/core-scache.h. + */ +#define FRAME_SIZE_MAX_ALLOW (1024*1024*16) + +PA_STATIC_FLIST_DECLARE(items, 0, pa_xfree); + +struct item_info { + enum { + PA_PSTREAM_ITEM_PACKET, + PA_PSTREAM_ITEM_MEMBLOCK, + PA_PSTREAM_ITEM_SHMRELEASE, + PA_PSTREAM_ITEM_SHMREVOKE + } type; + + /* packet info */ + pa_packet *packet; +#ifdef HAVE_CREDS + bool with_ancil_data; + pa_cmsg_ancil_data ancil_data; +#endif + + /* memblock info */ + pa_memchunk chunk; + uint32_t channel; + int64_t offset; + pa_seek_mode_t seek_mode; + + /* release/revoke info */ + uint32_t block_id; +}; + +struct pstream_read { + pa_pstream_descriptor descriptor; + pa_memblock *memblock; + pa_packet *packet; + uint32_t shm_info[PA_PSTREAM_SHM_MAX]; + void *data; + size_t index; +}; + +struct pa_pstream { + PA_REFCNT_DECLARE; + + pa_mainloop_api *mainloop; + pa_defer_event *defer_event; + pa_iochannel *io; + pa_srbchannel *srb, *srbpending; + bool is_srbpending; + + pa_queue *send_queue; + + bool dead; + + struct { + union { + uint8_t minibuf[MINIBUF_SIZE]; + pa_pstream_descriptor descriptor; + }; + struct item_info* current; + void *data; + size_t index; + int minibuf_validsize; + pa_memchunk memchunk; + } write; + + struct pstream_read readio, readsrb; + + /* @use_shm: beside copying the full audio data to the other + * PA end, this pipe supports just sending references of the + * same audio data blocks if they reside in a SHM pool. + * + * @use_memfd: pipe supports sending SHM memfd block references + * + * @registered_memfd_ids: registered memfd pools SHM IDs. Check + * pa_pstream_register_memfd_mempool() for more information. */ + bool use_shm, use_memfd; + pa_idxset *registered_memfd_ids; + + pa_memimport *import; + pa_memexport *export; + + pa_pstream_packet_cb_t receive_packet_callback; + void *receive_packet_callback_userdata; + + pa_pstream_memblock_cb_t receive_memblock_callback; + void *receive_memblock_callback_userdata; + + pa_pstream_notify_cb_t drain_callback; + void *drain_callback_userdata; + + pa_pstream_notify_cb_t die_callback; + void *die_callback_userdata; + + pa_pstream_block_id_cb_t revoke_callback; + void *revoke_callback_userdata; + + pa_pstream_block_id_cb_t release_callback; + void *release_callback_userdata; + + pa_mempool *mempool; + +#ifdef HAVE_CREDS + pa_cmsg_ancil_data read_ancil_data, *write_ancil_data; + bool send_ancil_data_now; +#endif +}; + +#ifdef HAVE_CREDS +/* + * memfd-backed SHM pools blocks transfer occur without passing the pool's + * fd every time, thus minimizing overhead and avoiding fd leaks. A + * REGISTER_MEMFD_SHMID command is sent, with the pool's memfd fd, very early + * on. This command has an ID that uniquely identifies the pool in question. + * Further pool's block references can then be exclusively done using such ID; + * the fd can be safely closed – on both ends – afterwards. + * + * On the sending side of this command, we want to close the passed fds + * directly after being sent. Meanwhile we're only allowed to asynchronously + * schedule packet writes to the pstream, so the job of closing passed fds is + * left to the pstream's actual writing function do_write(): it knows the + * exact point in time where the fds are passed to the other end through + * iochannels and the sendmsg() system call. + * + * Nonetheless not all code paths in the system desire their socket-passed + * fds to be closed after the send. srbchannel needs the passed fds to still + * be open for further communication. System-wide global memfd-backed pools + * also require the passed fd to be open: they pass the same fd, with the same + * ID registration mechanism, for each newly connected client to the system. + * + * So from all of the above, never close the ancillary fds by your own and + * always call below method instead. It takes care of closing the passed fds + * _only if allowed_ by the code paths that originally created them to do so. + * Moreover, it is multiple-invocations safe: failure handlers can, and + * should, call it for passed fds cleanup without worrying too much about + * the system state. + */ +void pa_cmsg_ancil_data_close_fds(struct pa_cmsg_ancil_data *ancil) { + if (ancil && ancil->nfd > 0 && ancil->close_fds_on_cleanup) { + int i; + + pa_assert(ancil->nfd <= MAX_ANCIL_DATA_FDS); + + for (i = 0; i < ancil->nfd; i++) + if (ancil->fds[i] != -1) { + pa_assert_se(pa_close(ancil->fds[i]) == 0); + ancil->fds[i] = -1; + } + + ancil->nfd = 0; + ancil->close_fds_on_cleanup = false; + } +} +#endif + +static int do_write(pa_pstream *p); +static int do_read(pa_pstream *p, struct pstream_read *re); + +static void do_pstream_read_write(pa_pstream *p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + pa_pstream_ref(p); + + p->mainloop->defer_enable(p->defer_event, 0); + + if (!p->dead && p->srb) { + int r = 0; + + if(do_write(p) < 0) + goto fail; + + while (!p->dead && r == 0) { + r = do_read(p, &p->readsrb); + if (r < 0) + goto fail; + } + } + + if (!p->dead && pa_iochannel_is_readable(p->io)) { + if (do_read(p, &p->readio) < 0) + goto fail; + } else if (!p->dead && pa_iochannel_is_hungup(p->io)) + goto fail; + + while (!p->dead && pa_iochannel_is_writable(p->io)) { + int r = do_write(p); + if (r < 0) + goto fail; + if (r == 0) + break; + } + + pa_pstream_unref(p); + return; + +fail: + + if (p->die_callback) + p->die_callback(p, p->die_callback_userdata); + + pa_pstream_unlink(p); + pa_pstream_unref(p); +} + +static bool srb_callback(pa_srbchannel *srb, void *userdata) { + bool b; + pa_pstream *p = userdata; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + pa_assert(p->srb == srb); + + pa_pstream_ref(p); + + do_pstream_read_write(p); + + /* If either pstream or the srb is going away, return false. + We need to check this before p is destroyed. */ + b = (PA_REFCNT_VALUE(p) > 1) && (p->srb == srb); + pa_pstream_unref(p); + + return b; +} + +static void io_callback(pa_iochannel*io, void *userdata) { + pa_pstream *p = userdata; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + pa_assert(p->io == io); + + do_pstream_read_write(p); +} + +static void defer_callback(pa_mainloop_api *m, pa_defer_event *e, void*userdata) { + pa_pstream *p = userdata; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + pa_assert(p->defer_event == e); + pa_assert(p->mainloop == m); + + do_pstream_read_write(p); +} + +static void memimport_release_cb(pa_memimport *i, uint32_t block_id, void *userdata); + +pa_pstream *pa_pstream_new(pa_mainloop_api *m, pa_iochannel *io, pa_mempool *pool) { + pa_pstream *p; + + pa_assert(m); + pa_assert(io); + pa_assert(pool); + + p = pa_xnew0(pa_pstream, 1); + PA_REFCNT_INIT(p); + p->io = io; + pa_iochannel_set_callback(io, io_callback, p); + + p->mainloop = m; + p->defer_event = m->defer_new(m, defer_callback, p); + m->defer_enable(p->defer_event, 0); + + p->send_queue = pa_queue_new(); + + p->mempool = pool; + + /* We do importing unconditionally */ + p->import = pa_memimport_new(p->mempool, memimport_release_cb, p); + + pa_iochannel_socket_set_rcvbuf(io, pa_mempool_block_size_max(p->mempool)); + pa_iochannel_socket_set_sndbuf(io, pa_mempool_block_size_max(p->mempool)); + + return p; +} + +/* Attach memfd<->SHM_ID mapping to given pstream and its memimport. + * Check pa_pstream_register_memfd_mempool() for further info. + * + * Caller owns the passed @memfd_fd and must close it down when appropriate. */ +int pa_pstream_attach_memfd_shmid(pa_pstream *p, unsigned shm_id, int memfd_fd) { + int err = -1; + + pa_assert(memfd_fd != -1); + + if (!p->use_memfd) { + pa_log_warn("Received memfd ID registration request over a pipe " + "that does not support memfds"); + return err; + } + + if (pa_idxset_get_by_data(p->registered_memfd_ids, PA_UINT32_TO_PTR(shm_id), NULL)) { + pa_log_warn("previously registered memfd SHM ID = %u", shm_id); + return err; + } + + if (pa_memimport_attach_memfd(p->import, shm_id, memfd_fd, true)) { + pa_log("Failed to create permanent mapping for memfd region with ID = %u", shm_id); + return err; + } + + pa_assert_se(pa_idxset_put(p->registered_memfd_ids, PA_UINT32_TO_PTR(shm_id), NULL) == 0); + return 0; +} + +static void item_free(void *item) { + struct item_info *i = item; + pa_assert(i); + + if (i->type == PA_PSTREAM_ITEM_MEMBLOCK) { + pa_assert(i->chunk.memblock); + pa_memblock_unref(i->chunk.memblock); + } else if (i->type == PA_PSTREAM_ITEM_PACKET) { + pa_assert(i->packet); + pa_packet_unref(i->packet); + } + +#ifdef HAVE_CREDS + /* On error recovery paths, there might be lingering items + * on the pstream send queue and they are usually freed with + * a call to 'pa_queue_free(p->send_queue, item_free)'. Make + * sure we do not leak any fds in that case! */ + if (i->with_ancil_data) + pa_cmsg_ancil_data_close_fds(&i->ancil_data); +#endif + + if (pa_flist_push(PA_STATIC_FLIST_GET(items), i) < 0) + pa_xfree(i); +} + +static void pstream_free(pa_pstream *p) { + pa_assert(p); + + pa_pstream_unlink(p); + + pa_queue_free(p->send_queue, item_free); + + if (p->write.current) + item_free(p->write.current); + + if (p->write.memchunk.memblock) + pa_memblock_unref(p->write.memchunk.memblock); + + if (p->readsrb.memblock) + pa_memblock_unref(p->readsrb.memblock); + + if (p->readsrb.packet) + pa_packet_unref(p->readsrb.packet); + + if (p->readio.memblock) + pa_memblock_unref(p->readio.memblock); + + if (p->readio.packet) + pa_packet_unref(p->readio.packet); + + if (p->registered_memfd_ids) + pa_idxset_free(p->registered_memfd_ids, NULL); + + pa_xfree(p); +} + +void pa_pstream_send_packet(pa_pstream*p, pa_packet *packet, pa_cmsg_ancil_data *ancil_data) { + struct item_info *i; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + pa_assert(packet); + + if (p->dead) { +#ifdef HAVE_CREDS + pa_cmsg_ancil_data_close_fds(ancil_data); +#endif + return; + } + + if (!(i = pa_flist_pop(PA_STATIC_FLIST_GET(items)))) + i = pa_xnew(struct item_info, 1); + + i->type = PA_PSTREAM_ITEM_PACKET; + i->packet = pa_packet_ref(packet); + +#ifdef HAVE_CREDS + if ((i->with_ancil_data = !!ancil_data)) { + i->ancil_data = *ancil_data; + if (ancil_data->creds_valid) + pa_assert(ancil_data->nfd == 0); + else + pa_assert(ancil_data->nfd > 0); + } +#endif + + pa_queue_push(p->send_queue, i); + + p->mainloop->defer_enable(p->defer_event, 1); +} + +void pa_pstream_send_memblock(pa_pstream*p, uint32_t channel, int64_t offset, pa_seek_mode_t seek_mode, const pa_memchunk *chunk) { + size_t length, idx; + size_t bsm; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + pa_assert(channel != (uint32_t) -1); + pa_assert(chunk); + + if (p->dead) + return; + + idx = 0; + length = chunk->length; + + bsm = pa_mempool_block_size_max(p->mempool); + + while (length > 0) { + struct item_info *i; + size_t n; + + if (!(i = pa_flist_pop(PA_STATIC_FLIST_GET(items)))) + i = pa_xnew(struct item_info, 1); + i->type = PA_PSTREAM_ITEM_MEMBLOCK; + + n = PA_MIN(length, bsm); + i->chunk.index = chunk->index + idx; + i->chunk.length = n; + i->chunk.memblock = pa_memblock_ref(chunk->memblock); + + i->channel = channel; + i->offset = offset; + i->seek_mode = seek_mode; +#ifdef HAVE_CREDS + i->with_ancil_data = false; +#endif + + pa_queue_push(p->send_queue, i); + + idx += n; + length -= n; + } + + p->mainloop->defer_enable(p->defer_event, 1); +} + +void pa_pstream_send_release(pa_pstream *p, uint32_t block_id) { + struct item_info *item; + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (p->dead) + return; + +/* pa_log("Releasing block %u", block_id); */ + + if (!(item = pa_flist_pop(PA_STATIC_FLIST_GET(items)))) + item = pa_xnew(struct item_info, 1); + item->type = PA_PSTREAM_ITEM_SHMRELEASE; + item->block_id = block_id; +#ifdef HAVE_CREDS + item->with_ancil_data = false; +#endif + + pa_queue_push(p->send_queue, item); + p->mainloop->defer_enable(p->defer_event, 1); +} + +/* might be called from thread context */ +static void memimport_release_cb(pa_memimport *i, uint32_t block_id, void *userdata) { + pa_pstream *p = userdata; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (p->dead) + return; + + if (p->release_callback) + p->release_callback(p, block_id, p->release_callback_userdata); + else + pa_pstream_send_release(p, block_id); +} + +void pa_pstream_send_revoke(pa_pstream *p, uint32_t block_id) { + struct item_info *item; + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (p->dead) + return; +/* pa_log("Revoking block %u", block_id); */ + + if (!(item = pa_flist_pop(PA_STATIC_FLIST_GET(items)))) + item = pa_xnew(struct item_info, 1); + item->type = PA_PSTREAM_ITEM_SHMREVOKE; + item->block_id = block_id; +#ifdef HAVE_CREDS + item->with_ancil_data = false; +#endif + + pa_queue_push(p->send_queue, item); + p->mainloop->defer_enable(p->defer_event, 1); +} + +/* might be called from thread context */ +static void memexport_revoke_cb(pa_memexport *e, uint32_t block_id, void *userdata) { + pa_pstream *p = userdata; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (p->revoke_callback) + p->revoke_callback(p, block_id, p->revoke_callback_userdata); + else + pa_pstream_send_revoke(p, block_id); +} + +static void prepare_next_write_item(pa_pstream *p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->write.current = pa_queue_pop(p->send_queue); + + if (!p->write.current) + return; + p->write.index = 0; + p->write.data = NULL; + p->write.minibuf_validsize = 0; + pa_memchunk_reset(&p->write.memchunk); + + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH] = 0; + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_CHANNEL] = htonl((uint32_t) -1); + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI] = 0; + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_LO] = 0; + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS] = 0; + + if (p->write.current->type == PA_PSTREAM_ITEM_PACKET) { + size_t plen; + + pa_assert(p->write.current->packet); + + p->write.data = (void *) pa_packet_data(p->write.current->packet, &plen); + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH] = htonl((uint32_t) plen); + + if (plen <= MINIBUF_SIZE - PA_PSTREAM_DESCRIPTOR_SIZE) { + memcpy(&p->write.minibuf[PA_PSTREAM_DESCRIPTOR_SIZE], p->write.data, plen); + p->write.minibuf_validsize = PA_PSTREAM_DESCRIPTOR_SIZE + plen; + } + + } else if (p->write.current->type == PA_PSTREAM_ITEM_SHMRELEASE) { + + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS] = htonl(PA_FLAG_SHMRELEASE); + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI] = htonl(p->write.current->block_id); + + } else if (p->write.current->type == PA_PSTREAM_ITEM_SHMREVOKE) { + + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS] = htonl(PA_FLAG_SHMREVOKE); + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI] = htonl(p->write.current->block_id); + + } else { + uint32_t flags; + bool send_payload = true; + + pa_assert(p->write.current->type == PA_PSTREAM_ITEM_MEMBLOCK); + pa_assert(p->write.current->chunk.memblock); + + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_CHANNEL] = htonl(p->write.current->channel); + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI] = htonl((uint32_t) (((uint64_t) p->write.current->offset) >> 32)); + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_LO] = htonl((uint32_t) ((uint64_t) p->write.current->offset)); + + flags = (uint32_t) (p->write.current->seek_mode & PA_FLAG_SEEKMASK); + + if (p->use_shm) { + pa_mem_type_t type; + uint32_t block_id, shm_id; + size_t offset, length; + uint32_t *shm_info = (uint32_t *) &p->write.minibuf[PA_PSTREAM_DESCRIPTOR_SIZE]; + size_t shm_size = sizeof(uint32_t) * PA_PSTREAM_SHM_MAX; + pa_mempool *current_pool = pa_memblock_get_pool(p->write.current->chunk.memblock); + pa_memexport *current_export; + + if (p->mempool == current_pool) + pa_assert_se(current_export = p->export); + else + pa_assert_se(current_export = pa_memexport_new(current_pool, memexport_revoke_cb, p)); + + if (pa_memexport_put(current_export, + p->write.current->chunk.memblock, + &type, + &block_id, + &shm_id, + &offset, + &length) >= 0) { + + if (type == PA_MEM_TYPE_SHARED_POSIX) + send_payload = false; + + if (type == PA_MEM_TYPE_SHARED_MEMFD && p->use_memfd) { + if (pa_idxset_get_by_data(p->registered_memfd_ids, PA_UINT32_TO_PTR(shm_id), NULL)) { + flags |= PA_FLAG_SHMDATA_MEMFD_BLOCK; + send_payload = false; + } else { + if (pa_log_ratelimit(PA_LOG_ERROR)) { + pa_log("Cannot send block reference with non-registered memfd ID = %u", shm_id); + pa_log("Fallig back to copying full block data over socket"); + } + } + } + + if (send_payload) { + pa_assert_se(pa_memexport_process_release(current_export, block_id) == 0); + } else { + flags |= PA_FLAG_SHMDATA; + if (pa_mempool_is_remote_writable(current_pool)) + flags |= PA_FLAG_SHMWRITABLE; + + shm_info[PA_PSTREAM_SHM_BLOCKID] = htonl(block_id); + shm_info[PA_PSTREAM_SHM_SHMID] = htonl(shm_id); + shm_info[PA_PSTREAM_SHM_INDEX] = htonl((uint32_t) (offset + p->write.current->chunk.index)); + shm_info[PA_PSTREAM_SHM_LENGTH] = htonl((uint32_t) p->write.current->chunk.length); + + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH] = htonl(shm_size); + p->write.minibuf_validsize = PA_PSTREAM_DESCRIPTOR_SIZE + shm_size; + } + } +/* else */ +/* FIXME: Avoid memexport slot leaks. Call pa_memexport_process_release() */ +/* pa_log_warn("Failed to export memory block."); */ + + if (current_export != p->export) + pa_memexport_free(current_export); + pa_mempool_unref(current_pool); + } + + if (send_payload) { + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH] = htonl((uint32_t) p->write.current->chunk.length); + p->write.memchunk = p->write.current->chunk; + pa_memblock_ref(p->write.memchunk.memblock); + } + + p->write.descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS] = htonl(flags); + } + +#ifdef HAVE_CREDS + if ((p->send_ancil_data_now = p->write.current->with_ancil_data)) + p->write_ancil_data = &p->write.current->ancil_data; +#endif +} + +static void check_srbpending(pa_pstream *p) { + if (!p->is_srbpending) + return; + + if (p->srb) + pa_srbchannel_free(p->srb); + + p->srb = p->srbpending; + p->is_srbpending = false; + + if (p->srb) + pa_srbchannel_set_callback(p->srb, srb_callback, p); +} + +static int do_write(pa_pstream *p) { + void *d; + size_t l; + ssize_t r; + pa_memblock *release_memblock = NULL; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (!p->write.current) + prepare_next_write_item(p); + + if (!p->write.current) { + /* The out queue is empty, so switching channels is safe */ + check_srbpending(p); + return 0; + } + + if (p->write.minibuf_validsize > 0) { + d = p->write.minibuf + p->write.index; + l = p->write.minibuf_validsize - p->write.index; + } else if (p->write.index < PA_PSTREAM_DESCRIPTOR_SIZE) { + d = (uint8_t*) p->write.descriptor + p->write.index; + l = PA_PSTREAM_DESCRIPTOR_SIZE - p->write.index; + } else { + pa_assert(p->write.data || p->write.memchunk.memblock); + + if (p->write.data) + d = p->write.data; + else { + d = pa_memblock_acquire_chunk(&p->write.memchunk); + release_memblock = p->write.memchunk.memblock; + } + + d = (uint8_t*) d + p->write.index - PA_PSTREAM_DESCRIPTOR_SIZE; + l = ntohl(p->write.descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH]) - (p->write.index - PA_PSTREAM_DESCRIPTOR_SIZE); + } + + pa_assert(l > 0); + +#ifdef HAVE_CREDS + if (p->send_ancil_data_now) { + if (p->write_ancil_data->creds_valid) { + pa_assert(p->write_ancil_data->nfd == 0); + if ((r = pa_iochannel_write_with_creds(p->io, d, l, &p->write_ancil_data->creds)) < 0) + goto fail; + } + else + if ((r = pa_iochannel_write_with_fds(p->io, d, l, p->write_ancil_data->nfd, p->write_ancil_data->fds)) < 0) + goto fail; + + pa_cmsg_ancil_data_close_fds(p->write_ancil_data); + p->send_ancil_data_now = false; + } else +#endif + if (p->srb) + r = pa_srbchannel_write(p->srb, d, l); + else if ((r = pa_iochannel_write(p->io, d, l)) < 0) + goto fail; + + if (release_memblock) + pa_memblock_release(release_memblock); + + p->write.index += (size_t) r; + + if (p->write.index >= PA_PSTREAM_DESCRIPTOR_SIZE + ntohl(p->write.descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH])) { + pa_assert(p->write.current); + item_free(p->write.current); + p->write.current = NULL; + + if (p->write.memchunk.memblock) + pa_memblock_unref(p->write.memchunk.memblock); + + pa_memchunk_reset(&p->write.memchunk); + + if (p->drain_callback && !pa_pstream_is_pending(p)) + p->drain_callback(p, p->drain_callback_userdata); + } + + return (size_t) r == l ? 1 : 0; + +fail: +#ifdef HAVE_CREDS + if (p->send_ancil_data_now) + pa_cmsg_ancil_data_close_fds(p->write_ancil_data); +#endif + + if (release_memblock) + pa_memblock_release(release_memblock); + + return -1; +} + +static void memblock_complete(pa_pstream *p, struct pstream_read *re) { + pa_memchunk chunk; + int64_t offset; + + if (!p->receive_memblock_callback) + return; + + chunk.memblock = re->memblock; + chunk.index = 0; + chunk.length = re->index - PA_PSTREAM_DESCRIPTOR_SIZE; + + offset = (int64_t) ( + (((uint64_t) ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI])) << 32) | + (((uint64_t) ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_LO])))); + + p->receive_memblock_callback( + p, + ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_CHANNEL]), + offset, + ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS]) & PA_FLAG_SEEKMASK, + &chunk, + p->receive_memblock_callback_userdata); +} + +static int do_read(pa_pstream *p, struct pstream_read *re) { + void *d; + size_t l; + ssize_t r; + pa_memblock *release_memblock = NULL; + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (re->index < PA_PSTREAM_DESCRIPTOR_SIZE) { + d = (uint8_t*) re->descriptor + re->index; + l = PA_PSTREAM_DESCRIPTOR_SIZE - re->index; + } else { + pa_assert(re->data || re->memblock); + + if (re->data) + d = re->data; + else { + d = pa_memblock_acquire(re->memblock); + release_memblock = re->memblock; + } + + d = (uint8_t*) d + re->index - PA_PSTREAM_DESCRIPTOR_SIZE; + l = ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH]) - (re->index - PA_PSTREAM_DESCRIPTOR_SIZE); + } + + if (re == &p->readsrb) { + r = pa_srbchannel_read(p->srb, d, l); + if (r == 0) { + if (release_memblock) + pa_memblock_release(release_memblock); + return 1; + } + } + else +#ifdef HAVE_CREDS + { + pa_cmsg_ancil_data b; + + if ((r = pa_iochannel_read_with_ancil_data(p->io, d, l, &b)) <= 0) + goto fail; + + if (b.creds_valid) { + p->read_ancil_data.creds_valid = true; + p->read_ancil_data.creds = b.creds; + } + if (b.nfd > 0) { + pa_assert(b.nfd <= MAX_ANCIL_DATA_FDS); + p->read_ancil_data.nfd = b.nfd; + memcpy(p->read_ancil_data.fds, b.fds, sizeof(int) * b.nfd); + p->read_ancil_data.close_fds_on_cleanup = b.close_fds_on_cleanup; + } + } +#else + if ((r = pa_iochannel_read(p->io, d, l)) <= 0) + goto fail; +#endif + + if (release_memblock) + pa_memblock_release(release_memblock); + + re->index += (size_t) r; + + if (re->index == PA_PSTREAM_DESCRIPTOR_SIZE) { + uint32_t flags, length, channel; + /* Reading of frame descriptor complete */ + + flags = ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS]); + + if (!p->use_shm && (flags & PA_FLAG_SHMMASK) != 0) { + pa_log_warn("Received SHM frame on a socket where SHM is disabled."); + return -1; + } + + if (flags == PA_FLAG_SHMRELEASE) { + + /* This is a SHM memblock release frame with no payload */ + +/* pa_log("Got release frame for %u", ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI])); */ + + pa_assert(p->export); + pa_memexport_process_release(p->export, ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI])); + + goto frame_done; + + } else if (flags == PA_FLAG_SHMREVOKE) { + + /* This is a SHM memblock revoke frame with no payload */ + +/* pa_log("Got revoke frame for %u", ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI])); */ + + pa_assert(p->import); + pa_memimport_process_revoke(p->import, ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI])); + + goto frame_done; + } + + length = ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH]); + + if (length > FRAME_SIZE_MAX_ALLOW || length <= 0) { + pa_log_warn("Received invalid frame size: %lu", (unsigned long) length); + return -1; + } + + pa_assert(!re->packet && !re->memblock); + + channel = ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_CHANNEL]); + + if (channel == (uint32_t) -1) { + size_t plen; + + if (flags != 0) { + pa_log_warn("Received packet frame with invalid flags value."); + return -1; + } + + /* Frame is a packet frame */ + re->packet = pa_packet_new(length); + re->data = (void *) pa_packet_data(re->packet, &plen); + + } else { + + if ((flags & PA_FLAG_SEEKMASK) > PA_SEEK_RELATIVE_END) { + pa_log_warn("Received memblock frame with invalid seek mode."); + return -1; + } + + if (((flags & PA_FLAG_SHMMASK) & PA_FLAG_SHMDATA) != 0) { + + if (length != sizeof(re->shm_info)) { + pa_log_warn("Received SHM memblock frame with invalid frame length."); + return -1; + } + + /* Frame is a memblock frame referencing an SHM memblock */ + re->data = re->shm_info; + + } else if ((flags & PA_FLAG_SHMMASK) == 0) { + + /* Frame is a memblock frame */ + + re->memblock = pa_memblock_new(p->mempool, length); + re->data = NULL; + } else { + + pa_log_warn("Received memblock frame with invalid flags value."); + return -1; + } + } + + } else if (re->index >= ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_LENGTH]) + PA_PSTREAM_DESCRIPTOR_SIZE) { + /* Frame complete */ + + if (re->memblock) { + memblock_complete(p, re); + + /* This was a memblock frame. We can unref the memblock now */ + pa_memblock_unref(re->memblock); + + } else if (re->packet) { + + if (p->receive_packet_callback) +#ifdef HAVE_CREDS + p->receive_packet_callback(p, re->packet, &p->read_ancil_data, p->receive_packet_callback_userdata); +#else + p->receive_packet_callback(p, re->packet, NULL, p->receive_packet_callback_userdata); +#endif + + pa_packet_unref(re->packet); + } else { + pa_memblock *b = NULL; + uint32_t flags = ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS]); + uint32_t shm_id = ntohl(re->shm_info[PA_PSTREAM_SHM_SHMID]); + pa_mem_type_t type = (flags & PA_FLAG_SHMDATA_MEMFD_BLOCK) ? + PA_MEM_TYPE_SHARED_MEMFD : PA_MEM_TYPE_SHARED_POSIX; + + pa_assert(((flags & PA_FLAG_SHMMASK) & PA_FLAG_SHMDATA) != 0); + pa_assert(p->import); + + if (type == PA_MEM_TYPE_SHARED_MEMFD && p->use_memfd && + !pa_idxset_get_by_data(p->registered_memfd_ids, PA_UINT32_TO_PTR(shm_id), NULL)) { + + if (pa_log_ratelimit(PA_LOG_ERROR)) + pa_log("Ignoring received block reference with non-registered memfd ID = %u", shm_id); + + } else if (!(b = pa_memimport_get(p->import, + type, + ntohl(re->shm_info[PA_PSTREAM_SHM_BLOCKID]), + shm_id, + ntohl(re->shm_info[PA_PSTREAM_SHM_INDEX]), + ntohl(re->shm_info[PA_PSTREAM_SHM_LENGTH]), + !!(flags & PA_FLAG_SHMWRITABLE)))) { + + if (pa_log_ratelimit(PA_LOG_DEBUG)) + pa_log_debug("Failed to import memory block."); + } + + if (p->receive_memblock_callback) { + int64_t offset; + pa_memchunk chunk; + + chunk.memblock = b; + chunk.index = 0; + chunk.length = b ? pa_memblock_get_length(b) : ntohl(re->shm_info[PA_PSTREAM_SHM_LENGTH]); + + offset = (int64_t) ( + (((uint64_t) ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_HI])) << 32) | + (((uint64_t) ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_OFFSET_LO])))); + + p->receive_memblock_callback( + p, + ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_CHANNEL]), + offset, + ntohl(re->descriptor[PA_PSTREAM_DESCRIPTOR_FLAGS]) & PA_FLAG_SEEKMASK, + &chunk, + p->receive_memblock_callback_userdata); + } + + if (b) + pa_memblock_unref(b); + } + + goto frame_done; + } + + return 0; + +frame_done: + re->memblock = NULL; + re->packet = NULL; + re->index = 0; + re->data = NULL; + +#ifdef HAVE_CREDS + /* FIXME: Close received ancillary data fds if the pstream's + * receive_packet_callback did not do so. + * + * Malicious clients can attach fds to unknown commands, or attach them + * to commands that does not expect fds. By doing so, server will reach + * its open fd limit and future clients' SHM transfers will always fail. + */ + p->read_ancil_data.creds_valid = false; + p->read_ancil_data.nfd = 0; +#endif + + return 0; + +fail: + if (release_memblock) + pa_memblock_release(release_memblock); + + return -1; +} + +void pa_pstream_set_die_callback(pa_pstream *p, pa_pstream_notify_cb_t cb, void *userdata) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->die_callback = cb; + p->die_callback_userdata = userdata; +} + +void pa_pstream_set_drain_callback(pa_pstream *p, pa_pstream_notify_cb_t cb, void *userdata) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->drain_callback = cb; + p->drain_callback_userdata = userdata; +} + +void pa_pstream_set_receive_packet_callback(pa_pstream *p, pa_pstream_packet_cb_t cb, void *userdata) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->receive_packet_callback = cb; + p->receive_packet_callback_userdata = userdata; +} + +void pa_pstream_set_receive_memblock_callback(pa_pstream *p, pa_pstream_memblock_cb_t cb, void *userdata) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->receive_memblock_callback = cb; + p->receive_memblock_callback_userdata = userdata; +} + +void pa_pstream_set_release_callback(pa_pstream *p, pa_pstream_block_id_cb_t cb, void *userdata) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->release_callback = cb; + p->release_callback_userdata = userdata; +} + +void pa_pstream_set_revoke_callback(pa_pstream *p, pa_pstream_block_id_cb_t cb, void *userdata) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->revoke_callback = cb; + p->revoke_callback_userdata = userdata; +} + +bool pa_pstream_is_pending(pa_pstream *p) { + bool b; + + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (p->dead) + b = false; + else + b = p->write.current || !pa_queue_isempty(p->send_queue); + + return b; +} + +void pa_pstream_unref(pa_pstream*p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + if (PA_REFCNT_DEC(p) <= 0) + pstream_free(p); +} + +pa_pstream* pa_pstream_ref(pa_pstream*p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + PA_REFCNT_INC(p); + return p; +} + +void pa_pstream_unlink(pa_pstream *p) { + pa_assert(p); + + if (p->dead) + return; + + p->dead = true; + + while (p->srb || p->is_srbpending) /* In theory there could be one active and one pending */ + pa_pstream_set_srbchannel(p, NULL); + + if (p->import) { + pa_memimport_free(p->import); + p->import = NULL; + } + + if (p->export) { + pa_memexport_free(p->export); + p->export = NULL; + } + + if (p->io) { + pa_iochannel_free(p->io); + p->io = NULL; + } + + if (p->defer_event) { + p->mainloop->defer_free(p->defer_event); + p->defer_event = NULL; + } + + p->die_callback = NULL; + p->drain_callback = NULL; + p->receive_packet_callback = NULL; + p->receive_memblock_callback = NULL; +} + +void pa_pstream_enable_shm(pa_pstream *p, bool enable) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + p->use_shm = enable; + + if (enable) { + + if (!p->export) + p->export = pa_memexport_new(p->mempool, memexport_revoke_cb, p); + + } else { + + if (p->export) { + pa_memexport_free(p->export); + p->export = NULL; + } + } +} + +void pa_pstream_enable_memfd(pa_pstream *p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + pa_assert(p->use_shm); + + p->use_memfd = true; + + if (!p->registered_memfd_ids) { + p->registered_memfd_ids = pa_idxset_new(NULL, NULL); + } +} + +bool pa_pstream_get_shm(pa_pstream *p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + return p->use_shm; +} + +bool pa_pstream_get_memfd(pa_pstream *p) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0); + + return p->use_memfd; +} + +void pa_pstream_set_srbchannel(pa_pstream *p, pa_srbchannel *srb) { + pa_assert(p); + pa_assert(PA_REFCNT_VALUE(p) > 0 || srb == NULL); + + if (srb == p->srb) + return; + + /* We can't handle quick switches between srbchannels. */ + pa_assert(!p->is_srbpending); + + p->srbpending = srb; + p->is_srbpending = true; + + /* Switch immediately, if possible. */ + if (p->dead) + check_srbpending(p); + else + do_write(p); +} |