freebsd-skq/sys/kern/kern_vimage.c

761 lines
19 KiB
C
Raw Normal View History

Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
/*-
* Copyright (c) 2004-2009 University of Zagreb
* Copyright (c) 2006-2009 FreeBSD Foundation
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
*
* This software was developed by the University of Zagreb and the
* FreeBSD Foundation under sponsorship by the Stichting NLnet and the
* FreeBSD Foundation.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
#include "opt_ddb.h"
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
#include <sys/param.h>
#include <sys/jail.h>
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
#include <sys/kernel.h>
#include <sys/linker.h>
#include <sys/lock.h>
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
#include <sys/malloc.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/sx.h>
#include <sys/priv.h>
#include <sys/refcount.h>
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
#include <sys/vimage.h>
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
#ifdef DDB
#include <ddb/ddb.h>
#endif
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
#include <net/if.h>
#include <net/route.h>
#include <net/vnet.h>
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
MALLOC_DEFINE(M_VIMAGE, "vimage", "vimage resource container");
Permit buiding kernels with options VIMAGE, restricted to only a single active network stack instance. Turning on options VIMAGE at compile time yields the following changes relative to default kernel build: 1) V_ accessor macros for virtualized variables resolve to structure fields via base pointers, instead of being resolved as fields in global structs or plain global variables. As an example, V_ifnet becomes: options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet default build: vnet_net_0._ifnet options VIMAGE_GLOBALS: ifnet 2) INIT_VNET_* macros will declare and set up base pointers to be used by V_ accessor macros, instead of resolving to whitespace: INIT_VNET_NET(ifp->if_vnet); becomes struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET]; 3) Memory for vnet modules registered via vnet_mod_register() is now allocated at run time in sys/kern/kern_vimage.c, instead of per vnet module structs being declared as globals. If required, vnet modules can now request the framework to provide them with allocated bzeroed memory by filling in the vmi_size field in their vmi_modinfo structures. 4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are extended to hold a pointer to the parent vnet. options VIMAGE builds will fill in those fields as required. 5) curvnet is introduced as a new global variable in options VIMAGE builds, always pointing to the default and only struct vnet. 6) struct sysctl_oid has been extended with additional two fields to store major and minor virtualization module identifiers, oid_v_subs and oid_v_mod. SYSCTL_V_* family of macros will fill in those fields accordingly, and store the offset in the appropriate vnet container struct in oid_arg1. In sysctl handlers dealing with virtualized sysctls, the SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target variable and make it available in arg1 variable for further processing. Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have been deleted. Reviewed by: bz, rwatson Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
MALLOC_DEFINE(M_VNET, "vnet", "network stack control block");
MALLOC_DEFINE(M_VPROCG, "vprocg", "process group control block");
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
static TAILQ_HEAD(vnet_modlink_head, vnet_modlink) vnet_modlink_head;
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
static TAILQ_HEAD(vnet_modpending_head, vnet_modlink) vnet_modpending_head;
static void vnet_mod_complete_registration(struct vnet_modlink *);
static int vnet_mod_constructor(struct vnet_modlink *);
static int vnet_mod_destructor(struct vnet_modlink *);
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
static struct vimage *vi_alloc(struct vimage *, char *);
static int vi_destroy(struct vimage *);
static struct vimage *vimage_get_next(struct vimage *, struct vimage *, int);
static void vimage_relative_name(struct vimage *, struct vimage *,
char *, int);
#define VNET_LIST_WLOCK() \
mtx_lock(&vnet_list_refc_mtx); \
while (vnet_list_refc != 0) \
cv_wait(&vnet_list_condvar, &vnet_list_refc_mtx);
#define VNET_LIST_WUNLOCK() \
mtx_unlock(&vnet_list_refc_mtx);
struct vimage_list_head vimage_head;
Permit buiding kernels with options VIMAGE, restricted to only a single active network stack instance. Turning on options VIMAGE at compile time yields the following changes relative to default kernel build: 1) V_ accessor macros for virtualized variables resolve to structure fields via base pointers, instead of being resolved as fields in global structs or plain global variables. As an example, V_ifnet becomes: options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet default build: vnet_net_0._ifnet options VIMAGE_GLOBALS: ifnet 2) INIT_VNET_* macros will declare and set up base pointers to be used by V_ accessor macros, instead of resolving to whitespace: INIT_VNET_NET(ifp->if_vnet); becomes struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET]; 3) Memory for vnet modules registered via vnet_mod_register() is now allocated at run time in sys/kern/kern_vimage.c, instead of per vnet module structs being declared as globals. If required, vnet modules can now request the framework to provide them with allocated bzeroed memory by filling in the vmi_size field in their vmi_modinfo structures. 4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are extended to hold a pointer to the parent vnet. options VIMAGE builds will fill in those fields as required. 5) curvnet is introduced as a new global variable in options VIMAGE builds, always pointing to the default and only struct vnet. 6) struct sysctl_oid has been extended with additional two fields to store major and minor virtualization module identifiers, oid_v_subs and oid_v_mod. SYSCTL_V_* family of macros will fill in those fields accordingly, and store the offset in the appropriate vnet container struct in oid_arg1. In sysctl handlers dealing with virtualized sysctls, the SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target variable and make it available in arg1 variable for further processing. Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have been deleted. Reviewed by: bz, rwatson Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
struct vnet_list_head vnet_head;
struct vprocg_list_head vprocg_head;
struct vprocg vprocg_0;
Permit buiding kernels with options VIMAGE, restricted to only a single active network stack instance. Turning on options VIMAGE at compile time yields the following changes relative to default kernel build: 1) V_ accessor macros for virtualized variables resolve to structure fields via base pointers, instead of being resolved as fields in global structs or plain global variables. As an example, V_ifnet becomes: options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet default build: vnet_net_0._ifnet options VIMAGE_GLOBALS: ifnet 2) INIT_VNET_* macros will declare and set up base pointers to be used by V_ accessor macros, instead of resolving to whitespace: INIT_VNET_NET(ifp->if_vnet); becomes struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET]; 3) Memory for vnet modules registered via vnet_mod_register() is now allocated at run time in sys/kern/kern_vimage.c, instead of per vnet module structs being declared as globals. If required, vnet modules can now request the framework to provide them with allocated bzeroed memory by filling in the vmi_size field in their vmi_modinfo structures. 4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are extended to hold a pointer to the parent vnet. options VIMAGE builds will fill in those fields as required. 5) curvnet is introduced as a new global variable in options VIMAGE builds, always pointing to the default and only struct vnet. 6) struct sysctl_oid has been extended with additional two fields to store major and minor virtualization module identifiers, oid_v_subs and oid_v_mod. SYSCTL_V_* family of macros will fill in those fields accordingly, and store the offset in the appropriate vnet container struct in oid_arg1. In sysctl handlers dealing with virtualized sysctls, the SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target variable and make it available in arg1 variable for further processing. Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have been deleted. Reviewed by: bz, rwatson Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
struct cv vnet_list_condvar;
struct mtx vnet_list_refc_mtx;
int vnet_list_refc = 0;
static u_int last_vi_id = 0;
static u_int last_vprocg_id = 0;
struct vnet *vnet0;
/*
* Move an ifnet to or from another vnet, specified by the jail id. If a
* vi_req is passed in, it is used to find the interface and a vimage
* containing the vnet (a vimage name of ".." stands for the parent vnet).
*/
int
vi_if_move(struct thread *td, struct ifnet *ifp, char *ifname, int jid,
struct vi_req *vi_req)
{
struct ifnet *t_ifp;
struct prison *pr;
struct vimage *new_vip, *my_vip;
struct vnet *new_vnet;
int error;
if (vi_req != NULL) {
/* SIOCSIFVIMAGE */
pr = NULL;
/* Check for API / ABI version mismatch. */
if (vi_req->vi_api_cookie != VI_API_COOKIE)
return (EDOOFUS);
/* Find the target vnet. */
my_vip = TD_TO_VIMAGE(td);
if (strcmp(vi_req->vi_name, "..") == 0) {
if (IS_DEFAULT_VIMAGE(my_vip))
return (ENXIO);
new_vnet = my_vip->vi_parent->v_net;
} else {
new_vip = vimage_by_name(my_vip, vi_req->vi_name);
if (new_vip == NULL)
return (ENXIO);
new_vnet = new_vip->v_net;
}
/* Try to find the target ifnet by name. */
ifname = vi_req->vi_if_xname;
ifp = ifunit(ifname);
if (ifp == NULL)
return (ENXIO);
} else {
sx_slock(&allprison_lock);
pr = prison_find_child(td->td_ucred->cr_prison, jid);
sx_sunlock(&allprison_lock);
if (pr == NULL)
return (ENXIO);
prison_hold_locked(pr);
mtx_unlock(&pr->pr_mtx);
if (ifp != NULL) {
/* SIOCSIFVNET */
new_vnet = pr->pr_vnet;
} else {
/* SIOCSIFRVNET */
new_vnet = TD_TO_VNET(td);
CURVNET_SET(pr->pr_vnet);
ifp = ifunit(ifname);
CURVNET_RESTORE();
if (ifp == NULL) {
prison_free(pr);
return (ENXIO);
}
}
}
error = 0;
if (new_vnet != ifp->if_vnet) {
/*
* Check for naming clashes in target vnet. Not locked so races
* are possible.
*/
CURVNET_SET_QUIET(new_vnet);
t_ifp = ifunit(ifname);
CURVNET_RESTORE();
if (t_ifp != NULL)
error = EEXIST;
else {
/* Detach from curvnet and attach to new_vnet. */
if_vmove(ifp, new_vnet);
/* Report the new if_xname back to the userland */
sprintf(ifname, "%s", ifp->if_xname);
}
}
if (pr != NULL)
prison_free(pr);
return (error);
}
/*
* Interim userspace interface - will be replaced by jail soon.
*/
int
vi_td_ioctl(u_long cmd, struct vi_req *vi_req, struct thread *td)
{
int error = 0;
struct vimage *vip = TD_TO_VIMAGE(td);
struct vimage *vip_r = NULL;
/* Check for API / ABI version mismatch. */
if (vi_req->vi_api_cookie != VI_API_COOKIE)
return (EDOOFUS);
error = priv_check(td, PRIV_REBOOT); /* XXX temp. priv abuse */
if (error)
return (error);
vip_r = vimage_by_name(vip, vi_req->vi_name);
if (vip_r == NULL && !(vi_req->vi_req_action & VI_CREATE))
return (ESRCH);
if (vip_r != NULL && vi_req->vi_req_action & VI_CREATE)
return (EADDRINUSE);
if (vi_req->vi_req_action == VI_GETNEXT) {
vip_r = vimage_get_next(vip, vip_r, 0);
if (vip_r == NULL)
return (ESRCH);
}
if (vi_req->vi_req_action == VI_GETNEXT_RECURSE) {
vip_r = vimage_get_next(vip, vip_r, 1);
if (vip_r == NULL)
return (ESRCH);
}
if (vip_r && !vi_child_of(vip, vip_r) && /* XXX delete the rest? */
vi_req->vi_req_action != VI_GET &&
vi_req->vi_req_action != VI_GETNEXT)
return (EPERM);
switch (cmd) {
case SIOCGPVIMAGE:
vimage_relative_name(vip, vip_r, vi_req->vi_name,
sizeof (vi_req->vi_name));
vi_req->vi_proc_count = vip_r->v_procg->nprocs;
vi_req->vi_if_count = vip_r->v_net->ifcnt;
vi_req->vi_sock_count = vip_r->v_net->sockcnt;
break;
case SIOCSPVIMAGE:
if (vi_req->vi_req_action == VI_DESTROY) {
error = vi_destroy(vip_r);
break;
}
if (vi_req->vi_req_action == VI_SWITCHTO) {
struct proc *p = td->td_proc;
struct ucred *oldcred, *newcred;
/*
* XXX priv_check()?
* XXX allow only a single td per proc here?
*/
newcred = crget();
PROC_LOCK(p);
oldcred = p->p_ucred;
setsugid(p);
crcopy(newcred, oldcred);
refcount_release(&newcred->cr_vimage->vi_ucredrefc);
newcred->cr_vimage = vip_r;
refcount_acquire(&newcred->cr_vimage->vi_ucredrefc);
p->p_ucred = newcred;
PROC_UNLOCK(p);
sx_xlock(&allproc_lock);
oldcred->cr_vimage->v_procg->nprocs--;
refcount_release(&oldcred->cr_vimage->vi_ucredrefc);
P_TO_VPROCG(p)->nprocs++;
sx_xunlock(&allproc_lock);
crfree(oldcred);
break;
}
if (vi_req->vi_req_action & VI_CREATE) {
char *dotpos;
dotpos = strrchr(vi_req->vi_name, '.');
if (dotpos != NULL) {
*dotpos = 0;
vip = vimage_by_name(vip, vi_req->vi_name);
if (vip == NULL)
return (ESRCH);
dotpos++;
vip_r = vi_alloc(vip, dotpos);
} else
vip_r = vi_alloc(vip, vi_req->vi_name);
if (vip_r == NULL)
return (ENOMEM);
}
}
return (error);
}
int
vi_child_of(struct vimage *parent, struct vimage *child)
{
if (child == parent)
return (0);
for (; child; child = child->vi_parent)
if (child == parent)
return (1);
return (0);
}
struct vimage *
vimage_by_name(struct vimage *top, char *name)
{
struct vimage *vip;
char *next_name;
int namelen;
next_name = strchr(name, '.');
if (next_name != NULL) {
namelen = next_name - name;
next_name++;
if (namelen == 0) {
if (strlen(next_name) == 0)
return (top); /* '.' == this vimage */
else
return (NULL);
}
} else
namelen = strlen(name);
if (namelen == 0)
return (NULL);
LIST_FOREACH(vip, &top->vi_child_head, vi_sibling) {
if (strlen(vip->vi_name) == namelen &&
strncmp(name, vip->vi_name, namelen) == 0) {
if (next_name != NULL)
return (vimage_by_name(vip, next_name));
else
return (vip);
}
}
return (NULL);
}
static void
vimage_relative_name(struct vimage *top, struct vimage *where,
char *buffer, int bufflen)
{
int used = 1;
if (where == top) {
sprintf(buffer, ".");
return;
} else
*buffer = 0;
do {
int namelen = strlen(where->vi_name);
if (namelen + used + 1 >= bufflen)
panic("buffer overflow");
if (used > 1) {
bcopy(buffer, &buffer[namelen + 1], used);
buffer[namelen] = '.';
used++;
} else
bcopy(buffer, &buffer[namelen], used);
bcopy(where->vi_name, buffer, namelen);
used += namelen;
where = where->vi_parent;
} while (where != top);
}
static struct vimage *
vimage_get_next(struct vimage *top, struct vimage *where, int recurse)
{
struct vimage *next;
if (recurse) {
/* Try to go deeper in the hierarchy */
next = LIST_FIRST(&where->vi_child_head);
if (next != NULL)
return (next);
}
do {
/* Try to find next sibling */
next = LIST_NEXT(where, vi_sibling);
if (!recurse || next != NULL)
return (next);
/* Nothing left on this level, go one level up */
where = where->vi_parent;
} while (where != top->vi_parent);
/* Nothing left to be visited, we are done */
return (NULL);
}
/*
* Kernel interfaces and handlers.
*/
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
void
vnet_mod_register(const struct vnet_modinfo *vmi)
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
{
vnet_mod_register_multi(vmi, NULL, NULL);
}
void
vnet_mod_register_multi(const struct vnet_modinfo *vmi, void *iarg,
char *iname)
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
{
struct vnet_modlink *vml, *vml_iter;
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
/* Do not register the same {module, iarg} pair more than once. */
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
TAILQ_FOREACH(vml_iter, &vnet_modlink_head, vml_mod_le)
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
if (vml_iter->vml_modinfo == vmi && vml_iter->vml_iarg == iarg)
break;
if (vml_iter != NULL)
panic("registering an already registered vnet module: %s",
vml_iter->vml_modinfo->vmi_name);
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
vml = malloc(sizeof(struct vnet_modlink), M_VIMAGE, M_NOWAIT);
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
/*
* XXX we support only statically assigned module IDs at the time.
* In principle modules should be able to get a dynamically
* assigned ID at registration time.
*
* If a module is registered in multiple instances, then each
* instance must have both iarg and iname set.
*/
if (vmi->vmi_id >= VNET_MOD_MAX)
panic("invalid vnet module ID: %d", vmi->vmi_id);
if (vmi->vmi_name == NULL)
panic("vnet module with no name: %d", vmi->vmi_id);
if ((iarg == NULL) ^ (iname == NULL))
panic("invalid vnet module instance: %s", vmi->vmi_name);
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
vml->vml_modinfo = vmi;
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
vml->vml_iarg = iarg;
vml->vml_iname = iname;
/* Check whether the module we depend on is already registered. */
if (vmi->vmi_dependson != vmi->vmi_id) {
TAILQ_FOREACH(vml_iter, &vnet_modlink_head, vml_mod_le)
if (vml_iter->vml_modinfo->vmi_id ==
vmi->vmi_dependson)
break; /* Depencency found, we are done. */
if (vml_iter == NULL) {
#ifdef DEBUG_ORDERING
printf("dependency %d missing for vnet mod %s,"
"postponing registration\n",
vmi->vmi_dependson, vmi->vmi_name);
#endif /* DEBUG_ORDERING */
TAILQ_INSERT_TAIL(&vnet_modpending_head, vml,
vml_mod_le);
return;
}
}
vnet_mod_complete_registration(vml);
}
void
vnet_mod_complete_registration(struct vnet_modlink *vml)
{
VNET_ITERATOR_DECL(vnet_iter);
struct vnet_modlink *vml_iter;
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
TAILQ_INSERT_TAIL(&vnet_modlink_head, vml, vml_mod_le);
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
VNET_FOREACH(vnet_iter) {
CURVNET_SET_QUIET(vnet_iter);
vnet_mod_constructor(vml);
CURVNET_RESTORE();
}
/* Check for pending modules depending on us. */
do {
TAILQ_FOREACH(vml_iter, &vnet_modpending_head, vml_mod_le)
if (vml_iter->vml_modinfo->vmi_dependson ==
vml->vml_modinfo->vmi_id)
break;
if (vml_iter != NULL) {
#ifdef DEBUG_ORDERING
printf("vnet mod %s now registering,"
"dependency %d loaded\n",
vml_iter->vml_modinfo->vmi_name,
vml->vml_modinfo->vmi_id);
#endif /* DEBUG_ORDERING */
TAILQ_REMOVE(&vnet_modpending_head, vml_iter,
vml_mod_le);
vnet_mod_complete_registration(vml_iter);
}
} while (vml_iter != NULL);
}
void
vnet_mod_deregister(const struct vnet_modinfo *vmi)
{
vnet_mod_deregister_multi(vmi, NULL, NULL);
}
void
vnet_mod_deregister_multi(const struct vnet_modinfo *vmi, void *iarg,
char *iname)
{
VNET_ITERATOR_DECL(vnet_iter);
struct vnet_modlink *vml;
TAILQ_FOREACH(vml, &vnet_modlink_head, vml_mod_le)
if (vml->vml_modinfo == vmi && vml->vml_iarg == iarg)
break;
if (vml == NULL)
panic("cannot deregister unregistered vnet module %s",
vmi->vmi_name);
VNET_FOREACH(vnet_iter) {
CURVNET_SET_QUIET(vnet_iter);
vnet_mod_destructor(vml);
CURVNET_RESTORE();
}
TAILQ_REMOVE(&vnet_modlink_head, vml, vml_mod_le);
free(vml, M_VIMAGE);
}
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
static int
vnet_mod_constructor(struct vnet_modlink *vml)
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
{
const struct vnet_modinfo *vmi = vml->vml_modinfo;
#ifdef DEBUG_ORDERING
printf("instantiating vnet_%s", vmi->vmi_name);
if (vml->vml_iarg)
printf("/%s", vml->vml_iname);
printf(": ");
if (vmi->vmi_iattach != NULL)
printf("iattach()");
printf("\n");
#endif
if (vmi->vmi_iattach != NULL)
vmi->vmi_iattach(vml->vml_iarg);
return (0);
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
}
static int
vnet_mod_destructor(struct vnet_modlink *vml)
{
const struct vnet_modinfo *vmi = vml->vml_modinfo;
#ifdef DEBUG_ORDERING
printf("destroying vnet_%s", vmi->vmi_name);
if (vml->vml_iarg)
printf("/%s", vml->vml_iname);
printf(": ");
if (vmi->vmi_idetach != NULL)
printf("idetach(); ");
printf("\n");
#endif
if (vmi->vmi_idetach)
vmi->vmi_idetach(vml->vml_iarg);
return (0);
}
struct vnet *
vnet_alloc(void)
{
struct vnet *vnet;
struct vnet_modlink *vml;
vnet = malloc(sizeof(struct vnet), M_VNET, M_WAITOK | M_ZERO);
vnet->vnet_magic_n = VNET_MAGIC_N;
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator (DPCPU), as suggested by Peter Wemm, and implement a new per-virtual network stack memory allocator. Modify vnet to use the allocator instead of monolithic global container structures (vinet, ...). This change solves many binary compatibility problems associated with VIMAGE, and restores ELF symbols for virtualized global variables. Each virtualized global variable exists as a "reference copy", and also once per virtual network stack. Virtualized global variables are tagged at compile-time, placing the in a special linker set, which is loaded into a contiguous region of kernel memory. Virtualized global variables in the base kernel are linked as normal, but those in modules are copied and relocated to a reserved portion of the kernel's vnet region with the help of a the kernel linker. Virtualized global variables exist in per-vnet memory set up when the network stack instance is created, and are initialized statically from the reference copy. Run-time access occurs via an accessor macro, which converts from the current vnet and requested symbol to a per-vnet address. When "options VIMAGE" is not compiled into the kernel, normal global ELF symbols will be used instead and indirection is avoided. This change restores static initialization for network stack global variables, restores support for non-global symbols and types, eliminates the need for many subsystem constructors, eliminates large per-subsystem structures that caused many binary compatibility issues both for monitoring applications (netstat) and kernel modules, removes the per-function INIT_VNET_*() macros throughout the stack, eliminates the need for vnet_symmap ksym(2) munging, and eliminates duplicate definitions of virtualized globals under VIMAGE_GLOBALS. Bump __FreeBSD_version and update UPDATING. Portions submitted by: bz Reviewed by: bz, zec Discussed with: gnn, jamie, jeff, jhb, julian, sam Suggested by: peter Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
vnet_data_init(vnet);
/* Initialize / attach vnet module instances. */
CURVNET_SET_QUIET(vnet);
TAILQ_FOREACH(vml, &vnet_modlink_head, vml_mod_le)
vnet_mod_constructor(vml);
CURVNET_RESTORE();
VNET_LIST_WLOCK();
LIST_INSERT_HEAD(&vnet_head, vnet, vnet_le);
VNET_LIST_WUNLOCK();
return (vnet);
}
void
vnet_destroy(struct vnet *vnet)
{
struct ifnet *ifp, *nifp;
struct vnet_modlink *vml;
KASSERT(vnet->sockcnt == 0, ("%s: vnet still has sockets", __func__));
VNET_LIST_WLOCK();
LIST_REMOVE(vnet, vnet_le);
VNET_LIST_WUNLOCK();
CURVNET_SET_QUIET(vnet);
/* Return all inherited interfaces to their parent vnets. */
TAILQ_FOREACH_SAFE(ifp, &V_ifnet, if_link, nifp) {
if (ifp->if_home_vnet != ifp->if_vnet)
if_vmove(ifp, ifp->if_home_vnet);
}
/* Detach / free per-module state instances. */
TAILQ_FOREACH_REVERSE(vml, &vnet_modlink_head,
vnet_modlink_head, vml_mod_le)
vnet_mod_destructor(vml);
CURVNET_RESTORE();
/* Hopefully, we are OK to free the vnet container itself. */
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator (DPCPU), as suggested by Peter Wemm, and implement a new per-virtual network stack memory allocator. Modify vnet to use the allocator instead of monolithic global container structures (vinet, ...). This change solves many binary compatibility problems associated with VIMAGE, and restores ELF symbols for virtualized global variables. Each virtualized global variable exists as a "reference copy", and also once per virtual network stack. Virtualized global variables are tagged at compile-time, placing the in a special linker set, which is loaded into a contiguous region of kernel memory. Virtualized global variables in the base kernel are linked as normal, but those in modules are copied and relocated to a reserved portion of the kernel's vnet region with the help of a the kernel linker. Virtualized global variables exist in per-vnet memory set up when the network stack instance is created, and are initialized statically from the reference copy. Run-time access occurs via an accessor macro, which converts from the current vnet and requested symbol to a per-vnet address. When "options VIMAGE" is not compiled into the kernel, normal global ELF symbols will be used instead and indirection is avoided. This change restores static initialization for network stack global variables, restores support for non-global symbols and types, eliminates the need for many subsystem constructors, eliminates large per-subsystem structures that caused many binary compatibility issues both for monitoring applications (netstat) and kernel modules, removes the per-function INIT_VNET_*() macros throughout the stack, eliminates the need for vnet_symmap ksym(2) munging, and eliminates duplicate definitions of virtualized globals under VIMAGE_GLOBALS. Bump __FreeBSD_version and update UPDATING. Portions submitted by: bz Reviewed by: bz, zec Discussed with: gnn, jamie, jeff, jhb, julian, sam Suggested by: peter Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
vnet_data_destroy(vnet);
vnet->vnet_magic_n = 0xdeadbeef;
free(vnet, M_VNET);
}
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator (DPCPU), as suggested by Peter Wemm, and implement a new per-virtual network stack memory allocator. Modify vnet to use the allocator instead of monolithic global container structures (vinet, ...). This change solves many binary compatibility problems associated with VIMAGE, and restores ELF symbols for virtualized global variables. Each virtualized global variable exists as a "reference copy", and also once per virtual network stack. Virtualized global variables are tagged at compile-time, placing the in a special linker set, which is loaded into a contiguous region of kernel memory. Virtualized global variables in the base kernel are linked as normal, but those in modules are copied and relocated to a reserved portion of the kernel's vnet region with the help of a the kernel linker. Virtualized global variables exist in per-vnet memory set up when the network stack instance is created, and are initialized statically from the reference copy. Run-time access occurs via an accessor macro, which converts from the current vnet and requested symbol to a per-vnet address. When "options VIMAGE" is not compiled into the kernel, normal global ELF symbols will be used instead and indirection is avoided. This change restores static initialization for network stack global variables, restores support for non-global symbols and types, eliminates the need for many subsystem constructors, eliminates large per-subsystem structures that caused many binary compatibility issues both for monitoring applications (netstat) and kernel modules, removes the per-function INIT_VNET_*() macros throughout the stack, eliminates the need for vnet_symmap ksym(2) munging, and eliminates duplicate definitions of virtualized globals under VIMAGE_GLOBALS. Bump __FreeBSD_version and update UPDATING. Portions submitted by: bz Reviewed by: bz, zec Discussed with: gnn, jamie, jeff, jhb, julian, sam Suggested by: peter Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
void
vnet_foreach(void (*vnet_foreach_fn)(struct vnet *, void *), void *arg)
{
struct vnet *vnet;
VNET_LIST_RLOCK();
LIST_FOREACH(vnet, &vnet_head, vnet_le)
vnet_foreach_fn(vnet, arg);
VNET_LIST_RUNLOCK();
}
static struct vimage *
vi_alloc(struct vimage *parent, char *name)
{
struct vimage *vip;
struct vprocg *vprocg;
Permit buiding kernels with options VIMAGE, restricted to only a single active network stack instance. Turning on options VIMAGE at compile time yields the following changes relative to default kernel build: 1) V_ accessor macros for virtualized variables resolve to structure fields via base pointers, instead of being resolved as fields in global structs or plain global variables. As an example, V_ifnet becomes: options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet default build: vnet_net_0._ifnet options VIMAGE_GLOBALS: ifnet 2) INIT_VNET_* macros will declare and set up base pointers to be used by V_ accessor macros, instead of resolving to whitespace: INIT_VNET_NET(ifp->if_vnet); becomes struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET]; 3) Memory for vnet modules registered via vnet_mod_register() is now allocated at run time in sys/kern/kern_vimage.c, instead of per vnet module structs being declared as globals. If required, vnet modules can now request the framework to provide them with allocated bzeroed memory by filling in the vmi_size field in their vmi_modinfo structures. 4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are extended to hold a pointer to the parent vnet. options VIMAGE builds will fill in those fields as required. 5) curvnet is introduced as a new global variable in options VIMAGE builds, always pointing to the default and only struct vnet. 6) struct sysctl_oid has been extended with additional two fields to store major and minor virtualization module identifiers, oid_v_subs and oid_v_mod. SYSCTL_V_* family of macros will fill in those fields accordingly, and store the offset in the appropriate vnet container struct in oid_arg1. In sysctl handlers dealing with virtualized sysctls, the SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target variable and make it available in arg1 variable for further processing. Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have been deleted. Reviewed by: bz, rwatson Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
vip = malloc(sizeof(struct vimage), M_VIMAGE, M_NOWAIT | M_ZERO);
if (vip == NULL)
panic("vi_alloc: malloc failed for vimage \"%s\"\n", name);
vip->vi_id = last_vi_id++;
LIST_INIT(&vip->vi_child_head);
sprintf(vip->vi_name, "%s", name);
vip->vi_parent = parent;
/* XXX locking */
if (parent != NULL)
LIST_INSERT_HEAD(&parent->vi_child_head, vip, vi_sibling);
else if (!LIST_EMPTY(&vimage_head))
panic("there can be only one default vimage!");
LIST_INSERT_HEAD(&vimage_head, vip, vi_le);
vip->v_net = vnet_alloc();
vprocg = malloc(sizeof(struct vprocg), M_VPROCG, M_NOWAIT | M_ZERO);
if (vprocg == NULL)
panic("vi_alloc: malloc failed for vprocg \"%s\"\n", name);
vip->v_procg = vprocg;
vprocg->vprocg_id = last_vprocg_id++;
/* XXX locking */
LIST_INSERT_HEAD(&vprocg_head, vprocg, vprocg_le);
return (vip);
}
/*
* Destroy a vnet - unlink all linked lists, hashtables etc., free all
* the memory, stop all the timers...
*/
static int
vi_destroy(struct vimage *vip)
{
struct vprocg *vprocg = vip->v_procg;
/* XXX Beware of races -> more locking to be done... */
if (!LIST_EMPTY(&vip->vi_child_head))
return (EBUSY);
if (vprocg->nprocs != 0)
return (EBUSY);
#ifdef INVARIANTS
if (vip->vi_ucredrefc != 0)
printf("vi_destroy: %s ucredrefc %d\n",
vip->vi_name, vip->vi_ucredrefc);
#endif
/* Point with no return - cleanup MUST succeed! */
vnet_destroy(vip->v_net);
LIST_REMOVE(vip, vi_le);
LIST_REMOVE(vip, vi_sibling);
LIST_REMOVE(vprocg, vprocg_le);
free(vprocg, M_VPROCG);
free(vip, M_VIMAGE);
return (0);
}
Permit buiding kernels with options VIMAGE, restricted to only a single active network stack instance. Turning on options VIMAGE at compile time yields the following changes relative to default kernel build: 1) V_ accessor macros for virtualized variables resolve to structure fields via base pointers, instead of being resolved as fields in global structs or plain global variables. As an example, V_ifnet becomes: options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet default build: vnet_net_0._ifnet options VIMAGE_GLOBALS: ifnet 2) INIT_VNET_* macros will declare and set up base pointers to be used by V_ accessor macros, instead of resolving to whitespace: INIT_VNET_NET(ifp->if_vnet); becomes struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET]; 3) Memory for vnet modules registered via vnet_mod_register() is now allocated at run time in sys/kern/kern_vimage.c, instead of per vnet module structs being declared as globals. If required, vnet modules can now request the framework to provide them with allocated bzeroed memory by filling in the vmi_size field in their vmi_modinfo structures. 4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are extended to hold a pointer to the parent vnet. options VIMAGE builds will fill in those fields as required. 5) curvnet is introduced as a new global variable in options VIMAGE builds, always pointing to the default and only struct vnet. 6) struct sysctl_oid has been extended with additional two fields to store major and minor virtualization module identifiers, oid_v_subs and oid_v_mod. SYSCTL_V_* family of macros will fill in those fields accordingly, and store the offset in the appropriate vnet container struct in oid_arg1. In sysctl handlers dealing with virtualized sysctls, the SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target variable and make it available in arg1 variable for further processing. Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have been deleted. Reviewed by: bz, rwatson Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
static void
vi_init(void *unused)
{
TAILQ_INIT(&vnet_modlink_head);
TAILQ_INIT(&vnet_modpending_head);
LIST_INIT(&vimage_head);
LIST_INIT(&vprocg_head);
LIST_INIT(&vnet_head);
mtx_init(&vnet_list_refc_mtx, "vnet_list_refc_mtx", NULL, MTX_DEF);
cv_init(&vnet_list_condvar, "vnet_list_condvar");
/* Default image has no parent and no name. */
vi_alloc(NULL, "");
/*
* We MUST clear curvnet in vi_init_done() before going SMP,
* otherwise CURVNET_SET() macros would scream about unnecessary
* curvnet recursions.
*/
curvnet = prison0.pr_vnet = vnet0 = LIST_FIRST(&vnet_head);
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
}
static void
vi_init_done(void *unused)
{
struct vnet_modlink *vml_iter;
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
curvnet = NULL;
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
if (TAILQ_EMPTY(&vnet_modpending_head))
return;
printf("vnet modules with unresolved dependencies:\n");
TAILQ_FOREACH(vml_iter, &vnet_modpending_head, vml_mod_le)
printf(" %d:%s depending on %d\n",
vml_iter->vml_modinfo->vmi_id,
vml_iter->vml_modinfo->vmi_name,
vml_iter->vml_modinfo->vmi_dependson);
panic("going nowhere without my vnet modules!");
Conditionally compile out V_ globals while instantiating the appropriate container structures, depending on VIMAGE_GLOBALS compile time option. Make VIMAGE_GLOBALS a new compile-time option, which by default will not be defined, resulting in instatiations of global variables selected for V_irtualization (enclosed in #ifdef VIMAGE_GLOBALS blocks) to be effectively compiled out. Instantiate new global container structures to hold V_irtualized variables: vnet_net_0, vnet_inet_0, vnet_inet6_0, vnet_ipsec_0, vnet_netgraph_0, and vnet_gif_0. Update the VSYM() macro so that depending on VIMAGE_GLOBALS the V_ macros resolve either to the original globals, or to fields inside container structures, i.e. effectively #ifdef VIMAGE_GLOBALS #define V_rt_tables rt_tables #else #define V_rt_tables vnet_net_0._rt_tables #endif Update SYSCTL_V_*() macros to operate either on globals or on fields inside container structs. Extend the internal kldsym() lookups with the ability to resolve selected fields inside the virtualization container structs. This applies only to the fields which are explicitly registered for kldsym() visibility via VNET_MOD_DECLARE() and vnet_mod_register(), currently this is done only in sys/net/if.c. Fix a few broken instances of MODULE_GLOBAL() macro use in SCTP code, and modify the MODULE_GLOBAL() macro to resolve to V_ macros, which in turn result in proper code being generated depending on VIMAGE_GLOBALS. De-virtualize local static variables in sys/contrib/pf/net/pf_subr.c which were prematurely V_irtualized by automated V_ prepending scripts during earlier merging steps. PF virtualization will be done separately, most probably after next PF import. Convert a few variable initializations at instantiation to initialization in init functions, most notably in ipfw. Also convert TUNABLE_INT() initializers for V_ variables to TUNABLE_FETCH_INT() in initializer functions. Discussed at: devsummit Strassburg Reviewed by: bz, julian Approved by: julian (mentor) Obtained from: //depot/projects/vimage-commit2/... X-MFC after: never Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-12-10 23:12:39 +00:00
}
SYSINIT(vimage, SI_SUB_VIMAGE, SI_ORDER_FIRST, vi_init, NULL);
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
SYSINIT(vimage_done, SI_SUB_VIMAGE_DONE, SI_ORDER_FIRST, vi_init_done, NULL);
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
#ifdef DDB
DB_SHOW_COMMAND(vnets, db_show_vnets)
{
VNET_ITERATOR_DECL(vnet_iter);
#if SIZE_MAX == UINT32_MAX /* 32-bit arch */
db_printf(" vnet ifs socks");
#else /* 64-bit arch, most probaly... */
db_printf(" vnet ifs socks");
#endif
VNET_FOREACH(vnet_iter) {
db_printf("%p %3d %5d",
vnet_iter, vnet_iter->ifcnt, vnet_iter->sockcnt);
Change the curvnet variable from a global const struct vnet *, previously always pointing to the default vnet context, to a dynamically changing thread-local one. The currvnet context should be set on entry to networking code via CURVNET_SET() macros, and reverted to previous state via CURVNET_RESTORE(). Recursions on curvnet are permitted, though strongly discuouraged. This change should have no functional impact on nooptions VIMAGE kernel builds, where CURVNET_* macros expand to whitespace. The curthread->td_vnet (aka curvnet) variable's purpose is to be an indicator of the vnet context in which the current network-related operation takes place, in case we cannot deduce the current vnet context from any other source, such as by looking at mbuf's m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so far curvnet has turned out to be an invaluable consistency checking aid: it helps to catch cases when sockets, ifnets or any other vnet-aware structures may have leaked from one vnet to another. The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros was a result of an empirical iterative process, whith an aim to reduce recursions on CURVNET_SET() to a minimum, while still reducing the scope of CURVNET_SET() to networking only operations - the alternative would be calling CURVNET_SET() on each system call entry. In general, curvnet has to be set in three typicall cases: when processing socket-related requests from userspace or from within the kernel; when processing inbound traffic flowing from device drivers to upper layers of the networking stack, and when executing timer-driven networking functions. This change also introduces a DDB subcommand to show the list of all vnet instances. Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
db_printf("\n");
}
}
#endif