September 2025 - gerrit-log - lists.osmocom.org

[XL] Change in osmo-ttcn3-hacks[master]: smdpp: es9p testsuite

by laforge

Attention is currently required from: Hoernchen. laforge has posted comments on this change by Hoernchen. ( https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/41119?usp=email ) Change subject: smdpp: es9p testsuite ...................................................................... Patch Set 4: Code-Review+1 (1 comment) Patchset: PS4: what I'm missign in both the main .ttcn file and the commit log is which exact version of which GSMA test suite document was used for implementingthe test suite. -- To view, visit https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/41119?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: comment Gerrit-Project: osmo-ttcn3-hacks Gerrit-Branch: master Gerrit-Change-Id: I3ba163d9155a3b019214dbccc3c8031dfd7deb6b Gerrit-Change-Number: 41119 Gerrit-PatchSet: 4 Gerrit-Owner: Hoernchen <ewild(a)sysmocom.de> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Attention: Hoernchen <ewild(a)sysmocom.de> Gerrit-Comment-Date: Thu, 18 Sep 2025 09:46:06 +0000 Gerrit-HasComments: Yes Gerrit-Has-Labels: Yes

3 days, 2 hours

1
0
0 0

[XL] Change in osmo-ttcn3-hacks[master]: euicc: clean up superfluous encdec cpp files

by laforge

Attention is currently required from: Hoernchen, dexter. laforge has posted comments on this change by Hoernchen. ( https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/40983?usp=email ) Change subject: euicc: clean up superfluous encdec cpp files ...................................................................... Patch Set 3: Code-Review+2 -- To view, visit https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/40983?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: comment Gerrit-Project: osmo-ttcn3-hacks Gerrit-Branch: master Gerrit-Change-Id: Ice4ab4328faf6ad34a060bc2004a44de10691cff Gerrit-Change-Number: 40983 Gerrit-PatchSet: 3 Gerrit-Owner: Hoernchen <ewild(a)sysmocom.de> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: dexter <pmaier(a)sysmocom.de> Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de> Gerrit-Attention: Hoernchen <ewild(a)sysmocom.de> Gerrit-Attention: dexter <pmaier(a)sysmocom.de> Gerrit-Comment-Date: Thu, 18 Sep 2025 09:43:52 +0000 Gerrit-HasComments: No Gerrit-Has-Labels: Yes

3 days, 2 hours

1
0
0 0

[M] Change in osmo-ttcn3-hacks[master]: euicc: changes and extensions to support es9p and es2p testing

by laforge

Attention is currently required from: Hoernchen, dexter, pespin. laforge has posted comments on this change by Hoernchen. ( https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/40982?usp=email ) Change subject: euicc: changes and extensions to support es9p and es2p testing ...................................................................... Patch Set 3: Code-Review+2 -- To view, visit https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/40982?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: comment Gerrit-Project: osmo-ttcn3-hacks Gerrit-Branch: master Gerrit-Change-Id: I9c622751e195d33f6f0a832265d6b10674d1c75c Gerrit-Change-Number: 40982 Gerrit-PatchSet: 3 Gerrit-Owner: Hoernchen <ewild(a)sysmocom.de> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: dexter <pmaier(a)sysmocom.de> Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de> Gerrit-Attention: Hoernchen <ewild(a)sysmocom.de> Gerrit-Attention: pespin <pespin(a)sysmocom.de> Gerrit-Attention: dexter <pmaier(a)sysmocom.de> Gerrit-Comment-Date: Thu, 18 Sep 2025 09:43:10 +0000 Gerrit-HasComments: No Gerrit-Has-Labels: Yes

3 days, 2 hours

1
0
0 0

[M] Change in libosmocore[master]: Put all io_uring related read and write states into sub structures

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40856?usp=email ) Change subject: Put all io_uring related read and write states into sub structures ...................................................................... Put all io_uring related read and write states into sub structures Related: OS#6705 Change-Id: I3cf4ab6d9aebf5532ad174c90c7b0e9633491c88 --- M src/core/osmo_io.c M src/core/osmo_io_internal.h M src/core/osmo_io_uring.c 3 files changed, 70 insertions(+), 63 deletions(-) Approvals: laforge: Looks good to me, approved Jenkins Builder: Verified pespin: Looks good to me, but someone else must approve diff --git a/src/core/osmo_io.c b/src/core/osmo_io.c index 2f3af06..399a604 100644 --- a/src/core/osmo_io.c +++ b/src/core/osmo_io.c @@ -901,7 +901,7 @@ if (sqes < 1 || sqes > IOFD_MSGHDR_MAX_READ_SQES) return -EINVAL; - iofd->u.uring.num_read_sqes = sqes; + iofd->u.uring.read.num_sqes = sqes; return 0; } diff --git a/src/core/osmo_io_internal.h b/src/core/osmo_io_internal.h index bf5bb76..f425da2 100644 --- a/src/core/osmo_io_internal.h +++ b/src/core/osmo_io_internal.h @@ -108,19 +108,26 @@ struct osmo_fd ofd; } poll; struct { - bool read_enabled; - bool write_enabled; - /*! requested number of simultaniously submitted read SQEs */ - uint8_t num_read_sqes; - /*! array of simultaneously submitted read SQEs */ - void *read_msghdr[IOFD_MSGHDR_MAX_READ_SQES]; - /*! ring the read SQEs have been submitted to */ - struct io_uring *read_ring; - /*! current number of simultaneously submitted read SQEs */ - uint8_t reads_submitted; - void *write_msghdr; - /*! ring the write SQE has been submitted to */ - struct io_uring *write_ring; + struct { + /*! read is enabled, due to registration of callback function */ + bool enabled; + /*! requested number of simultaniously submitted read SQEs */ + uint8_t num_sqes; + /*! array of simultaneously submitted read SQEs */ + void *msghdr[IOFD_MSGHDR_MAX_READ_SQES]; + /*! ring the read SQEs have been submitted to */ + struct io_uring *ring; + /*! current number of simultaneously submitted read SQEs */ + uint8_t sqes_submitted; + } read; + struct { + /*! write is enabled, due to pending msghdr in tx_queue */ + bool enabled; + /*! submitted write SQE */ + void *msghdr; + /*! ring the write SQE has been submitted to */ + struct io_uring *ring; + } write; /* TODO: index into array of registered fd's? */ /* osmo_fd for non-blocking connect handling */ struct osmo_fd connect_ofd; diff --git a/src/core/osmo_io_uring.c b/src/core/osmo_io_uring.c index 45a08fe..01b8063 100644 --- a/src/core/osmo_io_uring.c +++ b/src/core/osmo_io_uring.c @@ -196,7 +196,7 @@ struct io_uring_sqe *sqe; /* All subsequent read SQEs must be on the same ring. */ - if (read && iofd->u.uring.reads_submitted > 0 && iofd->u.uring.read_ring != &g_ring->ring) + if (read && iofd->u.uring.read.sqes_submitted > 0 && iofd->u.uring.read.ring != &g_ring->ring) return NULL; sqe = io_uring_get_sqe_and_count(g_ring); @@ -204,7 +204,7 @@ return sqe; /* The current ring is full, subsequent read SQEs on different ring are not allowed. */ - if (read && iofd->u.uring.reads_submitted > 0) + if (read && iofd->u.uring.read.sqes_submitted > 0) return NULL; if (g_io_uring_size < IOFD_URING_MAXIMUM_SIZE) { @@ -246,7 +246,7 @@ /* Tell iofd_uring_get_sqe() not to allocate a new ring, if we want to enqueue multiple read SQEs. */ sqe = iofd_uring_get_sqe(iofd, true); if (!sqe) { - if (iofd->u.uring.reads_submitted > 0) + if (iofd->u.uring.read.sqes_submitted > 0) return -EINVAL; LOGPIO(iofd, LOGL_ERROR, "Could not get io_uring_sqe\n"); OSMO_ASSERT(0); @@ -301,9 +301,9 @@ iofd_io_uring_submit(); - iofd->u.uring.read_msghdr[iofd->u.uring.reads_submitted] = msghdr; - iofd->u.uring.reads_submitted++; - iofd->u.uring.read_ring = &g_ring->ring; + iofd->u.uring.read.msghdr[iofd->u.uring.read.sqes_submitted] = msghdr; + iofd->u.uring.read.sqes_submitted++; + iofd->u.uring.read.ring = &g_ring->ring; return 0; } @@ -313,7 +313,7 @@ int rc; /* Submit more read SQEs in advance, if requested. */ - while (iofd->u.uring.reads_submitted < iofd->u.uring.num_read_sqes) { + while (iofd->u.uring.read.sqes_submitted < iofd->u.uring.read.num_sqes) { rc = iofd_uring_submit_recv_sqe(iofd, action); /* Stop, if we cannot enqueue multiple read SQEs in the same ring. */ if (rc < 0) @@ -328,19 +328,19 @@ uint8_t idx, i; /* Find which read_msghdr is completed and remove from list. */ - for (idx = 0; idx < iofd->u.uring.reads_submitted; idx++) { - if (iofd->u.uring.read_msghdr[idx] == msghdr) + for (idx = 0; idx < iofd->u.uring.read.sqes_submitted; idx++) { + if (iofd->u.uring.read.msghdr[idx] == msghdr) break; } - if (idx == iofd->u.uring.reads_submitted) { + if (idx == iofd->u.uring.read.sqes_submitted) { LOGP(DLIO, LOGL_FATAL, "Read SQE completion, but msghdr not found, please fix!\n"); return; } /* Remove entry at idx. */ - iofd->u.uring.reads_submitted--; - for (i = idx; i < iofd->u.uring.reads_submitted; i++) - iofd->u.uring.read_msghdr[i] = iofd->u.uring.read_msghdr[i + 1]; - iofd->u.uring.read_msghdr[i] = NULL; + iofd->u.uring.read.sqes_submitted--; + for (i = idx; i < iofd->u.uring.read.sqes_submitted; i++) + iofd->u.uring.read.msghdr[i] = iofd->u.uring.read.msghdr[i + 1]; + iofd->u.uring.read.msghdr[i] = NULL; for (idx = 0; idx < msghdr->io_len; idx++) { struct msgb *msg = msghdr->msg[idx]; @@ -359,7 +359,7 @@ } /* Check for every iteration, because iofd might get unregistered/closed during receive function. */ - if (iofd->u.uring.read_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) + if (iofd->u.uring.read.enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_handle_recv(iofd, msg, chunk, msghdr); else msgb_free(msg); @@ -372,7 +372,7 @@ msghdr->msg[idx] = NULL; } - if (iofd->u.uring.read_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) + if (iofd->u.uring.read.enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_uring_submit_recv(iofd, msghdr->action); iofd_msghdr_free(msghdr); @@ -388,12 +388,12 @@ /* Detach msghdr from iofd. It might get freed here or it is freed during iofd_handle_send_completion(). * If there is pending data to send, iofd_uring_submit_tx() will attach it again. * iofd_handle_send_completion() will invoke a callback function to signal the possibility of write/send. - * This callback function might close iofd, leading to the potential freeing of iofd->u.uring.write_msghdr if + * This callback function might close iofd, leading to the potential freeing of iofd->u.uring.write.msghdr if * still attached. Since iofd_handle_send_completion() frees msghdr at the end of the function, detaching * msghdr here prevents a double-free bug. */ - if (iofd->u.uring.write_msghdr == msghdr) { - iofd->u.uring.write_msghdr = NULL; - iofd->u.uring.write_ring = NULL; + if (iofd->u.uring.write.msghdr == msghdr) { + iofd->u.uring.write.msghdr = NULL; + iofd->u.uring.write.ring = NULL; } if (OSMO_UNLIKELY(IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED))) { @@ -407,7 +407,7 @@ } /* submit the next to-be-transmitted message for this file descriptor */ - if (iofd->u.uring.write_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) + if (iofd->u.uring.write.enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_uring_submit_tx(iofd); } @@ -435,7 +435,7 @@ IOFD_FLAG_UNSET(iofd, IOFD_FLAG_IN_CALLBACK); - if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.reads_submitted && !iofd->u.uring.write_msghdr) + if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.read.sqes_submitted && !iofd->u.uring.write.msghdr) talloc_free(iofd); } @@ -525,8 +525,8 @@ iofd_io_uring_submit(); - iofd->u.uring.write_msghdr = msghdr; - iofd->u.uring.write_ring = &g_ring->ring; + iofd->u.uring.write.msghdr = msghdr; + iofd->u.uring.write.ring = &g_ring->ring; return 0; } @@ -565,19 +565,19 @@ IOFD_FLAG_UNSET(iofd, IOFD_FLAG_IN_CALLBACK); /* If write/read notifications are pending, enable it now. */ - if (iofd->u.uring.write_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) + if (iofd->u.uring.write.enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_uring_write_enable(iofd); - if (iofd->u.uring.read_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) + if (iofd->u.uring.read.enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_uring_read_enable(iofd); - if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.reads_submitted && !iofd->u.uring.write_msghdr) + if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.read.sqes_submitted && !iofd->u.uring.write.msghdr) talloc_free(iofd); return 0; } static int iofd_uring_setup(struct osmo_io_fd *iofd) { - iofd->u.uring.num_read_sqes = g_io_uring_read_sqes; + iofd->u.uring.read.num_sqes = g_io_uring_read_sqes; return 0; } @@ -609,10 +609,10 @@ struct iofd_msghdr *msghdr; uint8_t idx; - for (idx = 0; idx < iofd->u.uring.reads_submitted; idx++) { - struct osmo_io_uring *ring = container_of(iofd->u.uring.read_ring, struct osmo_io_uring, ring); - msghdr = iofd->u.uring.read_msghdr[idx]; - iofd->u.uring.read_msghdr[idx] = NULL; + for (idx = 0; idx < iofd->u.uring.read.sqes_submitted; idx++) { + struct osmo_io_uring *ring = container_of(iofd->u.uring.read.ring, struct osmo_io_uring, ring); + msghdr = iofd->u.uring.read.msghdr[idx]; + iofd->u.uring.read.msghdr[idx] = NULL; /* Submit SQEs of the current ring, if needed. */ if (&ring->ring == &g_ring->ring) osmo_io_uring_submit(); @@ -630,15 +630,15 @@ talloc_steal(OTC_GLOBAL, msghdr); msghdr->iofd = NULL; } - if (iofd->u.uring.reads_submitted) { - iofd->u.uring.read_ring = NULL; - iofd->u.uring.reads_submitted = 0; + if (iofd->u.uring.read.sqes_submitted) { + iofd->u.uring.read.ring = NULL; + iofd->u.uring.read.sqes_submitted = 0; } - if (iofd->u.uring.write_msghdr) { - struct osmo_io_uring *ring = container_of(iofd->u.uring.write_ring, struct osmo_io_uring, ring); - msghdr = iofd->u.uring.write_msghdr; - iofd->u.uring.write_msghdr = NULL; + if (iofd->u.uring.write.msghdr) { + struct osmo_io_uring *ring = container_of(iofd->u.uring.write.ring, struct osmo_io_uring, ring); + msghdr = iofd->u.uring.write.msghdr; + iofd->u.uring.write.msghdr = NULL; for (int idx = 0; idx < msghdr->io_len; idx++) { msgb_free(msghdr->msg[idx]); msghdr->msg[idx] = NULL; @@ -659,7 +659,7 @@ } talloc_steal(OTC_GLOBAL, msghdr); msghdr->iofd = NULL; - iofd->u.uring.write_ring = NULL; + iofd->u.uring.write.ring = NULL; } if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_NOTIFY_CONNECTED)) { @@ -672,9 +672,9 @@ static void iofd_uring_write_enable(struct osmo_io_fd *iofd) { - iofd->u.uring.write_enabled = true; + iofd->u.uring.write.enabled = true; - if (iofd->u.uring.write_msghdr) + if (iofd->u.uring.write.msghdr) return; /* This function is called again, once the socket is connected. */ @@ -711,21 +711,21 @@ iofd_io_uring_submit(); - iofd->u.uring.write_msghdr = msghdr; - iofd->u.uring.write_ring = &g_ring->ring; + iofd->u.uring.write.msghdr = msghdr; + iofd->u.uring.write.ring = &g_ring->ring; } } static void iofd_uring_write_disable(struct osmo_io_fd *iofd) { - iofd->u.uring.write_enabled = false; + iofd->u.uring.write.enabled = false; } static void iofd_uring_read_enable(struct osmo_io_fd *iofd) { - iofd->u.uring.read_enabled = true; + iofd->u.uring.read.enabled = true; - if (iofd->u.uring.reads_submitted) + if (iofd->u.uring.read.sqes_submitted) return; /* This function is called again, once the socket is connected. */ @@ -749,7 +749,7 @@ static void iofd_uring_read_disable(struct osmo_io_fd *iofd) { - iofd->u.uring.read_enabled = false; + iofd->u.uring.read.enabled = false; } static int iofd_uring_close(struct osmo_io_fd *iofd) @@ -765,7 +765,7 @@ } /* OSMO_IO_FD_MODE_RECVMSG_SENDMSG: Don't call this function after enabling read or write. */ - OSMO_ASSERT(!iofd->u.uring.write_enabled && !iofd->u.uring.read_enabled); + OSMO_ASSERT(!iofd->u.uring.write.enabled && !iofd->u.uring.read.enabled); /* Set flag to enable temporary osmo_fd during register() time: */ IOFD_FLAG_SET(iofd, IOFD_FLAG_NOTIFY_CONNECTED); -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40856?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: I3cf4ab6d9aebf5532ad174c90c7b0e9633491c88 Gerrit-Change-Number: 40856 Gerrit-PatchSet: 6 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

[M] Change in libosmocore[master]: Automatically increase io_uring, if too small.

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40725?usp=email ) Change subject: Automatically increase io_uring, if too small. ...................................................................... Automatically increase io_uring, if too small. The ring may be too small to store all SQEs before the kernel can handle them. If this happens, a new ring is allocated with twice of the size of the old ring. The old ring will not be destroyed, as it still contains uncompleted elements. Some of them may never be completed. A pointer to the current ring will be stored within the msghdr structure. It is used when cancelling an SQE. The cancellation must be performed in the same ring where it was created. It is quite unlikely that the old ring cannot store the cancellation SQE. If this happens, the cancellation is queued and submitted, once the ring can store it. The old ring will not be removed, because there is currently no counter to determine when all submissions are completed. Related: OS#6705 Change-Id: Id9230146acc8d54bfd44834e783c31b37bd64bca --- M src/core/osmo_io_internal.h M src/core/osmo_io_uring.c 2 files changed, 148 insertions(+), 39 deletions(-) Approvals: Jenkins Builder: Verified pespin: Looks good to me, but someone else must approve laforge: Looks good to me, approved diff --git a/src/core/osmo_io_internal.h b/src/core/osmo_io_internal.h index f983b77..bf5bb76 100644 --- a/src/core/osmo_io_internal.h +++ b/src/core/osmo_io_internal.h @@ -114,9 +114,13 @@ uint8_t num_read_sqes; /*! array of simultaneously submitted read SQEs */ void *read_msghdr[IOFD_MSGHDR_MAX_READ_SQES]; + /*! ring the read SQEs have been submitted to */ + struct io_uring *read_ring; /*! current number of simultaneously submitted read SQEs */ uint8_t reads_submitted; void *write_msghdr; + /*! ring the write SQE has been submitted to */ + struct io_uring *write_ring; /* TODO: index into array of registered fd's? */ /* osmo_fd for non-blocking connect handling */ struct osmo_fd connect_ofd; @@ -157,6 +161,9 @@ /*! I/O file descriptor on which we perform this I/O operation */ struct osmo_io_fd *iofd; + /*! msghdr is in the cancel_queue list */ + bool in_cancel_queue; + /*! control message buffer for passing sctp_sndrcvinfo along */ char cmsg[0]; /* size is determined by iofd->cmsg_size on recvmsg, and by mcghdr->msg_controllen on sendmsg */ }; diff --git a/src/core/osmo_io_uring.c b/src/core/osmo_io_uring.c index 48f6f0d..12caa77 100644 --- a/src/core/osmo_io_uring.c +++ b/src/core/osmo_io_uring.c @@ -35,6 +35,7 @@ #include <string.h> #include <stdbool.h> #include <errno.h> +#include <limits.h> #include <netinet/in.h> #include <netinet/sctp.h> @@ -72,12 +73,15 @@ struct osmo_io_uring { struct osmo_fd event_ofd; struct io_uring ring; + struct llist_head cancel_queue; }; -static __thread struct osmo_io_uring g_ring; +static __thread struct osmo_io_uring *g_ring = NULL; static void iofd_uring_cqe(struct io_uring *ring); +void osmo_io_uring_submit(void); + /*! read call-back for eventfd notifying us if entries are in the completion queue */ static int iofd_uring_poll_cb(struct osmo_fd *ofd, unsigned int what) { @@ -109,7 +113,7 @@ if ((env = getenv(OSMO_IO_URING_BATCH))) g_io_uring_batch = true; - if ((env = getenv(OSMO_IO_URING_INITIAL_SIZE))) { + if (!g_ring && (env = getenv(OSMO_IO_URING_INITIAL_SIZE))) { int env_value; rc = osmo_str_to_int(&env_value, env, 10, 1, IOFD_URING_MAXIMUM_SIZE); if (rc < 0) { @@ -124,7 +128,10 @@ g_io_uring_size = env_value; } - rc = io_uring_queue_init(g_io_uring_size, &g_ring.ring, 0); + g_ring = talloc_zero(OTC_GLOBAL, struct osmo_io_uring); + INIT_LLIST_HEAD(&g_ring->cancel_queue); + + rc = io_uring_queue_init(g_io_uring_size, &g_ring->ring, 0); if (rc < 0) osmo_panic("failure during io_uring_queue_init(): %s\n", strerror(-rc)); @@ -139,42 +146,88 @@ rc = eventfd(0, 0); if (rc < 0) { - io_uring_queue_exit(&g_ring.ring); + io_uring_queue_exit(&g_ring->ring); osmo_panic("failure creating eventfd(0, 0) for io_uring: %s\n", strerror(-rc)); } evfd = rc; - osmo_fd_setup(&g_ring.event_ofd, evfd, OSMO_FD_READ, iofd_uring_poll_cb, &g_ring.ring, 0); - rc = osmo_fd_register(&g_ring.event_ofd); + osmo_fd_setup(&g_ring->event_ofd, evfd, OSMO_FD_READ, iofd_uring_poll_cb, &g_ring->ring, 0); + rc = osmo_fd_register(&g_ring->event_ofd); if (rc < 0) { close(evfd); - io_uring_queue_exit(&g_ring.ring); + io_uring_queue_exit(&g_ring->ring); osmo_panic("failure registering io_uring-eventfd as osmo_fd: %d\n", rc); } - rc = io_uring_register_eventfd(&g_ring.ring, evfd); + rc = io_uring_register_eventfd(&g_ring->ring, evfd); if (rc < 0) { - osmo_fd_unregister(&g_ring.event_ofd); + osmo_fd_unregister(&g_ring->event_ofd); close(evfd); - io_uring_queue_exit(&g_ring.ring); + io_uring_queue_exit(&g_ring->ring); osmo_panic("failure registering eventfd with io_uring: %s\n", strerror(-rc)); } } +static struct io_uring_sqe *iofd_uring_get_sqe(struct osmo_io_fd *iofd, bool read) +{ + struct io_uring_sqe *sqe; + + /* All subsequent read SQEs must be on the same ring. */ + if (read && iofd->u.uring.reads_submitted > 0 && iofd->u.uring.read_ring != &g_ring->ring) + return NULL; + + sqe = io_uring_get_sqe(&g_ring->ring); + if (sqe) + return sqe; + + /* The current ring is full, subsequent read SQEs on different ring are not allowed. */ + if (read && iofd->u.uring.reads_submitted > 0) + return NULL; + + if (g_io_uring_size < IOFD_URING_MAXIMUM_SIZE) { + LOGP(DLIO, LOGL_NOTICE, "io_uring too small to handle all SQEs with its current size of %d. " + "Increasing io_uring size to %d.\n", g_io_uring_size, g_io_uring_size * 2); + g_io_uring_size <<= 1; + } else { + LOGP(DLIO, LOGL_NOTICE, "io_uring too small to handle all SQEs with its maximum size of %d. " + "adding another one.\n", g_io_uring_size); + } + + /* Submit all SQEs of current ring and create a new ring, if needed. + * The old ring will be kept, as there are uncompleted submissions. + * TODO: Destroy old ring, once all submissions are completed. */ + osmo_io_uring_submit(); + + osmo_iofd_uring_init(); + + sqe = io_uring_get_sqe(&g_ring->ring); + OSMO_ASSERT(sqe); + return sqe; +} + static inline void iofd_io_uring_submit(void) { if (OSMO_LIKELY(!g_io_uring_batch)) - io_uring_submit(&g_ring.ring); + io_uring_submit(&g_ring->ring); else g_io_uring_submit_needed = true; } -static void iofd_uring_submit_recv_sqe(struct osmo_io_fd *iofd, enum iofd_msg_action action) +static inline int iofd_uring_submit_recv_sqe(struct osmo_io_fd *iofd, enum iofd_msg_action action) { struct msgb *msg; struct iofd_msghdr *msghdr; struct io_uring_sqe *sqe; uint8_t idx; + /* Tell iofd_uring_get_sqe() not to allocate a new ring, if we want to enqueue multiple read SQEs. */ + sqe = iofd_uring_get_sqe(iofd, true); + if (!sqe) { + if (iofd->u.uring.reads_submitted > 0) + return -EINVAL; + LOGPIO(iofd, LOGL_ERROR, "Could not get io_uring_sqe\n"); + OSMO_ASSERT(0); + } + msg = iofd_msgb_alloc(iofd); if (!msg) { LOGPIO(iofd, LOGL_ERROR, "Could not allocate msgb for reading\n"); @@ -209,12 +262,6 @@ OSMO_ASSERT(0); } - sqe = io_uring_get_sqe(&g_ring.ring); - if (!sqe) { - LOGPIO(iofd, LOGL_ERROR, "Could not get io_uring_sqe\n"); - OSMO_ASSERT(0); - } - switch (action) { case IOFD_ACT_READ: io_uring_prep_readv(sqe, iofd->fd, msghdr->iov, msghdr->hdr.msg_iovlen, -1); @@ -232,13 +279,22 @@ iofd->u.uring.read_msghdr[iofd->u.uring.reads_submitted] = msghdr; iofd->u.uring.reads_submitted++; + iofd->u.uring.read_ring = &g_ring->ring; + + return 0; } static void iofd_uring_submit_recv(struct osmo_io_fd *iofd, enum iofd_msg_action action) { + int rc; + /* Submit more read SQEs in advance, if requested. */ - while (iofd->u.uring.reads_submitted < iofd->u.uring.num_read_sqes) - iofd_uring_submit_recv_sqe(iofd, action); + while (iofd->u.uring.reads_submitted < iofd->u.uring.num_read_sqes) { + rc = iofd_uring_submit_recv_sqe(iofd, action); + /* Stop, if we cannot enqueue multiple read SQEs in the same ring. */ + if (rc < 0) + break; + } } /*! completion call-back for READ/RECVFROM */ @@ -311,8 +367,10 @@ * This callback function might close iofd, leading to the potential freeing of iofd->u.uring.write_msghdr if * still attached. Since iofd_handle_send_completion() frees msghdr at the end of the function, detaching * msghdr here prevents a double-free bug. */ - if (iofd->u.uring.write_msghdr == msghdr) + if (iofd->u.uring.write_msghdr == msghdr) { iofd->u.uring.write_msghdr = NULL; + iofd->u.uring.write_ring = NULL; + } if (OSMO_UNLIKELY(IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED))) { for (int idx = 0; idx < msghdr->io_len; idx++) { @@ -362,7 +420,8 @@ { int rc; struct io_uring_cqe *cqe; - struct iofd_msghdr *msghdr; + struct iofd_msghdr *msghdr, *msghdr2; + struct osmo_io_uring *orig_ring = container_of(ring, struct osmo_io_uring, ring); while (io_uring_peek_cqe(ring, &cqe) == 0) { @@ -374,6 +433,8 @@ } if (!msghdr->iofd) { io_uring_cqe_seen(ring, cqe); + if (msghdr->in_cancel_queue) + llist_del(&msghdr->list); iofd_msghdr_free(msghdr); continue; } @@ -385,6 +446,23 @@ iofd_uring_handle_completion(msghdr, rc); } + + /* If there are unsubmitted cancel SQEs, try to queue them now. */ + if (OSMO_LIKELY(llist_empty(&orig_ring->cancel_queue))) + return; + llist_for_each_entry_safe(msghdr, msghdr2, &orig_ring->cancel_queue, list) { + struct io_uring_sqe *sqe; + sqe = io_uring_get_sqe(&orig_ring->ring); + if (!sqe) + break; + io_uring_sqe_set_data(sqe, NULL); + LOGP(DLIO, LOGL_DEBUG, "Cancelling queued read/write\n"); + io_uring_prep_cancel(sqe, msghdr, 0); + llist_del(&msghdr->list); + msghdr->in_cancel_queue = false; + } + io_uring_submit(&orig_ring->ring); + } /*! will submit the next to-be-transmitted message for given iofd */ @@ -397,7 +475,7 @@ if (!msghdr) return -ENODATA; - sqe = io_uring_get_sqe(&g_ring.ring); + sqe = iofd_uring_get_sqe(iofd, false); if (!sqe) { LOGPIO(iofd, LOGL_ERROR, "Could not get io_uring_sqe\n"); OSMO_ASSERT(0); @@ -420,6 +498,7 @@ iofd_io_uring_submit(); iofd->u.uring.write_msghdr = msghdr; + iofd->u.uring.write_ring = &g_ring->ring; return 0; } @@ -503,36 +582,58 @@ uint8_t idx; for (idx = 0; idx < iofd->u.uring.reads_submitted; idx++) { + struct osmo_io_uring *ring = container_of(iofd->u.uring.read_ring, struct osmo_io_uring, ring); msghdr = iofd->u.uring.read_msghdr[idx]; iofd->u.uring.read_msghdr[idx] = NULL; - sqe = io_uring_get_sqe(&g_ring.ring); - OSMO_ASSERT(sqe != NULL); - io_uring_sqe_set_data(sqe, NULL); - LOGPIO(iofd, LOGL_DEBUG, "Cancelling read\n"); + /* Submit SQEs of the current ring, if needed. */ + if (&ring->ring == &g_ring->ring) + osmo_io_uring_submit(); + /* If the submission queue is full, use cancel queue. We cannot cancel SQEs on the new ring. */ + sqe = io_uring_get_sqe(&ring->ring); + if (sqe) { + io_uring_sqe_set_data(sqe, NULL); + LOGPIO(iofd, LOGL_DEBUG, "Cancelling read\n"); + io_uring_prep_cancel(sqe, msghdr, 0); + io_uring_submit(&ring->ring); + } else { + llist_add_tail(&msghdr->list, &ring->cancel_queue); + msghdr->in_cancel_queue = true; + } talloc_steal(OTC_GLOBAL, msghdr); msghdr->iofd = NULL; - io_uring_prep_cancel(sqe, msghdr, 0); } - iofd->u.uring.reads_submitted = 0; + if (iofd->u.uring.reads_submitted) { + iofd->u.uring.read_ring = NULL; + iofd->u.uring.reads_submitted = 0; + } if (iofd->u.uring.write_msghdr) { + struct osmo_io_uring *ring = container_of(iofd->u.uring.write_ring, struct osmo_io_uring, ring); msghdr = iofd->u.uring.write_msghdr; - sqe = io_uring_get_sqe(&g_ring.ring); - OSMO_ASSERT(sqe != NULL); - io_uring_sqe_set_data(sqe, NULL); - LOGPIO(iofd, LOGL_DEBUG, "Cancelling write\n"); iofd->u.uring.write_msghdr = NULL; - talloc_steal(OTC_GLOBAL, msghdr); for (int idx = 0; idx < msghdr->io_len; idx++) { msgb_free(msghdr->msg[idx]); msghdr->msg[idx] = NULL; } + /* Submit SQEs of the current ring, if needed. */ + if (&ring->ring == &g_ring->ring) + osmo_io_uring_submit(); + /* If the submission queue is full, use cancel queue. We cannot cancel SQEs on the new ring. */ + sqe = io_uring_get_sqe(&ring->ring); + if (sqe) { + io_uring_sqe_set_data(sqe, NULL); + LOGPIO(iofd, LOGL_DEBUG, "Cancelling write\n"); + io_uring_prep_cancel(sqe, msghdr, 0); + io_uring_submit(&ring->ring); + } else { + llist_add_tail(&msghdr->list, &ring->cancel_queue); + msghdr->in_cancel_queue = true; + } + talloc_steal(OTC_GLOBAL, msghdr); msghdr->iofd = NULL; - io_uring_prep_cancel(sqe, msghdr, 0); + iofd->u.uring.write_ring = NULL; } - iofd_io_uring_submit(); - if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_NOTIFY_CONNECTED)) { osmo_fd_unregister(&iofd->u.uring.connect_ofd); IOFD_FLAG_UNSET(iofd, IOFD_FLAG_NOTIFY_CONNECTED); @@ -572,7 +673,7 @@ msghdr->iov[0].iov_base = msgb_data(msg); msghdr->iov[0].iov_len = msgb_length(msg); - sqe = io_uring_get_sqe(&g_ring.ring); + sqe = iofd_uring_get_sqe(iofd, false); if (!sqe) { LOGPIO(iofd, LOGL_ERROR, "Could not get io_uring_sqe\n"); OSMO_ASSERT(0); @@ -583,6 +684,7 @@ iofd_io_uring_submit(); iofd->u.uring.write_msghdr = msghdr; + iofd->u.uring.write_ring = &g_ring->ring; } } @@ -656,7 +758,7 @@ void osmo_io_uring_submit(void) { if (OSMO_LIKELY(g_io_uring_submit_needed)) { - io_uring_submit(&g_ring.ring); + io_uring_submit(&g_ring->ring); g_io_uring_submit_needed = false; } } -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40725?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: Id9230146acc8d54bfd44834e783c31b37bd64bca Gerrit-Change-Number: 40725 Gerrit-PatchSet: 15 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

[S] Change in libosmocore[master]: Remove old empty io_uring

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40855?usp=email ) Change subject: Remove old empty io_uring ...................................................................... Remove old empty io_uring A previous patch creates a new io_uring, if it becomes too small to store all SQEs. When all SQEs of the old ring are completed, the old ring will be destroyed. A counter is incremented whenever an SQE is submitted to an io_uring. The counter is decremented whenever a CQE is received and handled. This counter will determine when a ring is empty and can be destroyed. Related: OS#6705 Change-Id: Id2d2a0400ad442198c684ea0ead4eaeaead4c53d --- M src/core/osmo_io_uring.c 1 file changed, 36 insertions(+), 8 deletions(-) Approvals: pespin: Looks good to me, but someone else must approve Jenkins Builder: Verified laforge: Looks good to me, approved diff --git a/src/core/osmo_io_uring.c b/src/core/osmo_io_uring.c index 12caa77..45a08fe 100644 --- a/src/core/osmo_io_uring.c +++ b/src/core/osmo_io_uring.c @@ -66,7 +66,7 @@ bool g_io_uring_batch = false; bool g_io_uring_submit_needed = false; -static unsigned int g_io_uring_size = IOFD_URING_INITIAL_SIZE; +static int g_io_uring_size = IOFD_URING_INITIAL_SIZE; static int g_io_uring_read_sqes = 1; @@ -74,6 +74,7 @@ struct osmo_fd event_ofd; struct io_uring ring; struct llist_head cancel_queue; + unsigned int num_pending_submissions; }; static __thread struct osmo_io_uring *g_ring = NULL; @@ -167,6 +168,29 @@ } } +static void osmo_iofd_uring_exit(struct osmo_io_uring *ring) +{ + LOGP(DLIO, LOGL_DEBUG, "Old empty io_uring will be destroyed."); + + io_uring_queue_exit(&ring->ring); + + osmo_fd_unregister(&ring->event_ofd); + close(ring->event_ofd.fd); + + talloc_free(ring); +} + +static inline struct io_uring_sqe *io_uring_get_sqe_and_count(struct osmo_io_uring *ring) +{ + struct io_uring_sqe *sqe; + + sqe = io_uring_get_sqe(&ring->ring); + if (sqe) + ring->num_pending_submissions++; + + return sqe; +} + static struct io_uring_sqe *iofd_uring_get_sqe(struct osmo_io_fd *iofd, bool read) { struct io_uring_sqe *sqe; @@ -175,7 +199,7 @@ if (read && iofd->u.uring.reads_submitted > 0 && iofd->u.uring.read_ring != &g_ring->ring) return NULL; - sqe = io_uring_get_sqe(&g_ring->ring); + sqe = io_uring_get_sqe_and_count(g_ring); if (sqe) return sqe; @@ -199,7 +223,7 @@ osmo_iofd_uring_init(); - sqe = io_uring_get_sqe(&g_ring->ring); + sqe = io_uring_get_sqe_and_count(g_ring); OSMO_ASSERT(sqe); return sqe; } @@ -424,6 +448,7 @@ struct osmo_io_uring *orig_ring = container_of(ring, struct osmo_io_uring, ring); while (io_uring_peek_cqe(ring, &cqe) == 0) { + orig_ring->num_pending_submissions--; msghdr = io_uring_cqe_get_data(cqe); if (!msghdr) { @@ -448,11 +473,15 @@ } /* If there are unsubmitted cancel SQEs, try to queue them now. */ - if (OSMO_LIKELY(llist_empty(&orig_ring->cancel_queue))) + if (OSMO_LIKELY(llist_empty(&orig_ring->cancel_queue))) { + /* Old ring is empty, remove it. */ + if (OSMO_UNLIKELY(orig_ring != g_ring && orig_ring->num_pending_submissions == 0)) + osmo_iofd_uring_exit(orig_ring); return; + } llist_for_each_entry_safe(msghdr, msghdr2, &orig_ring->cancel_queue, list) { struct io_uring_sqe *sqe; - sqe = io_uring_get_sqe(&orig_ring->ring); + sqe = io_uring_get_sqe_and_count(orig_ring); if (!sqe) break; io_uring_sqe_set_data(sqe, NULL); @@ -462,7 +491,6 @@ msghdr->in_cancel_queue = false; } io_uring_submit(&orig_ring->ring); - } /*! will submit the next to-be-transmitted message for given iofd */ @@ -589,7 +617,7 @@ if (&ring->ring == &g_ring->ring) osmo_io_uring_submit(); /* If the submission queue is full, use cancel queue. We cannot cancel SQEs on the new ring. */ - sqe = io_uring_get_sqe(&ring->ring); + sqe = io_uring_get_sqe_and_count(ring); if (sqe) { io_uring_sqe_set_data(sqe, NULL); LOGPIO(iofd, LOGL_DEBUG, "Cancelling read\n"); @@ -619,7 +647,7 @@ if (&ring->ring == &g_ring->ring) osmo_io_uring_submit(); /* If the submission queue is full, use cancel queue. We cannot cancel SQEs on the new ring. */ - sqe = io_uring_get_sqe(&ring->ring); + sqe = io_uring_get_sqe_and_count(ring); if (sqe) { io_uring_sqe_set_data(sqe, NULL); LOGPIO(iofd, LOGL_DEBUG, "Cancelling write\n"); -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40855?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: Id2d2a0400ad442198c684ea0ead4eaeaead4c53d Gerrit-Change-Number: 40855 Gerrit-PatchSet: 7 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

[M] Change in libosmocore[master]: Send multiple read/recvfrom/recvmsg SQEs in advance

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40494?usp=email ) Change subject: Send multiple read/recvfrom/recvmsg SQEs in advance ...................................................................... Send multiple read/recvfrom/recvmsg SQEs in advance Multiple read or recvfrom operations can be submitted via SQEs when using io_uring. This allows reading multiple packet / more data between calls of osmo_select_main() the main loop. Having more than one SQE submitted is optional and the number can be controlled via environment variable. Related: OS#6705 Change-Id: Id50a8900fa2fe6de553e5025feae7e1e8d501e30 --- M include/osmocom/core/osmo_io.h M src/core/libosmocore.map M src/core/osmo_io.c M src/core/osmo_io_internal.h M src/core/osmo_io_uring.c 5 files changed, 103 insertions(+), 14 deletions(-) Approvals: pespin: Looks good to me, approved laforge: Looks good to me, but someone else must approve Jenkins Builder: Verified diff --git a/include/osmocom/core/osmo_io.h b/include/osmocom/core/osmo_io.h index 0b3dc83..0733240 100644 --- a/include/osmocom/core/osmo_io.h +++ b/include/osmocom/core/osmo_io.h @@ -213,6 +213,7 @@ enum osmo_io_fd_mode mode, const struct osmo_io_ops *ioops, void *data); int osmo_iofd_set_cmsg_size(struct osmo_io_fd *iofd, size_t cmsg_size); int osmo_iofd_set_io_buffers(struct osmo_io_fd *iofd, enum osmo_io_op op, uint8_t buffers); +int osmo_iofd_set_sqes(struct osmo_io_fd *iofd, enum osmo_io_op op, uint8_t sqes); int osmo_iofd_register(struct osmo_io_fd *iofd, int fd); int osmo_iofd_unregister(struct osmo_io_fd *iofd); unsigned int osmo_iofd_txqueue_len(struct osmo_io_fd *iofd); diff --git a/src/core/libosmocore.map b/src/core/libosmocore.map index 239597c..198b31e 100644 --- a/src/core/libosmocore.map +++ b/src/core/libosmocore.map @@ -276,6 +276,7 @@ osmo_iofd_set_alloc_info; osmo_iofd_set_cmsg_size; osmo_iofd_set_io_buffers; +osmo_iofd_set_sqes; osmo_iofd_set_data; osmo_iofd_set_ioops; osmo_iofd_set_priv_nr; diff --git a/src/core/osmo_io.c b/src/core/osmo_io.c index 62d2362..2f3af06 100644 --- a/src/core/osmo_io.c +++ b/src/core/osmo_io.c @@ -816,6 +816,14 @@ iofd->io_read_buffers = 1; iofd->io_write_buffers = 1; + if (osmo_iofd_ops.setup) { + int rc = osmo_iofd_ops.setup(iofd); + if (rc < 0) { + osmo_iofd_free(iofd); + return NULL; + } + } + return iofd; } @@ -868,6 +876,35 @@ return 0; } +/*! Set the number of SQEs that are submitted to an io_unring before completion is received. + * + * If the io_using backend is selected, this API function can be used to tell the osmo_io process how many SQE are + * scheduled in advance. + * The feature is currently supports scheduling read SQEs only. + * + * \param[in] iofd the iofd file descriptor + * \param[in] op the osmo_io_op (read) to set the number of IO buffers for + * \param[in] number of scheduled SQEs + * \returns zero on success, a negative value on error + */ +int osmo_iofd_set_sqes(struct osmo_io_fd *iofd, enum osmo_io_op op, uint8_t sqes) +{ + if (iofd->mode != OSMO_IO_FD_MODE_READ_WRITE) + return -EINVAL; + + if (g_io_backend != OSMO_IO_BACKEND_IO_URING) + return -EINVAL; + + if (op != OSMO_IO_OP_READ) + return -EINVAL; + + if (sqes < 1 || sqes > IOFD_MSGHDR_MAX_READ_SQES) + return -EINVAL; + + iofd->u.uring.num_read_sqes = sqes; + return 0; +} + /*! Register the osmo_io_fd for active I/O. * * Calling this function will register a previously initialized osmo_io_fd for performing I/O. diff --git a/src/core/osmo_io_internal.h b/src/core/osmo_io_internal.h index d28e3d7..f983b77 100644 --- a/src/core/osmo_io_internal.h +++ b/src/core/osmo_io_internal.h @@ -25,6 +25,7 @@ #endif struct iofd_backend_ops { + int (*setup)(struct osmo_io_fd *iofd); int (*register_fd)(struct osmo_io_fd *iofd); int (*unregister_fd)(struct osmo_io_fd *iofd); int (*close)(struct osmo_io_fd *iofd); @@ -49,6 +50,8 @@ #define IOFD_FLAG_ISSET(iofd, flag) ((iofd)->flags & (flag)) +#define IOFD_MSGHDR_MAX_READ_SQES 32 + struct osmo_io_fd { /*! linked list for internal management */ struct llist_head list; @@ -107,7 +110,12 @@ struct { bool read_enabled; bool write_enabled; - void *read_msghdr; + /*! requested number of simultaniously submitted read SQEs */ + uint8_t num_read_sqes; + /*! array of simultaneously submitted read SQEs */ + void *read_msghdr[IOFD_MSGHDR_MAX_READ_SQES]; + /*! current number of simultaneously submitted read SQEs */ + uint8_t reads_submitted; void *write_msghdr; /* TODO: index into array of registered fd's? */ /* osmo_fd for non-blocking connect handling */ diff --git a/src/core/osmo_io_uring.c b/src/core/osmo_io_uring.c index 018bdab..1eaa5e6 100644 --- a/src/core/osmo_io_uring.c +++ b/src/core/osmo_io_uring.c @@ -56,9 +56,13 @@ #define OSMO_IO_URING_BATCH "LIBOSMO_IO_URING_BATCH" +#define OSMO_IO_URING_READ_SQE "LIBOSMO_IO_URING_READ_SQE" + bool g_io_uring_batch = false; bool g_io_uring_submit_needed = false; +static int g_io_uring_read_sqes = 1; + struct osmo_io_uring { struct osmo_fd event_ofd; struct io_uring ring; @@ -103,6 +107,15 @@ if (rc < 0) osmo_panic("failure during io_uring_queue_init(): %s\n", strerror(-rc)); + if ((env = getenv(OSMO_IO_URING_READ_SQE))) { + g_io_uring_read_sqes = atoi(env); + if (g_io_uring_read_sqes < 1 || g_io_uring_read_sqes > IOFD_MSGHDR_MAX_READ_SQES) { + fprintf(stderr, "Invalid osmo_uring read SQEs requested: \"%s\"\n Allowed range: 1..%d\n", + env, IOFD_MSGHDR_MAX_READ_SQES); + exit(1); + } + } + rc = eventfd(0, 0); if (rc < 0) { io_uring_queue_exit(&g_ring.ring); @@ -134,7 +147,7 @@ g_io_uring_submit_needed = true; } -static void iofd_uring_submit_recv(struct osmo_io_fd *iofd, enum iofd_msg_action action) +static void iofd_uring_submit_recv_sqe(struct osmo_io_fd *iofd, enum iofd_msg_action action) { struct msgb *msg; struct iofd_msghdr *msghdr; @@ -196,15 +209,37 @@ iofd_io_uring_submit(); - /* NOTE: This only works if we have one read per fd */ - iofd->u.uring.read_msghdr = msghdr; + iofd->u.uring.read_msghdr[iofd->u.uring.reads_submitted] = msghdr; + iofd->u.uring.reads_submitted++; +} + +static void iofd_uring_submit_recv(struct osmo_io_fd *iofd, enum iofd_msg_action action) +{ + /* Submit more read SQEs in advance, if requested. */ + while (iofd->u.uring.reads_submitted < iofd->u.uring.num_read_sqes) + iofd_uring_submit_recv_sqe(iofd, action); } /*! completion call-back for READ/RECVFROM */ static void iofd_uring_handle_recv(struct iofd_msghdr *msghdr, int rc) { struct osmo_io_fd *iofd = msghdr->iofd; - uint8_t idx; + uint8_t idx, i; + + /* Find which read_msghdr is completed and remove from list. */ + for (idx = 0; idx < iofd->u.uring.reads_submitted; idx++) { + if (iofd->u.uring.read_msghdr[idx] == msghdr) + break; + } + if (idx == iofd->u.uring.reads_submitted) { + LOGP(DLIO, LOGL_FATAL, "Read SQE completion, but msghdr not found, please fix!\n"); + return; + } + /* Remove entry at idx. */ + iofd->u.uring.reads_submitted--; + for (i = idx; i < iofd->u.uring.reads_submitted; i++) + iofd->u.uring.read_msghdr[i] = iofd->u.uring.read_msghdr[i + 1]; + iofd->u.uring.read_msghdr[i] = NULL; for (idx = 0; idx < msghdr->io_len; idx++) { struct msgb *msg = msghdr->msg[idx]; @@ -238,9 +273,6 @@ if (iofd->u.uring.read_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_uring_submit_recv(iofd, msghdr->action); - else - iofd->u.uring.read_msghdr = NULL; - iofd_msghdr_free(msghdr); } @@ -300,7 +332,7 @@ IOFD_FLAG_UNSET(iofd, IOFD_FLAG_IN_CALLBACK); - if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.read_msghdr && !iofd->u.uring.write_msghdr) + if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.reads_submitted && !iofd->u.uring.write_msghdr) talloc_free(iofd); } @@ -410,11 +442,18 @@ if (iofd->u.uring.read_enabled && !IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED)) iofd_uring_read_enable(iofd); - if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.read_msghdr && !iofd->u.uring.write_msghdr) + if (IOFD_FLAG_ISSET(iofd, IOFD_FLAG_TO_FREE) && !iofd->u.uring.reads_submitted && !iofd->u.uring.write_msghdr) talloc_free(iofd); return 0; } +static int iofd_uring_setup(struct osmo_io_fd *iofd) +{ + iofd->u.uring.num_read_sqes = g_io_uring_read_sqes; + + return 0; +} + static int iofd_uring_register(struct osmo_io_fd *iofd) { if (iofd->mode != OSMO_IO_FD_MODE_RECVMSG_SENDMSG) @@ -440,18 +479,20 @@ { struct io_uring_sqe *sqe; struct iofd_msghdr *msghdr; + uint8_t idx; - if (iofd->u.uring.read_msghdr) { - msghdr = iofd->u.uring.read_msghdr; + for (idx = 0; idx < iofd->u.uring.reads_submitted; idx++) { + msghdr = iofd->u.uring.read_msghdr[idx]; + iofd->u.uring.read_msghdr[idx] = NULL; sqe = io_uring_get_sqe(&g_ring.ring); OSMO_ASSERT(sqe != NULL); io_uring_sqe_set_data(sqe, NULL); LOGPIO(iofd, LOGL_DEBUG, "Cancelling read\n"); - iofd->u.uring.read_msghdr = NULL; talloc_steal(OTC_GLOBAL, msghdr); msghdr->iofd = NULL; io_uring_prep_cancel(sqe, msghdr, 0); } + iofd->u.uring.reads_submitted = 0; if (iofd->u.uring.write_msghdr) { msghdr = iofd->u.uring.write_msghdr; @@ -533,7 +574,7 @@ { iofd->u.uring.read_enabled = true; - if (iofd->u.uring.read_msghdr) + if (iofd->u.uring.reads_submitted) return; /* This function is called again, once the socket is connected. */ @@ -580,6 +621,7 @@ } const struct iofd_backend_ops iofd_uring_ops = { + .setup = iofd_uring_setup, .register_fd = iofd_uring_register, .unregister_fd = iofd_uring_unregister, .close = iofd_uring_close, -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40494?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: Id50a8900fa2fe6de553e5025feae7e1e8d501e30 Gerrit-Change-Number: 40494 Gerrit-PatchSet: 16 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

[S] Change in libosmocore[master]: Add environment variable to set io_uring size

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40782?usp=email ) Change subject: Add environment variable to set io_uring size ...................................................................... Add environment variable to set io_uring size The term "LIBOSMO_IO_URING_INITIAL_SIZE" is related to the following patch, which will increment the size of the io_uring automatically if the initial size is too small. Related: OS#6705 Change-Id: I55289d9282e13aa1bf82f3931c85c196752f1484 --- M src/core/osmo_io_uring.c 1 file changed, 23 insertions(+), 2 deletions(-) Approvals: pespin: Looks good to me, but someone else must approve lynxis lazus: Looks good to me, but someone else must approve laforge: Looks good to me, approved Jenkins Builder: Verified diff --git a/src/core/osmo_io_uring.c b/src/core/osmo_io_uring.c index 1eaa5e6..48f6f0d 100644 --- a/src/core/osmo_io_uring.c +++ b/src/core/osmo_io_uring.c @@ -52,15 +52,21 @@ #include "osmo_io_internal.h" -#define IOFD_URING_ENTRIES 4096 +#define IOFD_URING_INITIAL_SIZE 4096 +/* 32768 refers to the IORING_MAX_ENTRIES of the kernel (io_uring/io_uring.h). */ +#define IOFD_URING_MAXIMUM_SIZE 32768 #define OSMO_IO_URING_BATCH "LIBOSMO_IO_URING_BATCH" +#define OSMO_IO_URING_INITIAL_SIZE "LIBOSMO_IO_URING_INITIAL_SIZE" + #define OSMO_IO_URING_READ_SQE "LIBOSMO_IO_URING_READ_SQE" bool g_io_uring_batch = false; bool g_io_uring_submit_needed = false; +static unsigned int g_io_uring_size = IOFD_URING_INITIAL_SIZE; + static int g_io_uring_read_sqes = 1; struct osmo_io_uring { @@ -103,7 +109,22 @@ if ((env = getenv(OSMO_IO_URING_BATCH))) g_io_uring_batch = true; - rc = io_uring_queue_init(IOFD_URING_ENTRIES, &g_ring.ring, 0); + if ((env = getenv(OSMO_IO_URING_INITIAL_SIZE))) { + int env_value; + rc = osmo_str_to_int(&env_value, env, 10, 1, IOFD_URING_MAXIMUM_SIZE); + if (rc < 0) { + fprintf(stderr, "Error: Initial io_uring size out of range (1..%d).\n", + IOFD_URING_MAXIMUM_SIZE); + exit(1); + } + if ((env_value & (env_value - 1))) { + fprintf(stderr, "Error: Initial io_uring size must be a positive power of two.\n"); + exit(1); + } + g_io_uring_size = env_value; + } + + rc = io_uring_queue_init(g_io_uring_size, &g_ring.ring, 0); if (rc < 0) osmo_panic("failure during io_uring_queue_init(): %s\n", strerror(-rc)); -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40782?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: I55289d9282e13aa1bf82f3931c85c196752f1484 Gerrit-Change-Number: 40782 Gerrit-PatchSet: 8 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: lynxis lazus <lynxis(a)fe80.eu> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

[M] Change in libosmocore[master]: Add multiple messages buffers to io_uring write operations

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40493?usp=email ) ( 10 is the latest approved patch-set. No files were changed between the latest approved patch-set and the submitted one. )Change subject: Add multiple messages buffers to io_uring write operations ...................................................................... Add multiple messages buffers to io_uring write operations Multiple message buffers can be writen by sending a single SQE when using io_uring. If there is less data written, the completely written buffers are removed and the partly written buffers are truncated. Afterwards they are re-queued for next write operation. Having more than one buffer is optional and the number can be controlled via environment variable. Related: OS#6705 Change-Id: I8c4e0a785cf66becd7fb5b2caf718c9724b56686 --- M src/core/osmo_io.c M src/core/osmo_io_uring.c 2 files changed, 103 insertions(+), 46 deletions(-) Approvals: laforge: Looks good to me, but someone else must approve pespin: Looks good to me, approved Jenkins Builder: Verified diff --git a/src/core/osmo_io.c b/src/core/osmo_io.c index 46885aa..bfc75c6 100644 --- a/src/core/osmo_io.c +++ b/src/core/osmo_io.c @@ -417,42 +417,74 @@ */ void iofd_handle_send_completion(struct osmo_io_fd *iofd, int rc, struct iofd_msghdr *msghdr) { - struct msgb *msg = msghdr->msg[0]; + int idx, i; - /* Incomplete write */ - if (rc > 0 && rc < msgb_length(msg)) { - /* Re-enqueue remaining data */ - msgb_pull(msg, rc); - msghdr->iov[0].iov_len = msgb_length(msg); - iofd_txqueue_enqueue_front(iofd, msghdr); - return; - } - - /* Reenqueue the complete msgb */ + /* Re-enqueue the complete msgb. */ if (rc == -EAGAIN) { iofd_txqueue_enqueue_front(iofd, msghdr); return; } - /* All other failure and success cases are handled here */ - switch (msghdr->action) { - case IOFD_ACT_WRITE: - if (iofd->io_ops.write_cb) - iofd->io_ops.write_cb(iofd, rc, msg); - break; - case IOFD_ACT_SENDTO: - if (iofd->io_ops.sendto_cb) - iofd->io_ops.sendto_cb(iofd, rc, msg, &msghdr->osa); - break; - case IOFD_ACT_SENDMSG: - if (iofd->io_ops.sendmsg_cb) - iofd->io_ops.sendmsg_cb(iofd, rc, msg); - break; - default: - OSMO_ASSERT(0); - } + for (idx = 0; idx < msghdr->io_len; idx++) { + struct msgb *msg = msghdr->msg[idx]; + int chunk; - msgb_free(msghdr->msg[0]); + /* Incomplete write */ + if (rc > 0 && rc < msgb_length(msg)) { + /* Keep msg with unsent data only. */ + msgb_pull(msg, rc); + msghdr->iov[idx].iov_len = msgb_length(msg); + /* Shift all existing buffers down. */ + if (idx) { + msghdr->io_len -= idx; + for (i = 0; i < msghdr->io_len; i++) { + msghdr->iov[i] = msghdr->iov[idx + i]; + msghdr->msg[i] = msghdr->msg[idx + i]; + } + for (i = 0; i < idx; i++) { + memset(&msghdr->iov[msghdr->io_len + i], 0, sizeof(struct iovec)); + msghdr->msg[msghdr->io_len + i] = NULL; + } + msghdr->hdr.msg_iovlen = msghdr->io_len; + } + /* Re-enqueue remaining buffers. */ + iofd_txqueue_enqueue_front(iofd, msghdr); + return; + } + + if (rc >= 0) { + chunk = msgb_length(msg); + if (rc < chunk) + chunk = rc; + } else { + chunk = rc; + } + + /* All other failure and success cases are handled here */ + switch (msghdr->action) { + case IOFD_ACT_WRITE: + if (iofd->io_ops.write_cb) + iofd->io_ops.write_cb(iofd, chunk, msg); + break; + case IOFD_ACT_SENDTO: + if (iofd->io_ops.sendto_cb) + iofd->io_ops.sendto_cb(iofd, chunk, msg, &msghdr->osa); + break; + case IOFD_ACT_SENDMSG: + if (iofd->io_ops.sendmsg_cb) + iofd->io_ops.sendmsg_cb(iofd, chunk, msg); + break; + default: + OSMO_ASSERT(0); + } + + msgb_free(msghdr->msg[idx]); + msghdr->msg[idx] = NULL; + + /* The user can unregister/close the iofd during callback above. */ + if (!IOFD_FLAG_ISSET(iofd, IOFD_FLAG_FD_REGISTERED)) + break; + } iofd_msghdr_free(msghdr); } @@ -475,6 +507,8 @@ int osmo_iofd_write_msgb(struct osmo_io_fd *iofd, struct msgb *msg) { int rc; + struct iofd_msghdr *msghdr; + int idx; if (OSMO_UNLIKELY(msgb_length(msg) == 0)) { LOGPIO(iofd, LOGL_ERROR, "Length is 0, rejecting msgb.\n"); @@ -483,21 +517,35 @@ OSMO_ASSERT(iofd->mode == OSMO_IO_FD_MODE_READ_WRITE); - struct iofd_msghdr *msghdr = iofd_msghdr_alloc(iofd, IOFD_ACT_WRITE, msg, 0); - if (!msghdr) - return -ENOMEM; + /* Always try to add msg to last msghdr. Only if it is completely filled, allocate a new msghdr. + * This way all the previous meghdrs in the queue are completely filled. */ + msghdr = llist_last_entry_or_null(&iofd->tx_queue.msg_queue, struct iofd_msghdr, list); + if (msghdr && msghdr->io_len < iofd->io_write_buffers) { + /* Add msg to existing msghdr. */ + msghdr->msg[msghdr->io_len++] = msg; + } else { + /* Create new msghdr with msg. */ + msghdr = iofd_msghdr_alloc(iofd, IOFD_ACT_WRITE, msg, 0); + if (!msghdr) + return -ENOMEM; + msghdr->hdr.msg_iov = &msghdr->iov[0]; + msghdr->flags = MSG_NOSIGNAL; + } - msghdr->flags = MSG_NOSIGNAL; - msghdr->iov[0].iov_base = msgb_data(msghdr->msg[0]); - msghdr->iov[0].iov_len = msgb_length(msghdr->msg[0]); - msghdr->hdr.msg_iov = &msghdr->iov[0]; - msghdr->hdr.msg_iovlen = 1; + /* Add set IO vector to msg. */ + idx = msghdr->io_len - 1; + msghdr->iov[idx].iov_base = msgb_data(msg); + msghdr->iov[idx].iov_len = msgb_length(msg); + msghdr->hdr.msg_iovlen = msghdr->io_len; - rc = iofd_txqueue_enqueue(iofd, msghdr); - if (rc < 0) { - iofd_msghdr_free(msghdr); - LOGPIO(iofd, LOGL_ERROR, "enqueueing message failed (%d). Rejecting msgb\n", rc); - return rc; + /* Only new msghdr will be enqueued. */ + if (msghdr->io_len == 1) { + rc = iofd_txqueue_enqueue(iofd, msghdr); + if (rc < 0) { + iofd_msghdr_free(msghdr); + LOGPIO(iofd, LOGL_ERROR, "enqueueing message failed (%d). Rejecting msgb\n", rc); + return rc; + } } return 0; @@ -884,7 +932,10 @@ { struct iofd_msghdr *hdr; while ((hdr = iofd_txqueue_dequeue(iofd))) { - msgb_free(hdr->msg[0]); + for (int idx = 0; idx < hdr->io_len; idx++) { + msgb_free(hdr->msg[idx]); + hdr->msg[idx] = NULL; + } iofd_msghdr_free(hdr); } } diff --git a/src/core/osmo_io_uring.c b/src/core/osmo_io_uring.c index 754a39a..018bdab 100644 --- a/src/core/osmo_io_uring.c +++ b/src/core/osmo_io_uring.c @@ -262,7 +262,10 @@ iofd->u.uring.write_msghdr = NULL; if (OSMO_UNLIKELY(IOFD_FLAG_ISSET(iofd, IOFD_FLAG_CLOSED))) { - msgb_free(msghdr->msg[0]); + for (int idx = 0; idx < msghdr->io_len; idx++) { + msgb_free(msghdr->msg[idx]); + msghdr->msg[idx] = NULL; + } iofd_msghdr_free(msghdr); } else { iofd_handle_send_completion(iofd, rc, msghdr); @@ -351,7 +354,7 @@ switch (msghdr->action) { case IOFD_ACT_WRITE: - io_uring_prep_writev(sqe, msghdr->iofd->fd, msghdr->iov, 1, -1); + io_uring_prep_writev(sqe, msghdr->iofd->fd, msghdr->iov, msghdr->io_len, -1); break; case IOFD_ACT_SENDTO: case IOFD_ACT_SENDMSG: @@ -458,7 +461,10 @@ LOGPIO(iofd, LOGL_DEBUG, "Cancelling write\n"); iofd->u.uring.write_msghdr = NULL; talloc_steal(OTC_GLOBAL, msghdr); - msgb_free(msghdr->msg[0]); + for (int idx = 0; idx < msghdr->io_len; idx++) { + msgb_free(msghdr->msg[idx]); + msghdr->msg[idx] = NULL; + } msghdr->iofd = NULL; io_uring_prep_cancel(sqe, msghdr, 0); } -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40493?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: I8c4e0a785cf66becd7fb5b2caf718c9724b56686 Gerrit-Change-Number: 40493 Gerrit-PatchSet: 14 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

[S] Change in libosmocore[master]: osmo-io: Put together message buffers when dequeued from tx queue

by laforge

laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/40760?usp=email ) Change subject: osmo-io: Put together message buffers when dequeued from tx queue ...................................................................... osmo-io: Put together message buffers when dequeued from tx queue Write operations may be incomplete. osmo-io processs will remove complete message buffers after a write operation from msghdr and put the msghdr with the remaining buffers back into tx_queue. If the user requests multiple messages buffers per write operation, the msghdr of an incomplete write may have less message buffers than the user requested. The remaining buffers are buffers are taken from next msghdr in the queue, if exists. Change-Id: I97c366211dd266fd58ec252890ec017d6d334534 --- M src/core/osmo_io.c 1 file changed, 36 insertions(+), 5 deletions(-) Approvals: pespin: Looks good to me, approved laforge: Looks good to me, but someone else must approve Jenkins Builder: Verified diff --git a/src/core/osmo_io.c b/src/core/osmo_io.c index bfc75c6..62d2362 100644 --- a/src/core/osmo_io.c +++ b/src/core/osmo_io.c @@ -247,21 +247,52 @@ */ struct iofd_msghdr *iofd_txqueue_dequeue(struct osmo_io_fd *iofd) { - struct llist_head *lh; + struct iofd_msghdr *msghdr; if (iofd->tx_queue.current_length == 0) return NULL; - lh = iofd->tx_queue.msg_queue.next; + msghdr = llist_first_entry_or_null(&iofd->tx_queue.msg_queue, struct iofd_msghdr, list); - OSMO_ASSERT(lh); + OSMO_ASSERT(msghdr); iofd->tx_queue.current_length--; - llist_del(lh); + llist_del(&msghdr->list); + + /* Fill up empty buffers in dequeued msghdr with buffers from the next msghdr. + * There can be empty buffers, when a msghdr is queued to the front with incomplete write. */ + while (OSMO_UNLIKELY(msghdr->io_len < iofd->io_write_buffers)) { + struct iofd_msghdr *next; + int i; + + if (iofd->tx_queue.current_length == 0) + break; + next = llist_first_entry_or_null(&iofd->tx_queue.msg_queue, struct iofd_msghdr, list); + OSMO_ASSERT(next->io_len > 0); + /* Get first message buffer from next msghdr and store them in the dequeued one. */ + msghdr->iov[msghdr->io_len] = next->iov[0]; + msghdr->msg[msghdr->io_len] = next->msg[0]; + msghdr->hdr.msg_iovlen = ++msghdr->io_len; + /* Remove the message buffer from the next msghdr and free, if empty. */ + next->io_len--; + for (i = 0; i < next->io_len; i++) { + next->iov[i] = next->iov[i + 1]; + next->msg[i] = next->msg[i + 1]; + } + if (next->io_len == 0) { + iofd->tx_queue.current_length--; + llist_del(&next->list); + iofd_msghdr_free(next); + } else { + memset(&next->iov[next->io_len], 0, sizeof(struct iovec)); + next->msg[next->io_len] = NULL; + next->hdr.msg_iovlen = --next->io_len; + } + } if (iofd->tx_queue.current_length == 0) osmo_iofd_ops.write_disable(iofd); - return llist_entry(lh, struct iofd_msghdr, list); + return msghdr; } /*! Handle segmentation of the msg. If this function returns *_HANDLE_ONE or MORE then the data in msg will contain -- To view, visit https://gerrit.osmocom.org/c/libosmocore/+/40760?usp=email To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings?usp=email Gerrit-MessageType: merged Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Change-Id: I97c366211dd266fd58ec252890ec017d6d334534 Gerrit-Change-Number: 40760 Gerrit-PatchSet: 7 Gerrit-Owner: jolly <andreas(a)eversberg.eu> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <laforge(a)osmocom.org> Gerrit-Reviewer: pespin <pespin(a)sysmocom.de>

3 days, 2 hours

1
0
0 0

2025

2024

2023

2022

gerrit-log September 2025