From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 9ABEEA328D for ; Wed, 23 Oct 2019 00:51:58 +0200 (CEST) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id 98B0C1BF2F; Wed, 23 Oct 2019 00:51:57 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by dpdk.org (Postfix) with ESMTP id 1A6D01BF2E for ; Wed, 23 Oct 2019 00:51:55 +0200 (CEST) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga003.fm.intel.com ([10.253.24.29]) by orsmga102.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 22 Oct 2019 15:51:54 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.68,218,1569308400"; d="scan'208";a="203793492" Received: from irsmsx108.ger.corp.intel.com ([163.33.3.3]) by FMSMGA003.fm.intel.com with ESMTP; 22 Oct 2019 15:51:52 -0700 Received: from irsmsx111.ger.corp.intel.com (10.108.20.4) by IRSMSX108.ger.corp.intel.com (163.33.3.3) with Microsoft SMTP Server (TLS) id 14.3.439.0; Tue, 22 Oct 2019 23:51:51 +0100 Received: from irsmsx104.ger.corp.intel.com ([169.254.5.252]) by irsmsx111.ger.corp.intel.com ([169.254.2.20]) with mapi id 14.03.0439.000; Tue, 22 Oct 2019 23:51:51 +0100 From: "Ananyev, Konstantin" To: Olivier Matz , "dev@dpdk.org" CC: Andrew Rybchenko , "Richardson, Bruce" , "Wang, Haiyue" , "Jerin Jacob Kollanukkaran" , "Wiles, Keith" , =?iso-8859-1?Q?Morten_Br=F8rup?= , Stephen Hemminger , Thomas Monjalon Thread-Topic: [PATCH v2] mbuf: support dynamic fields and flags Thread-Index: AQHVhPkmj10y/x3Gj0mcCTjWlAS5/6dnRDsA Date: Tue, 22 Oct 2019 22:51:51 +0000 Message-ID: <2601191342CEEE43887BDE71AB97725801A8C6E2BD@IRSMSX104.ger.corp.intel.com> References: <20190710092907.5565-1-olivier.matz@6wind.com> <20191017144219.32708-1-olivier.matz@6wind.com> In-Reply-To: <20191017144219.32708-1-olivier.matz@6wind.com> Accept-Language: en-IE, en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-titus-metadata-40: eyJDYXRlZ29yeUxhYmVscyI6IiIsIk1ldGFkYXRhIjp7Im5zIjoiaHR0cDpcL1wvd3d3LnRpdHVzLmNvbVwvbnNcL0ludGVsMyIsImlkIjoiY2I1Y2E3NjktMDY5OS00ZjZkLWI3NzAtY2M4OThmNjg5ZTMwIiwicHJvcHMiOlt7Im4iOiJDVFBDbGFzc2lmaWNhdGlvbiIsInZhbHMiOlt7InZhbHVlIjoiQ1RQX05UIn1dfV19LCJTdWJqZWN0TGFiZWxzIjpbXSwiVE1DVmVyc2lvbiI6IjE3LjEwLjE4MDQuNDkiLCJUcnVzdGVkTGFiZWxIYXNoIjoiNzdQcHJ6WUR0M3lGODN1d0FcLzIzZ2R3NGxiWGtydVFYQkRWNFwvSDl6Sk82U2N4bk1qYjlMb2tZNm9FcFhNRTdWIn0= x-ctpclassification: CTP_NT dlp-product: dlpe-windows dlp-version: 11.2.0.6 dlp-reaction: no-action x-originating-ip: [163.33.239.180] Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Subject: Re: [dpdk-dev] [PATCH v2] mbuf: support dynamic fields and flags X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" > Many features require to store data inside the mbuf. As the room in mbuf > structure is limited, it is not possible to have a field for each > feature. Also, changing fields in the mbuf structure can break the API > or ABI. >=20 > This commit addresses these issues, by enabling the dynamic registration > of fields or flags: >=20 > - a dynamic field is a named area in the rte_mbuf structure, with a > given size (>=3D 1 byte) and alignment constraint. > - a dynamic flag is a named bit in the rte_mbuf structure. >=20 > The typical use case is a PMD that registers space for an offload > feature, when the application requests to enable this feature. As > the space in mbuf is limited, the space should only be reserved if it > is going to be used (i.e when the application explicitly asks for it). >=20 > The registration can be done at any moment, but it is not possible > to unregister fields or flags for now. >=20 > Signed-off-by: Olivier Matz > Acked-by: Thomas Monjalon > --- >=20 > v2 >=20 > * Rebase on top of master: solve conflict with Stephen's patchset > (packet copy) > * Add new apis to register a dynamic field/flag at a specific place > * Add a dump function (sugg by David) > * Enhance field registration function to select the best offset, keeping > large aligned zones as much as possible (sugg by Konstantin) > * Use a size_t and unsigned int instead of int when relevant > (sugg by Konstantin) > * Use "uint64_t dynfield1[2]" in mbuf instead of 2 uint64_t fields > (sugg by Konstantin) > * Remove unused argument in private function (sugg by Konstantin) > * Fix and simplify locking (sugg by Konstantin) > * Fix minor typo >=20 > rfc -> v1 >=20 > * Rebase on top of master > * Change registration API to use a structure instead of > variables, getting rid of #defines (Stephen's comment) > * Update flag registration to use a similar API as fields. > * Change max name length from 32 to 64 (sugg. by Thomas) > * Enhance API documentation (Haiyue's and Andrew's comments) > * Add a debug log at registration > * Add some words in release note > * Did some performance tests (sugg. by Andrew): > On my platform, reading a dynamic field takes ~3 cycles more > than a static field, and ~2 cycles more for writing. >=20 > app/test/test_mbuf.c | 145 ++++++- > doc/guides/rel_notes/release_19_11.rst | 7 + > lib/librte_mbuf/Makefile | 2 + > lib/librte_mbuf/meson.build | 6 +- > lib/librte_mbuf/rte_mbuf.h | 23 +- > lib/librte_mbuf/rte_mbuf_dyn.c | 548 +++++++++++++++++++++++++ > lib/librte_mbuf/rte_mbuf_dyn.h | 226 ++++++++++ > lib/librte_mbuf/rte_mbuf_version.map | 7 + > 8 files changed, 959 insertions(+), 5 deletions(-) > create mode 100644 lib/librte_mbuf/rte_mbuf_dyn.c > create mode 100644 lib/librte_mbuf/rte_mbuf_dyn.h >=20 > diff --git a/app/test/test_mbuf.c b/app/test/test_mbuf.c > index b9c2b2500..01cafad59 100644 > --- a/app/test/test_mbuf.c > +++ b/app/test/test_mbuf.c > @@ -28,6 +28,7 @@ > #include > #include > #include > +#include >=20 > #include "test.h" >=20 > @@ -657,7 +658,6 @@ test_attach_from_different_pool(struct rte_mempool *p= ktmbuf_pool, > rte_pktmbuf_free(clone2); > return -1; > } > -#undef GOTO_FAIL >=20 > /* > * test allocation and free of mbufs > @@ -1276,6 +1276,143 @@ test_tx_offload(void) > return (v1 =3D=3D v2) ? 0 : -EINVAL; > } >=20 > +static int > +test_mbuf_dyn(struct rte_mempool *pktmbuf_pool) > +{ > + const struct rte_mbuf_dynfield dynfield =3D { > + .name =3D "test-dynfield", > + .size =3D sizeof(uint8_t), > + .align =3D __alignof__(uint8_t), > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynfield dynfield2 =3D { > + .name =3D "test-dynfield2", > + .size =3D sizeof(uint16_t), > + .align =3D __alignof__(uint16_t), > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynfield dynfield3 =3D { > + .name =3D "test-dynfield3", > + .size =3D sizeof(uint8_t), > + .align =3D __alignof__(uint8_t), > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynfield dynfield_fail_big =3D { > + .name =3D "test-dynfield-fail-big", > + .size =3D 256, > + .align =3D 1, > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynfield dynfield_fail_align =3D { > + .name =3D "test-dynfield-fail-align", > + .size =3D 1, > + .align =3D 3, > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynflag dynflag =3D { > + .name =3D "test-dynflag", > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynflag dynflag2 =3D { > + .name =3D "test-dynflag2", > + .flags =3D 0, > + }; > + const struct rte_mbuf_dynflag dynflag3 =3D { > + .name =3D "test-dynflag3", > + .flags =3D 0, > + }; > + struct rte_mbuf *m =3D NULL; > + int offset, offset2, offset3; > + int flag, flag2, flag3; > + int ret; > + > + printf("Test mbuf dynamic fields and flags\n"); > + rte_mbuf_dyn_dump(stdout); > + > + offset =3D rte_mbuf_dynfield_register(&dynfield); > + if (offset =3D=3D -1) > + GOTO_FAIL("failed to register dynamic field, offset=3D%d: %s", > + offset, strerror(errno)); > + > + ret =3D rte_mbuf_dynfield_register(&dynfield); > + if (ret !=3D offset) > + GOTO_FAIL("failed to lookup dynamic field, ret=3D%d: %s", > + ret, strerror(errno)); > + > + offset2 =3D rte_mbuf_dynfield_register(&dynfield2); > + if (offset2 =3D=3D -1 || offset2 =3D=3D offset || (offset2 & 1)) > + GOTO_FAIL("failed to register dynamic field 2, offset2=3D%d: %s", > + offset2, strerror(errno)); > + > + offset3 =3D rte_mbuf_dynfield_register_offset(&dynfield3, > + offsetof(struct rte_mbuf, dynfield1[1])); > + if (offset3 !=3D offsetof(struct rte_mbuf, dynfield1[1])) > + GOTO_FAIL("failed to register dynamic field 3, offset=3D%d: %s", > + offset3, strerror(errno)); > + > + printf("dynfield: offset=3D%d, offset2=3D%d, offset3=3D%d\n", > + offset, offset2, offset3); > + > + ret =3D rte_mbuf_dynfield_register(&dynfield_fail_big); > + if (ret !=3D -1) > + GOTO_FAIL("dynamic field creation should fail (too big)"); > + > + ret =3D rte_mbuf_dynfield_register(&dynfield_fail_align); > + if (ret !=3D -1) > + GOTO_FAIL("dynamic field creation should fail (bad alignment)"); > + > + ret =3D rte_mbuf_dynfield_register_offset(&dynfield_fail_align, > + offsetof(struct rte_mbuf, ol_flags)); > + if (ret !=3D -1) > + GOTO_FAIL("dynamic field creation should fail (not avail)"); > + > + flag =3D rte_mbuf_dynflag_register(&dynflag); > + if (flag =3D=3D -1) > + GOTO_FAIL("failed to register dynamic flag, flag=3D%d: %s", > + flag, strerror(errno)); > + > + ret =3D rte_mbuf_dynflag_register(&dynflag); > + if (ret !=3D flag) > + GOTO_FAIL("failed to lookup dynamic flag, ret=3D%d: %s", > + ret, strerror(errno)); > + > + flag2 =3D rte_mbuf_dynflag_register(&dynflag2); > + if (flag2 =3D=3D -1 || flag2 =3D=3D flag) > + GOTO_FAIL("failed to register dynamic flag 2, flag2=3D%d: %s", > + flag2, strerror(errno)); > + > + flag3 =3D rte_mbuf_dynflag_register_bitnum(&dynflag3, > + rte_bsf64(PKT_LAST_FREE)); > + if (flag3 !=3D rte_bsf64(PKT_LAST_FREE)) > + GOTO_FAIL("failed to register dynamic flag 3, flag2=3D%d: %s", > + flag3, strerror(errno)); > + > + printf("dynflag: flag=3D%d, flag2=3D%d, flag3=3D%d\n", flag, flag2, fla= g3); > + > + /* set, get dynamic field */ > + m =3D rte_pktmbuf_alloc(pktmbuf_pool); > + if (m =3D=3D NULL) > + GOTO_FAIL("Cannot allocate mbuf"); > + > + *RTE_MBUF_DYNFIELD(m, offset, uint8_t *) =3D 1; > + if (*RTE_MBUF_DYNFIELD(m, offset, uint8_t *) !=3D 1) > + GOTO_FAIL("failed to read dynamic field"); > + *RTE_MBUF_DYNFIELD(m, offset2, uint16_t *) =3D 1000; > + if (*RTE_MBUF_DYNFIELD(m, offset2, uint16_t *) !=3D 1000) > + GOTO_FAIL("failed to read dynamic field"); > + > + /* set a dynamic flag */ > + m->ol_flags |=3D (1ULL << flag); > + > + rte_mbuf_dyn_dump(stdout); > + rte_pktmbuf_free(m); > + return 0; > +fail: > + rte_pktmbuf_free(m); > + return -1; > +} > +#undef GOTO_FAIL > + > static int > test_mbuf(void) > { > @@ -1295,6 +1432,12 @@ test_mbuf(void) > goto err; > } >=20 > + /* test registration of dynamic fields and flags */ > + if (test_mbuf_dyn(pktmbuf_pool) < 0) { > + printf("mbuf dynflag test failed\n"); > + goto err; > + } > + > /* create a specific pktmbuf pool with a priv_size !=3D 0 and no data > * room size */ > pktmbuf_pool2 =3D rte_pktmbuf_pool_create("test_pktmbuf_pool2", > diff --git a/doc/guides/rel_notes/release_19_11.rst b/doc/guides/rel_note= s/release_19_11.rst > index 85953b962..9e9c94554 100644 > --- a/doc/guides/rel_notes/release_19_11.rst > +++ b/doc/guides/rel_notes/release_19_11.rst > @@ -21,6 +21,13 @@ DPDK Release 19.11 >=20 > xdg-open build/doc/html/guides/rel_notes/release_19_11.html >=20 > +* **Add support of support dynamic fields and flags in mbuf.** > + > + This new feature adds the ability to dynamically register some room > + for a field or a flag in the mbuf structure. This is typically used > + for specific offload features, where adding a static field or flag > + in the mbuf is not justified. > + >=20 > New Features > ------------ > diff --git a/lib/librte_mbuf/Makefile b/lib/librte_mbuf/Makefile > index c8f6d2689..5a9bcee73 100644 > --- a/lib/librte_mbuf/Makefile > +++ b/lib/librte_mbuf/Makefile > @@ -17,8 +17,10 @@ LIBABIVER :=3D 5 >=20 > # all source are stored in SRCS-y > SRCS-$(CONFIG_RTE_LIBRTE_MBUF) :=3D rte_mbuf.c rte_mbuf_ptype.c rte_mbuf= _pool_ops.c > +SRCS-$(CONFIG_RTE_LIBRTE_MBUF) +=3D rte_mbuf_dyn.c >=20 > # install includes > SYMLINK-$(CONFIG_RTE_LIBRTE_MBUF)-include :=3D rte_mbuf.h rte_mbuf_ptype= .h rte_mbuf_pool_ops.h > +SYMLINK-$(CONFIG_RTE_LIBRTE_MBUF)-include +=3D rte_mbuf_dyn.h >=20 > include $(RTE_SDK)/mk/rte.lib.mk > diff --git a/lib/librte_mbuf/meson.build b/lib/librte_mbuf/meson.build > index 6cc11ebb4..9137e8f26 100644 > --- a/lib/librte_mbuf/meson.build > +++ b/lib/librte_mbuf/meson.build > @@ -2,8 +2,10 @@ > # Copyright(c) 2017 Intel Corporation >=20 > version =3D 5 > -sources =3D files('rte_mbuf.c', 'rte_mbuf_ptype.c', 'rte_mbuf_pool_ops.c= ') > -headers =3D files('rte_mbuf.h', 'rte_mbuf_ptype.h', 'rte_mbuf_pool_ops.h= ') > +sources =3D files('rte_mbuf.c', 'rte_mbuf_ptype.c', 'rte_mbuf_pool_ops.c= ', > + 'rte_mbuf_dyn.c') > +headers =3D files('rte_mbuf.h', 'rte_mbuf_ptype.h', 'rte_mbuf_pool_ops.h= ', > + 'rte_mbuf_dyn.h') > deps +=3D ['mempool'] >=20 > allow_experimental_apis =3D true > diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib/librte_mbuf/rte_mbuf.h > index fb0849ac1..5740b1e93 100644 > --- a/lib/librte_mbuf/rte_mbuf.h > +++ b/lib/librte_mbuf/rte_mbuf.h > @@ -198,9 +198,12 @@ extern "C" { > #define PKT_RX_OUTER_L4_CKSUM_GOOD (1ULL << 22) > #define PKT_RX_OUTER_L4_CKSUM_INVALID ((1ULL << 21) | (1ULL << 22)) >=20 > -/* add new RX flags here */ > +/* add new RX flags here, don't forget to update PKT_FIRST_FREE */ >=20 > -/* add new TX flags here */ > +#define PKT_FIRST_FREE (1ULL << 23) > +#define PKT_LAST_FREE (1ULL << 39) > + > +/* add new TX flags here, don't forget to update PKT_LAST_FREE */ >=20 > /** > * Indicate that the metadata field in the mbuf is in use. > @@ -738,6 +741,7 @@ struct rte_mbuf { > */ > struct rte_mbuf_ext_shared_info *shinfo; >=20 > + uint64_t dynfield1[2]; /**< Reserved for dynamic fields. */ > } __rte_cache_aligned; >=20 > /** > @@ -1684,6 +1688,20 @@ rte_pktmbuf_attach_extbuf(struct rte_mbuf *m, void= *buf_addr, > */ > #define rte_pktmbuf_detach_extbuf(m) rte_pktmbuf_detach(m) >=20 > +/** > + * Copy dynamic fields from m_src to m_dst. > + * > + * @param m_dst > + * The destination mbuf. > + * @param m_src > + * The source mbuf. > + */ > +static inline void > +rte_mbuf_dynfield_copy(struct rte_mbuf *mdst, const struct rte_mbuf *msr= c) > +{ > + memcpy(&mdst->dynfield1, msrc->dynfield1, sizeof(mdst->dynfield1)); > +} > + > /* internal */ > static inline void > __rte_pktmbuf_copy_hdr(struct rte_mbuf *mdst, const struct rte_mbuf *msr= c) > @@ -1695,6 +1713,7 @@ __rte_pktmbuf_copy_hdr(struct rte_mbuf *mdst, const= struct rte_mbuf *msrc) > mdst->hash =3D msrc->hash; > mdst->packet_type =3D msrc->packet_type; > mdst->timestamp =3D msrc->timestamp; > + rte_mbuf_dynfield_copy(mdst, msrc); > } >=20 > /** > diff --git a/lib/librte_mbuf/rte_mbuf_dyn.c b/lib/librte_mbuf/rte_mbuf_dy= n.c > new file mode 100644 > index 000000000..9ef235483 > --- /dev/null > +++ b/lib/librte_mbuf/rte_mbuf_dyn.c > @@ -0,0 +1,548 @@ > +/* SPDX-License-Identifier: BSD-3-Clause > + * Copyright 2019 6WIND S.A. > + */ > + > +#include > +#include > +#include > + > +#include > +#include > +#include > +#include > +#include > +#include > +#include > +#include > +#include > + > +#define RTE_MBUF_DYN_MZNAME "rte_mbuf_dyn" > + > +struct mbuf_dynfield_elt { > + TAILQ_ENTRY(mbuf_dynfield_elt) next; > + struct rte_mbuf_dynfield params; > + size_t offset; > +}; > +TAILQ_HEAD(mbuf_dynfield_list, rte_tailq_entry); > + > +static struct rte_tailq_elem mbuf_dynfield_tailq =3D { > + .name =3D "RTE_MBUF_DYNFIELD", > +}; > +EAL_REGISTER_TAILQ(mbuf_dynfield_tailq); > + > +struct mbuf_dynflag_elt { > + TAILQ_ENTRY(mbuf_dynflag_elt) next; > + struct rte_mbuf_dynflag params; > + unsigned int bitnum; > +}; > +TAILQ_HEAD(mbuf_dynflag_list, rte_tailq_entry); > + > +static struct rte_tailq_elem mbuf_dynflag_tailq =3D { > + .name =3D "RTE_MBUF_DYNFLAG", > +}; > +EAL_REGISTER_TAILQ(mbuf_dynflag_tailq); > + > +struct mbuf_dyn_shm { > + /** > + * For each mbuf byte, free_space[i] !=3D 0 if space is free. > + * The value is the size of the biggest aligned element that > + * can fit in the zone. > + */ > + uint8_t free_space[sizeof(struct rte_mbuf)]; > + /** Bitfield of available flags. */ > + uint64_t free_flags; > +}; > +static struct mbuf_dyn_shm *shm; > + > +/* Set the value of free_space[] according to the size and alignment of > + * the free areas. This helps to select the best place when reserving a > + * dynamic field. Assume tailq is locked. > + */ > +static void > +process_score(void) > +{ > + size_t off, align, size, i; > + > + /* first, erase previous info */ > + for (i =3D 0; i < sizeof(struct rte_mbuf); i++) { > + if (shm->free_space[i]) > + shm->free_space[i] =3D 1; > + } > + > + for (off =3D 0; off < sizeof(struct rte_mbuf); off++) { > + /* get the size of the free zone */ > + for (size =3D 0; shm->free_space[off + size]; size++) > + ; > + if (size =3D=3D 0) > + continue; > + > + /* get the alignment of biggest object that can fit in > + * the zone at this offset. > + */ > + for (align =3D 1; > + (off % (align << 1)) =3D=3D 0 && (align << 1) <=3D size; > + align <<=3D 1) > + ; > + > + /* save it in free_space[] */ > + for (i =3D off; i < off + size; i++) > + shm->free_space[i] =3D RTE_MAX(align, shm->free_space[i]); > + } > +} > + > +/* Allocate and initialize the shared memory. Assume tailq is locked */ > +static int > +init_shared_mem(void) > +{ > + const struct rte_memzone *mz; > + uint64_t mask; > + > + if (rte_eal_process_type() =3D=3D RTE_PROC_PRIMARY) { > + mz =3D rte_memzone_reserve_aligned(RTE_MBUF_DYN_MZNAME, > + sizeof(struct mbuf_dyn_shm), > + SOCKET_ID_ANY, 0, > + RTE_CACHE_LINE_SIZE); > + } else { > + mz =3D rte_memzone_lookup(RTE_MBUF_DYN_MZNAME); > + } > + if (mz =3D=3D NULL) > + return -1; > + > + shm =3D mz->addr; > + > +#define mark_free(field) \ > + memset(&shm->free_space[offsetof(struct rte_mbuf, field)], \ > + 1, sizeof(((struct rte_mbuf *)0)->field)) Still think it would look nicer without multi-line macro defines/undef in t= he middle of the function. > + > + if (rte_eal_process_type() =3D=3D RTE_PROC_PRIMARY) { > + /* init free_space, keep it sync'd with > + * rte_mbuf_dynfield_copy(). > + */ > + memset(shm, 0, sizeof(*shm)); > + mark_free(dynfield1); > + > + /* init free_flags */ > + for (mask =3D PKT_FIRST_FREE; mask <=3D PKT_LAST_FREE; mask <<=3D 1) > + shm->free_flags |=3D mask; > + > + process_score(); > + } > +#undef mark_free > + > + return 0; > +} > + > +/* check if this offset can be used */ > +static int > +check_offset(size_t offset, size_t size, size_t align) > +{ > + size_t i; > + > + if ((offset & (align - 1)) !=3D 0) > + return -1; > + if (offset + size > sizeof(struct rte_mbuf)) > + return -1; > + > + for (i =3D 0; i < size; i++) { > + if (!shm->free_space[i + offset]) > + return -1; > + } > + > + return 0; > +} > + > +/* assume tailq is locked */ > +static struct mbuf_dynfield_elt * > +__mbuf_dynfield_lookup(const char *name) > +{ > + struct mbuf_dynfield_list *mbuf_dynfield_list; > + struct mbuf_dynfield_elt *mbuf_dynfield; > + struct rte_tailq_entry *te; > + > + mbuf_dynfield_list =3D RTE_TAILQ_CAST( > + mbuf_dynfield_tailq.head, mbuf_dynfield_list); > + > + TAILQ_FOREACH(te, mbuf_dynfield_list, next) { > + mbuf_dynfield =3D (struct mbuf_dynfield_elt *)te->data; > + if (strcmp(name, mbuf_dynfield->params.name) =3D=3D 0) > + break; > + } > + > + if (te =3D=3D NULL) { > + rte_errno =3D ENOENT; > + return NULL; > + } > + > + return mbuf_dynfield; > +} > + > +int > +rte_mbuf_dynfield_lookup(const char *name, struct rte_mbuf_dynfield *par= ams) > +{ > + struct mbuf_dynfield_elt *mbuf_dynfield; > + > + if (shm =3D=3D NULL) { > + rte_errno =3D ENOENT; > + return -1; > + } > + > + rte_mcfg_tailq_read_lock(); > + mbuf_dynfield =3D __mbuf_dynfield_lookup(name); > + rte_mcfg_tailq_read_unlock(); > + > + if (mbuf_dynfield =3D=3D NULL) { > + rte_errno =3D ENOENT; > + return -1; > + } > + > + if (params !=3D NULL) > + memcpy(params, &mbuf_dynfield->params, sizeof(*params)); > + > + return mbuf_dynfield->offset; > +} > + > +static int mbuf_dynfield_cmp(const struct rte_mbuf_dynfield *params1, > + const struct rte_mbuf_dynfield *params2) > +{ > + if (strcmp(params1->name, params2->name)) > + return -1; > + if (params1->size !=3D params2->size) > + return -1; > + if (params1->align !=3D params2->align) > + return -1; > + if (params1->flags !=3D params2->flags) > + return -1; > + return 0; > +} > + > +/* assume tailq is locked */ > +static int > +__rte_mbuf_dynfield_register_offset(const struct rte_mbuf_dynfield *para= ms, > + size_t req) > +{ > + struct mbuf_dynfield_list *mbuf_dynfield_list; > + struct mbuf_dynfield_elt *mbuf_dynfield =3D NULL; > + struct rte_tailq_entry *te =3D NULL; > + unsigned int best_zone =3D UINT_MAX; > + size_t i, offset; > + int ret; > + > + if (shm =3D=3D NULL && init_shared_mem() < 0) > + return -1; > + > + mbuf_dynfield =3D __mbuf_dynfield_lookup(params->name); > + if (mbuf_dynfield !=3D NULL) { > + if (req !=3D SIZE_MAX && req !=3D mbuf_dynfield->offset) { > + rte_errno =3D EEXIST; > + return -1; > + } > + if (mbuf_dynfield_cmp(params, &mbuf_dynfield->params) < 0) { > + rte_errno =3D EEXIST; > + return -1; > + } > + return mbuf_dynfield->offset; > + } > + > + if (rte_eal_process_type() !=3D RTE_PROC_PRIMARY) { > + rte_errno =3D EPERM; > + return -1; > + } > + > + if (req =3D=3D SIZE_MAX) { > + for (offset =3D 0; > + offset < sizeof(struct rte_mbuf); > + offset++) { > + if (check_offset(offset, params->size, > + params->align) =3D=3D 0 && > + shm->free_space[offset] < best_zone) { Probably worth to explain a bit more here about best_zone logic - trying to find offset with minimal score (minimal continuous length), etc. > + best_zone =3D shm->free_space[offset]; > + req =3D offset; > + } > + } > + if (req =3D=3D SIZE_MAX) { > + rte_errno =3D ENOENT; > + return -1; > + } > + } else { > + if (check_offset(req, params->size, params->align) < 0) { > + rte_errno =3D EBUSY; > + return -1; > + } > + } > + > + offset =3D req; > + mbuf_dynfield_list =3D RTE_TAILQ_CAST( > + mbuf_dynfield_tailq.head, mbuf_dynfield_list); > + > + te =3D rte_zmalloc("MBUF_DYNFIELD_TAILQ_ENTRY", sizeof(*te), 0); > + if (te =3D=3D NULL) > + return -1; > + > + mbuf_dynfield =3D rte_zmalloc("mbuf_dynfield", sizeof(*mbuf_dynfield), = 0); > + if (mbuf_dynfield =3D=3D NULL) { > + rte_free(te); > + return -1; > + } > + > + ret =3D strlcpy(mbuf_dynfield->params.name, params->name, > + sizeof(mbuf_dynfield->params.name)); > + if (ret < 0 || ret >=3D (int)sizeof(mbuf_dynfield->params.name)) { > + rte_errno =3D ENAMETOOLONG; > + rte_free(mbuf_dynfield); > + rte_free(te); > + return -1; > + } > + memcpy(&mbuf_dynfield->params, params, sizeof(mbuf_dynfield->params)); > + mbuf_dynfield->offset =3D offset; > + te->data =3D mbuf_dynfield; > + > + TAILQ_INSERT_TAIL(mbuf_dynfield_list, te, next); > + > + for (i =3D offset; i < offset + params->size; i++) > + shm->free_space[i] =3D 0; > + process_score(); > + > + RTE_LOG(DEBUG, MBUF, "Registered dynamic field %s (sz=3D%zu, al=3D%zu, = fl=3D0x%x) -> %zd\n", > + params->name, params->size, params->align, params->flags, > + offset); > + > + return offset; > +} > + > +int > +rte_mbuf_dynfield_register_offset(const struct rte_mbuf_dynfield *params= , > + size_t req) > +{ > + int ret; > + > + if (params->size >=3D sizeof(struct rte_mbuf)) { > + rte_errno =3D EINVAL; > + return -1; > + } > + if (!rte_is_power_of_2(params->align)) { > + rte_errno =3D EINVAL; > + return -1; > + } > + if (params->flags !=3D 0) { > + rte_errno =3D EINVAL; > + return -1; > + } > + > + rte_mcfg_tailq_write_lock(); > + ret =3D __rte_mbuf_dynfield_register_offset(params, req); > + rte_mcfg_tailq_write_unlock(); > + > + return ret; > +} > + > +int > +rte_mbuf_dynfield_register(const struct rte_mbuf_dynfield *params) > +{ > + return rte_mbuf_dynfield_register_offset(params, SIZE_MAX); > +} > + > +/* assume tailq is locked */ > +static struct mbuf_dynflag_elt * > +__mbuf_dynflag_lookup(const char *name) > +{ > + struct mbuf_dynflag_list *mbuf_dynflag_list; > + struct mbuf_dynflag_elt *mbuf_dynflag; > + struct rte_tailq_entry *te; > + > + mbuf_dynflag_list =3D RTE_TAILQ_CAST( > + mbuf_dynflag_tailq.head, mbuf_dynflag_list); > + > + TAILQ_FOREACH(te, mbuf_dynflag_list, next) { > + mbuf_dynflag =3D (struct mbuf_dynflag_elt *)te->data; > + if (strncmp(name, mbuf_dynflag->params.name, > + RTE_MBUF_DYN_NAMESIZE) =3D=3D 0) > + break; > + } > + > + if (te =3D=3D NULL) { > + rte_errno =3D ENOENT; > + return NULL; > + } > + > + return mbuf_dynflag; > +} > + > +int > +rte_mbuf_dynflag_lookup(const char *name, > + struct rte_mbuf_dynflag *params) > +{ > + struct mbuf_dynflag_elt *mbuf_dynflag; > + > + if (shm =3D=3D NULL) { > + rte_errno =3D ENOENT; > + return -1; > + } > + > + rte_mcfg_tailq_read_lock(); > + mbuf_dynflag =3D __mbuf_dynflag_lookup(name); > + rte_mcfg_tailq_read_unlock(); > + > + if (mbuf_dynflag =3D=3D NULL) { > + rte_errno =3D ENOENT; > + return -1; > + } > + > + if (params !=3D NULL) > + memcpy(params, &mbuf_dynflag->params, sizeof(*params)); > + > + return mbuf_dynflag->bitnum; > +} > + > +static int mbuf_dynflag_cmp(const struct rte_mbuf_dynflag *params1, > + const struct rte_mbuf_dynflag *params2) > +{ > + if (strcmp(params1->name, params2->name)) > + return -1; > + if (params1->flags !=3D params2->flags) > + return -1; > + return 0; > +} > + > +/* assume tailq is locked */ > +static int > +__rte_mbuf_dynflag_register_bitnum(const struct rte_mbuf_dynflag *params= , > + unsigned int req) > +{ > + struct mbuf_dynflag_list *mbuf_dynflag_list; > + struct mbuf_dynflag_elt *mbuf_dynflag =3D NULL; > + struct rte_tailq_entry *te =3D NULL; > + unsigned int bitnum; > + int ret; > + > + if (shm =3D=3D NULL && init_shared_mem() < 0) > + return -1; > + > + mbuf_dynflag =3D __mbuf_dynflag_lookup(params->name); > + if (mbuf_dynflag !=3D NULL) { > + if (req !=3D UINT_MAX && req !=3D mbuf_dynflag->bitnum) { > + rte_errno =3D EEXIST; > + return -1; > + } > + if (mbuf_dynflag_cmp(params, &mbuf_dynflag->params) < 0) { > + rte_errno =3D EEXIST; > + return -1; > + } > + return mbuf_dynflag->bitnum; > + } > + > + if (rte_eal_process_type() !=3D RTE_PROC_PRIMARY) { > + rte_errno =3D EPERM; > + return -1; > + } > + > + if (req =3D=3D UINT_MAX) { > + if (shm->free_flags =3D=3D 0) { > + rte_errno =3D ENOENT; > + return -1; > + } > + bitnum =3D rte_bsf64(shm->free_flags); > + } else { > + if ((shm->free_flags & (1ULL << req)) =3D=3D 0) { > + rte_errno =3D EBUSY; > + return -1; > + } > + bitnum =3D req; > + } > + > + mbuf_dynflag_list =3D RTE_TAILQ_CAST( > + mbuf_dynflag_tailq.head, mbuf_dynflag_list); > + > + te =3D rte_zmalloc("MBUF_DYNFLAG_TAILQ_ENTRY", sizeof(*te), 0); > + if (te =3D=3D NULL) > + return -1; > + > + mbuf_dynflag =3D rte_zmalloc("mbuf_dynflag", sizeof(*mbuf_dynflag), 0); > + if (mbuf_dynflag =3D=3D NULL) { > + rte_free(te); > + return -1; > + } > + > + ret =3D strlcpy(mbuf_dynflag->params.name, params->name, > + sizeof(mbuf_dynflag->params.name)); > + if (ret < 0 || ret >=3D (int)sizeof(mbuf_dynflag->params.name)) { > + rte_free(mbuf_dynflag); > + rte_free(te); > + rte_errno =3D ENAMETOOLONG; > + return -1; > + } > + mbuf_dynflag->bitnum =3D bitnum; > + te->data =3D mbuf_dynflag; > + > + TAILQ_INSERT_TAIL(mbuf_dynflag_list, te, next); > + > + shm->free_flags &=3D ~(1ULL << bitnum); > + > + RTE_LOG(DEBUG, MBUF, "Registered dynamic flag %s (fl=3D0x%x) -> %u\n", > + params->name, params->flags, bitnum); > + > + return bitnum; > +} > + > +int > +rte_mbuf_dynflag_register_bitnum(const struct rte_mbuf_dynflag *params, > + unsigned int req) > +{ > + int ret; > + > + if (req !=3D UINT_MAX && req >=3D 64) { Might be better to replace 64 with something like sizeof(mbuf->ol_flags) * = CHAR_BIT or so. Apart from that: Acked-by: Konstantin Ananyev > + rte_errno =3D EINVAL; > + return -1; > + } > + > + rte_mcfg_tailq_write_lock(); > + ret =3D __rte_mbuf_dynflag_register_bitnum(params, req); > + rte_mcfg_tailq_write_unlock(); > + > + return ret; > +} > +