postgresql/contrib/sepgsql/label.c

951 lines
24 KiB
C
Raw Normal View History

/* -------------------------------------------------------------------------
*
* contrib/sepgsql/label.c
*
* Routines to support SELinux labels (security context)
*
* Copyright (c) 2010-2019, PostgreSQL Global Development Group
*
* -------------------------------------------------------------------------
*/
#include "postgres.h"
#include <selinux/label.h>
#include "access/genam.h"
#include "access/htup_details.h"
#include "access/table.h"
#include "access/xact.h"
#include "catalog/catalog.h"
#include "catalog/dependency.h"
#include "catalog/indexing.h"
#include "catalog/pg_attribute.h"
#include "catalog/pg_class.h"
#include "catalog/pg_database.h"
#include "catalog/pg_namespace.h"
#include "catalog/pg_proc.h"
#include "commands/dbcommands.h"
#include "commands/seclabel.h"
#include "libpq/auth.h"
#include "libpq/libpq-be.h"
#include "miscadmin.h"
#include "sepgsql.h"
#include "utils/builtins.h"
#include "utils/fmgroids.h"
#include "utils/guc.h"
#include "utils/lsyscache.h"
#include "utils/memutils.h"
#include "utils/rel.h"
/*
* Saved hook entries (if stacked)
*/
static ClientAuthentication_hook_type next_client_auth_hook = NULL;
static needs_fmgr_hook_type next_needs_fmgr_hook = NULL;
static fmgr_hook_type next_fmgr_hook = NULL;
/*
* client_label_*
*
* security label of the database client. Initially the client security label
* is equal to client_label_peer, and can be changed by one or more calls to
* sepgsql_setcon(), and also be temporarily overridden during execution of a
* trusted-procedure.
*
* sepgsql_setcon() is a transaction-aware operation; a (sub-)transaction
* rollback should also rollback the current client security label. Therefore
* we use the list client_label_pending of pending_label to keep track of which
* labels were set during the (sub-)transactions.
*/
static char *client_label_peer = NULL; /* set by getpeercon(3) */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
static List *client_label_pending = NIL; /* pending list being set by
* sepgsql_setcon() */
static char *client_label_committed = NULL; /* set by sepgsql_setcon(), and
* already committed */
static char *client_label_func = NULL; /* set by trusted procedure */
typedef struct
{
SubTransactionId subid;
char *label;
2017-06-21 20:39:04 +02:00
} pending_label;
/*
* sepgsql_get_client_label
*
* Returns the current security label of the client. All code should use this
* routine to get the current label, instead of referring to the client_label_*
* variables above.
*/
char *
sepgsql_get_client_label(void)
{
/* trusted procedure client label override */
if (client_label_func)
return client_label_func;
/* uncommitted sepgsql_setcon() value */
if (client_label_pending)
{
pending_label *plabel = llast(client_label_pending);
if (plabel->label)
return plabel->label;
}
else if (client_label_committed)
return client_label_committed; /* set by sepgsql_setcon() committed */
/* default label */
Assert(client_label_peer != NULL);
return client_label_peer;
}
/*
* sepgsql_set_client_label
*
* This routine tries to switch the current security label of the client, and
* checks related permissions. The supplied new label shall be added to the
* client_label_pending list, then saved at transaction-commit time to ensure
* transaction-awareness.
*/
static void
sepgsql_set_client_label(const char *new_label)
{
const char *tcontext;
MemoryContext oldcxt;
pending_label *plabel;
/* Reset to the initial client label, if NULL */
if (!new_label)
tcontext = client_label_peer;
else
{
if (security_check_context_raw((security_context_t) new_label) < 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_NAME),
errmsg("SELinux: invalid security label: \"%s\"",
new_label)));
tcontext = new_label;
}
/* Check process:{setcurrent} permission. */
sepgsql_avc_check_perms_label(sepgsql_get_client_label(),
SEPG_CLASS_PROCESS,
SEPG_PROCESS__SETCURRENT,
NULL,
true);
/* Check process:{dyntransition} permission. */
sepgsql_avc_check_perms_label(tcontext,
SEPG_CLASS_PROCESS,
SEPG_PROCESS__DYNTRANSITION,
NULL,
true);
/*
* Append the supplied new_label on the pending list until the current
* transaction is committed.
*/
oldcxt = MemoryContextSwitchTo(CurTransactionContext);
plabel = palloc0(sizeof(pending_label));
plabel->subid = GetCurrentSubTransactionId();
if (new_label)
plabel->label = pstrdup(new_label);
client_label_pending = lappend(client_label_pending, plabel);
MemoryContextSwitchTo(oldcxt);
}
/*
* sepgsql_xact_callback
*
2016-03-15 23:06:11 +01:00
* A callback routine of transaction commit/abort/prepare. Commit or abort
* changes in the client_label_pending list.
*/
static void
sepgsql_xact_callback(XactEvent event, void *arg)
{
if (event == XACT_EVENT_COMMIT)
{
if (client_label_pending != NIL)
{
pending_label *plabel = llast(client_label_pending);
char *new_label;
if (plabel->label)
new_label = MemoryContextStrdup(TopMemoryContext,
plabel->label);
else
new_label = NULL;
if (client_label_committed)
pfree(client_label_committed);
client_label_committed = new_label;
/*
* XXX - Note that items of client_label_pending are allocated on
* CurTransactionContext, thus, all acquired memory region shall
* be released implicitly.
*/
client_label_pending = NIL;
}
}
else if (event == XACT_EVENT_ABORT)
client_label_pending = NIL;
}
/*
* sepgsql_subxact_callback
*
* A callback routine of sub-transaction start/abort/commit. Releases all
* security labels that are set within the sub-transaction that is aborted.
*/
static void
sepgsql_subxact_callback(SubXactEvent event, SubTransactionId mySubid,
SubTransactionId parentSubid, void *arg)
{
ListCell *cell;
if (event == SUBXACT_EVENT_ABORT_SUB)
{
Represent Lists as expansible arrays, not chains of cons-cells. Originally, Postgres Lists were a more or less exact reimplementation of Lisp lists, which consist of chains of separately-allocated cons cells, each having a value and a next-cell link. We'd hacked that once before (commit d0b4399d8) to add a separate List header, but the data was still in cons cells. That makes some operations -- notably list_nth() -- O(N), and it's bulky because of the next-cell pointers and per-cell palloc overhead, and it's very cache-unfriendly if the cons cells end up scattered around rather than being adjacent. In this rewrite, we still have List headers, but the data is in a resizable array of values, with no next-cell links. Now we need at most two palloc's per List, and often only one, since we can allocate some values in the same palloc call as the List header. (Of course, extending an existing List may require repalloc's to enlarge the array. But this involves just O(log N) allocations not O(N).) Of course this is not without downsides. The key difficulty is that addition or deletion of a list entry may now cause other entries to move, which it did not before. For example, that breaks foreach() and sister macros, which historically used a pointer to the current cons-cell as loop state. We can repair those macros transparently by making their actual loop state be an integer list index; the exposed "ListCell *" pointer is no longer state carried across loop iterations, but is just a derived value. (In practice, modern compilers can optimize things back to having just one loop state value, at least for simple cases with inline loop bodies.) In principle, this is a semantics change for cases where the loop body inserts or deletes list entries ahead of the current loop index; but I found no such cases in the Postgres code. The change is not at all transparent for code that doesn't use foreach() but chases lists "by hand" using lnext(). The largest share of such code in the backend is in loops that were maintaining "prev" and "next" variables in addition to the current-cell pointer, in order to delete list cells efficiently using list_delete_cell(). However, we no longer need a previous-cell pointer to delete a list cell efficiently. Keeping a next-cell pointer doesn't work, as explained above, but we can improve matters by changing such code to use a regular foreach() loop and then using the new macro foreach_delete_current() to delete the current cell. (This macro knows how to update the associated foreach loop's state so that no cells will be missed in the traversal.) There remains a nontrivial risk of code assuming that a ListCell * pointer will remain good over an operation that could now move the list contents. To help catch such errors, list.c can be compiled with a new define symbol DEBUG_LIST_MEMORY_USAGE that forcibly moves list contents whenever that could possibly happen. This makes list operations significantly more expensive so it's not normally turned on (though it is on by default if USE_VALGRIND is on). There are two notable API differences from the previous code: * lnext() now requires the List's header pointer in addition to the current cell's address. * list_delete_cell() no longer requires a previous-cell argument. These changes are somewhat unfortunate, but on the other hand code using either function needs inspection to see if it is assuming anything it shouldn't, so it's not all bad. Programmers should be aware of these significant performance changes: * list_nth() and related functions are now O(1); so there's no major access-speed difference between a list and an array. * Inserting or deleting a list element now takes time proportional to the distance to the end of the list, due to moving the array elements. (However, it typically *doesn't* require palloc or pfree, so except in long lists it's probably still faster than before.) Notably, lcons() used to be about the same cost as lappend(), but that's no longer true if the list is long. Code that uses lcons() and list_delete_first() to maintain a stack might usefully be rewritten to push and pop at the end of the list rather than the beginning. * There are now list_insert_nth...() and list_delete_nth...() functions that add or remove a list cell identified by index. These have the data-movement penalty explained above, but there's no search penalty. * list_concat() and variants now copy the second list's data into storage belonging to the first list, so there is no longer any sharing of cells between the input lists. The second argument is now declared "const List *" to reflect that it isn't changed. This patch just does the minimum needed to get the new implementation in place and fix bugs exposed by the regression tests. As suggested by the foregoing, there's a fair amount of followup work remaining to do. Also, the ENABLE_LIST_COMPAT macros are finally removed in this commit. Code using those should have been gone a dozen years ago. Patch by me; thanks to David Rowley, Jesper Pedersen, and others for review. Discussion: https://postgr.es/m/11587.1550975080@sss.pgh.pa.us
2019-07-15 19:41:58 +02:00
foreach(cell, client_label_pending)
{
pending_label *plabel = lfirst(cell);
if (plabel->subid == mySubid)
client_label_pending
Represent Lists as expansible arrays, not chains of cons-cells. Originally, Postgres Lists were a more or less exact reimplementation of Lisp lists, which consist of chains of separately-allocated cons cells, each having a value and a next-cell link. We'd hacked that once before (commit d0b4399d8) to add a separate List header, but the data was still in cons cells. That makes some operations -- notably list_nth() -- O(N), and it's bulky because of the next-cell pointers and per-cell palloc overhead, and it's very cache-unfriendly if the cons cells end up scattered around rather than being adjacent. In this rewrite, we still have List headers, but the data is in a resizable array of values, with no next-cell links. Now we need at most two palloc's per List, and often only one, since we can allocate some values in the same palloc call as the List header. (Of course, extending an existing List may require repalloc's to enlarge the array. But this involves just O(log N) allocations not O(N).) Of course this is not without downsides. The key difficulty is that addition or deletion of a list entry may now cause other entries to move, which it did not before. For example, that breaks foreach() and sister macros, which historically used a pointer to the current cons-cell as loop state. We can repair those macros transparently by making their actual loop state be an integer list index; the exposed "ListCell *" pointer is no longer state carried across loop iterations, but is just a derived value. (In practice, modern compilers can optimize things back to having just one loop state value, at least for simple cases with inline loop bodies.) In principle, this is a semantics change for cases where the loop body inserts or deletes list entries ahead of the current loop index; but I found no such cases in the Postgres code. The change is not at all transparent for code that doesn't use foreach() but chases lists "by hand" using lnext(). The largest share of such code in the backend is in loops that were maintaining "prev" and "next" variables in addition to the current-cell pointer, in order to delete list cells efficiently using list_delete_cell(). However, we no longer need a previous-cell pointer to delete a list cell efficiently. Keeping a next-cell pointer doesn't work, as explained above, but we can improve matters by changing such code to use a regular foreach() loop and then using the new macro foreach_delete_current() to delete the current cell. (This macro knows how to update the associated foreach loop's state so that no cells will be missed in the traversal.) There remains a nontrivial risk of code assuming that a ListCell * pointer will remain good over an operation that could now move the list contents. To help catch such errors, list.c can be compiled with a new define symbol DEBUG_LIST_MEMORY_USAGE that forcibly moves list contents whenever that could possibly happen. This makes list operations significantly more expensive so it's not normally turned on (though it is on by default if USE_VALGRIND is on). There are two notable API differences from the previous code: * lnext() now requires the List's header pointer in addition to the current cell's address. * list_delete_cell() no longer requires a previous-cell argument. These changes are somewhat unfortunate, but on the other hand code using either function needs inspection to see if it is assuming anything it shouldn't, so it's not all bad. Programmers should be aware of these significant performance changes: * list_nth() and related functions are now O(1); so there's no major access-speed difference between a list and an array. * Inserting or deleting a list element now takes time proportional to the distance to the end of the list, due to moving the array elements. (However, it typically *doesn't* require palloc or pfree, so except in long lists it's probably still faster than before.) Notably, lcons() used to be about the same cost as lappend(), but that's no longer true if the list is long. Code that uses lcons() and list_delete_first() to maintain a stack might usefully be rewritten to push and pop at the end of the list rather than the beginning. * There are now list_insert_nth...() and list_delete_nth...() functions that add or remove a list cell identified by index. These have the data-movement penalty explained above, but there's no search penalty. * list_concat() and variants now copy the second list's data into storage belonging to the first list, so there is no longer any sharing of cells between the input lists. The second argument is now declared "const List *" to reflect that it isn't changed. This patch just does the minimum needed to get the new implementation in place and fix bugs exposed by the regression tests. As suggested by the foregoing, there's a fair amount of followup work remaining to do. Also, the ENABLE_LIST_COMPAT macros are finally removed in this commit. Code using those should have been gone a dozen years ago. Patch by me; thanks to David Rowley, Jesper Pedersen, and others for review. Discussion: https://postgr.es/m/11587.1550975080@sss.pgh.pa.us
2019-07-15 19:41:58 +02:00
= foreach_delete_current(client_label_pending, cell);
}
}
}
/*
* sepgsql_client_auth
*
* Entrypoint of the client authentication hook.
* It switches the client label according to getpeercon(), and the current
* performing mode according to the GUC setting.
*/
static void
sepgsql_client_auth(Port *port, int status)
{
if (next_client_auth_hook)
(*next_client_auth_hook) (port, status);
/*
* In the case when authentication failed, the supplied socket shall be
* closed soon, so we don't need to do anything here.
*/
if (status != STATUS_OK)
return;
/*
* Getting security label of the peer process using API of libselinux.
*/
if (getpeercon_raw(port->sock, &client_label_peer) < 0)
ereport(FATAL,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: unable to get peer label: %m")));
/*
* Switch the current performing mode from INTERNAL to either DEFAULT or
* PERMISSIVE.
*/
if (sepgsql_get_permissive())
sepgsql_set_mode(SEPGSQL_MODE_PERMISSIVE);
else
sepgsql_set_mode(SEPGSQL_MODE_DEFAULT);
}
/*
* sepgsql_needs_fmgr_hook
*
* It informs the core whether the supplied function is trusted procedure,
* or not. If true, sepgsql_fmgr_hook shall be invoked at start, end, and
* abort time of function invocation.
*/
static bool
sepgsql_needs_fmgr_hook(Oid functionId)
{
ObjectAddress object;
if (next_needs_fmgr_hook &&
(*next_needs_fmgr_hook) (functionId))
return true;
/*
* SELinux needs the function to be called via security_definer wrapper,
* if this invocation will take a domain-transition. We call these
* functions as trusted-procedure, if the security policy has a rule that
* switches security label of the client on execution.
*/
if (sepgsql_avc_trusted_proc(functionId) != NULL)
return true;
/*
* Even if not a trusted-procedure, this function should not be inlined
* unless the client has db_procedure:{execute} permission. Please note
* that it shall be actually failed later because of same reason with
* ACL_EXECUTE.
*/
object.classId = ProcedureRelationId;
object.objectId = functionId;
object.objectSubId = 0;
if (!sepgsql_avc_check_perms(&object,
SEPG_CLASS_DB_PROCEDURE,
SEPG_DB_PROCEDURE__EXECUTE |
SEPG_DB_PROCEDURE__ENTRYPOINT,
SEPGSQL_AVC_NOAUDIT, false))
return true;
return false;
}
/*
* sepgsql_fmgr_hook
*
* It switches security label of the client on execution of trusted
* procedures.
*/
static void
sepgsql_fmgr_hook(FmgrHookEventType event,
FmgrInfo *flinfo, Datum *private)
{
struct
{
char *old_label;
char *new_label;
Datum next_private;
} *stack;
switch (event)
{
case FHET_START:
stack = (void *) DatumGetPointer(*private);
if (!stack)
{
MemoryContext oldcxt;
oldcxt = MemoryContextSwitchTo(flinfo->fn_mcxt);
stack = palloc(sizeof(*stack));
stack->old_label = NULL;
stack->new_label = sepgsql_avc_trusted_proc(flinfo->fn_oid);
stack->next_private = 0;
MemoryContextSwitchTo(oldcxt);
/*
* process:transition permission between old and new label,
* when user tries to switch security label of the client on
* execution of trusted procedure.
*
* Also, db_procedure:entrypoint permission should be checked
* whether this procedure can perform as an entrypoint of the
* trusted procedure, or not. Note that db_procedure:execute
* permission shall be checked individually.
*/
if (stack->new_label)
{
ObjectAddress object;
object.classId = ProcedureRelationId;
object.objectId = flinfo->fn_oid;
object.objectSubId = 0;
sepgsql_avc_check_perms(&object,
SEPG_CLASS_DB_PROCEDURE,
SEPG_DB_PROCEDURE__ENTRYPOINT,
getObjectDescription(&object),
true);
sepgsql_avc_check_perms_label(stack->new_label,
SEPG_CLASS_PROCESS,
SEPG_PROCESS__TRANSITION,
NULL, true);
}
*private = PointerGetDatum(stack);
}
Assert(!stack->old_label);
if (stack->new_label)
{
stack->old_label = client_label_func;
client_label_func = stack->new_label;
}
if (next_fmgr_hook)
(*next_fmgr_hook) (event, flinfo, &stack->next_private);
break;
case FHET_END:
case FHET_ABORT:
stack = (void *) DatumGetPointer(*private);
if (next_fmgr_hook)
(*next_fmgr_hook) (event, flinfo, &stack->next_private);
if (stack->new_label)
{
client_label_func = stack->old_label;
stack->old_label = NULL;
}
break;
default:
elog(ERROR, "unexpected event type: %d", (int) event);
break;
}
}
/*
* sepgsql_init_client_label
*
* Initializes the client security label and sets up related hooks for client
* label management.
*/
void
sepgsql_init_client_label(void)
{
/*
* Set up dummy client label.
*
* XXX - note that PostgreSQL launches background worker process like
* autovacuum without authentication steps. So, we initialize sepgsql_mode
* with SEPGSQL_MODE_INTERNAL, and client_label with the security context
* of server process. Later, it also launches background of user session.
* In this case, the process is always hooked on post-authentication, and
* we can initialize the sepgsql_mode and client_label correctly.
*/
if (getcon_raw(&client_label_peer) < 0)
ereport(ERROR,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: failed to get server security label: %m")));
/* Client authentication hook */
next_client_auth_hook = ClientAuthentication_hook;
ClientAuthentication_hook = sepgsql_client_auth;
/* Trusted procedure hooks */
next_needs_fmgr_hook = needs_fmgr_hook;
needs_fmgr_hook = sepgsql_needs_fmgr_hook;
next_fmgr_hook = fmgr_hook;
fmgr_hook = sepgsql_fmgr_hook;
/* Transaction/Sub-transaction callbacks */
RegisterXactCallback(sepgsql_xact_callback, NULL);
RegisterSubXactCallback(sepgsql_subxact_callback, NULL);
}
/*
* sepgsql_get_label
*
* It returns a security context of the specified database object.
* If unlabeled or incorrectly labeled, the system "unlabeled" label
* shall be returned.
*/
char *
sepgsql_get_label(Oid classId, Oid objectId, int32 subId)
{
2011-04-10 17:42:00 +02:00
ObjectAddress object;
char *label;
2011-04-10 17:42:00 +02:00
object.classId = classId;
object.objectId = objectId;
object.objectSubId = subId;
label = GetSecurityLabel(&object, SEPGSQL_LABEL_TAG);
2011-04-10 17:42:00 +02:00
if (!label || security_check_context_raw((security_context_t) label))
{
2011-04-10 17:42:00 +02:00
security_context_t unlabeled;
if (security_get_initial_context_raw("unlabeled", &unlabeled) < 0)
ereport(ERROR,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: failed to get initial security label: %m")));
PG_TRY();
{
label = pstrdup(unlabeled);
}
PG_CATCH();
{
freecon(unlabeled);
PG_RE_THROW();
}
PG_END_TRY();
freecon(unlabeled);
}
return label;
}
/*
* sepgsql_object_relabel
*
* An entrypoint of SECURITY LABEL statement
*/
void
sepgsql_object_relabel(const ObjectAddress *object, const char *seclabel)
{
/*
2011-04-10 17:42:00 +02:00
* validate format of the supplied security label, if it is security
* context of selinux.
*/
if (seclabel &&
security_check_context_raw((security_context_t) seclabel) < 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_NAME),
errmsg("SELinux: invalid security label: \"%s\"", seclabel)));
2011-04-10 17:42:00 +02:00
/*
* Do actual permission checks for each object classes
*/
switch (object->classId)
{
case DatabaseRelationId:
sepgsql_database_relabel(object->objectId, seclabel);
break;
case NamespaceRelationId:
2011-04-10 17:42:00 +02:00
sepgsql_schema_relabel(object->objectId, seclabel);
break;
case RelationRelationId:
if (object->objectSubId == 0)
sepgsql_relation_relabel(object->objectId,
seclabel);
else
sepgsql_attribute_relabel(object->objectId,
object->objectSubId,
seclabel);
break;
case ProcedureRelationId:
sepgsql_proc_relabel(object->objectId, seclabel);
break;
default:
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("sepgsql provider does not support labels on %s",
getObjectTypeDescription(object))));
break;
}
}
/*
* TEXT sepgsql_getcon(VOID)
*
* It returns the security label of the client.
*/
PG_FUNCTION_INFO_V1(sepgsql_getcon);
Datum
sepgsql_getcon(PG_FUNCTION_ARGS)
{
2011-04-10 17:42:00 +02:00
char *client_label;
if (!sepgsql_is_enabled())
PG_RETURN_NULL();
client_label = sepgsql_get_client_label();
PG_RETURN_TEXT_P(cstring_to_text(client_label));
}
/*
* BOOL sepgsql_setcon(TEXT)
*
* It switches the security label of the client.
*/
PG_FUNCTION_INFO_V1(sepgsql_setcon);
Datum
sepgsql_setcon(PG_FUNCTION_ARGS)
{
const char *new_label;
if (PG_ARGISNULL(0))
new_label = NULL;
else
new_label = TextDatumGetCString(PG_GETARG_DATUM(0));
sepgsql_set_client_label(new_label);
PG_RETURN_BOOL(true);
}
/*
* TEXT sepgsql_mcstrans_in(TEXT)
*
* It translate the given qualified MLS/MCS range into raw format
* when mcstrans daemon is working.
*/
PG_FUNCTION_INFO_V1(sepgsql_mcstrans_in);
Datum
sepgsql_mcstrans_in(PG_FUNCTION_ARGS)
{
text *label = PG_GETARG_TEXT_PP(0);
2011-04-10 17:42:00 +02:00
char *raw_label;
char *result;
if (!sepgsql_is_enabled())
ereport(ERROR,
(errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
errmsg("sepgsql is not enabled")));
if (selinux_trans_to_raw_context(text_to_cstring(label),
&raw_label) < 0)
ereport(ERROR,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: could not translate security label: %m")));
PG_TRY();
{
result = pstrdup(raw_label);
}
PG_CATCH();
{
freecon(raw_label);
PG_RE_THROW();
}
PG_END_TRY();
freecon(raw_label);
PG_RETURN_TEXT_P(cstring_to_text(result));
}
/*
* TEXT sepgsql_mcstrans_out(TEXT)
*
* It translate the given raw MLS/MCS range into qualified format
* when mcstrans daemon is working.
*/
PG_FUNCTION_INFO_V1(sepgsql_mcstrans_out);
Datum
sepgsql_mcstrans_out(PG_FUNCTION_ARGS)
{
text *label = PG_GETARG_TEXT_PP(0);
2011-04-10 17:42:00 +02:00
char *qual_label;
char *result;
if (!sepgsql_is_enabled())
ereport(ERROR,
(errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
errmsg("sepgsql is not currently enabled")));
if (selinux_raw_to_trans_context(text_to_cstring(label),
&qual_label) < 0)
ereport(ERROR,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: could not translate security label: %m")));
PG_TRY();
{
result = pstrdup(qual_label);
}
PG_CATCH();
{
freecon(qual_label);
PG_RE_THROW();
}
PG_END_TRY();
freecon(qual_label);
PG_RETURN_TEXT_P(cstring_to_text(result));
}
/*
* quote_object_name
*
* It tries to quote the supplied identifiers
*/
static char *
quote_object_name(const char *src1, const char *src2,
const char *src3, const char *src4)
{
2011-04-10 17:42:00 +02:00
StringInfoData result;
const char *temp;
initStringInfo(&result);
if (src1)
{
temp = quote_identifier(src1);
appendStringInfoString(&result, temp);
if (src1 != temp)
2011-04-10 17:42:00 +02:00
pfree((void *) temp);
}
if (src2)
{
temp = quote_identifier(src2);
appendStringInfo(&result, ".%s", temp);
if (src2 != temp)
2011-04-10 17:42:00 +02:00
pfree((void *) temp);
}
if (src3)
{
temp = quote_identifier(src3);
appendStringInfo(&result, ".%s", temp);
if (src3 != temp)
2011-04-10 17:42:00 +02:00
pfree((void *) temp);
}
if (src4)
{
temp = quote_identifier(src4);
appendStringInfo(&result, ".%s", temp);
if (src4 != temp)
2011-04-10 17:42:00 +02:00
pfree((void *) temp);
}
return result.data;
}
/*
* exec_object_restorecon
*
* This routine is a helper called by sepgsql_restorecon; it set up
* initial security labels of database objects within the supplied
* catalog OID.
*/
static void
2017-06-21 20:39:04 +02:00
exec_object_restorecon(struct selabel_handle *sehnd, Oid catalogId)
{
2011-04-10 17:42:00 +02:00
Relation rel;
SysScanDesc sscan;
HeapTuple tuple;
char *database_name = get_database_name(MyDatabaseId);
char *namespace_name;
Oid namespace_id;
char *relation_name;
/*
2011-04-10 17:42:00 +02:00
* Open the target catalog. We don't want to allow writable accesses by
* other session during initial labeling.
*/
rel = table_open(catalogId, AccessShareLock);
sscan = systable_beginscan(rel, InvalidOid, false,
NULL, 0, NULL);
while (HeapTupleIsValid(tuple = systable_getnext(sscan)))
{
Form_pg_database datForm;
2011-04-10 17:42:00 +02:00
Form_pg_namespace nspForm;
Form_pg_class relForm;
Form_pg_attribute attForm;
Form_pg_proc proForm;
char *objname;
int objtype = 1234;
ObjectAddress object;
security_context_t context;
/*
2011-04-10 17:42:00 +02:00
* The way to determine object name depends on object classes. So, any
* branches set up `objtype', `objname' and `object' here.
*/
switch (catalogId)
{
case DatabaseRelationId:
datForm = (Form_pg_database) GETSTRUCT(tuple);
objtype = SELABEL_DB_DATABASE;
objname = quote_object_name(NameStr(datForm->datname),
NULL, NULL, NULL);
object.classId = DatabaseRelationId;
Remove WITH OIDS support, change oid catalog column visibility. Previously tables declared WITH OIDS, including a significant fraction of the catalog tables, stored the oid column not as a normal column, but as part of the tuple header. This special column was not shown by default, which was somewhat odd, as it's often (consider e.g. pg_class.oid) one of the more important parts of a row. Neither pg_dump nor COPY included the contents of the oid column by default. The fact that the oid column was not an ordinary column necessitated a significant amount of special case code to support oid columns. That already was painful for the existing, but upcoming work aiming to make table storage pluggable, would have required expanding and duplicating that "specialness" significantly. WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0). Remove it. Removing includes: - CREATE TABLE and ALTER TABLE syntax for declaring the table to be WITH OIDS has been removed (WITH (oids[ = true]) will error out) - pg_dump does not support dumping tables declared WITH OIDS and will issue a warning when dumping one (and ignore the oid column). - restoring an pg_dump archive with pg_restore will warn when restoring a table with oid contents (and ignore the oid column) - COPY will refuse to load binary dump that includes oids. - pg_upgrade will error out when encountering tables declared WITH OIDS, they have to be altered to remove the oid column first. - Functionality to access the oid of the last inserted row (like plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed. The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false) for CREATE TABLE) is still supported. While that requires a bit of support code, it seems unnecessary to break applications / dumps that do not use oids, and are explicit about not using them. The biggest user of WITH OID columns was postgres' catalog. This commit changes all 'magic' oid columns to be columns that are normally declared and stored. To reduce unnecessary query breakage all the newly added columns are still named 'oid', even if a table's column naming scheme would indicate 'reloid' or such. This obviously requires adapting a lot code, mostly replacing oid access via HeapTupleGetOid() with access to the underlying Form_pg_*->oid column. The bootstrap process now assigns oids for all oid columns in genbki.pl that do not have an explicit value (starting at the largest oid previously used), only oids assigned later by oids will be above FirstBootstrapObjectId. As the oid column now is a normal column the special bootstrap syntax for oids has been removed. Oids are not automatically assigned during insertion anymore, all backend code explicitly assigns oids with GetNewOidWithIndex(). For the rare case that insertions into the catalog via SQL are called for the new pg_nextoid() function can be used (which only works on catalog tables). The fact that oid columns on system tables are now normal columns means that they will be included in the set of columns expanded by * (i.e. SELECT * FROM pg_class will now include the table's oid, previously it did not). It'd not technically be hard to hide oid column by default, but that'd mean confusing behavior would either have to be carried forward forever, or it'd cause breakage down the line. While it's not unlikely that further adjustments are needed, the scope/invasiveness of the patch makes it worthwhile to get merge this now. It's painful to maintain externally, too complicated to commit after the code code freeze, and a dependency of a number of other patches. Catversion bump, for obvious reasons. Author: Andres Freund, with contributions by John Naylor Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
object.objectId = datForm->oid;
object.objectSubId = 0;
break;
case NamespaceRelationId:
nspForm = (Form_pg_namespace) GETSTRUCT(tuple);
objtype = SELABEL_DB_SCHEMA;
objname = quote_object_name(database_name,
NameStr(nspForm->nspname),
NULL, NULL);
object.classId = NamespaceRelationId;
Remove WITH OIDS support, change oid catalog column visibility. Previously tables declared WITH OIDS, including a significant fraction of the catalog tables, stored the oid column not as a normal column, but as part of the tuple header. This special column was not shown by default, which was somewhat odd, as it's often (consider e.g. pg_class.oid) one of the more important parts of a row. Neither pg_dump nor COPY included the contents of the oid column by default. The fact that the oid column was not an ordinary column necessitated a significant amount of special case code to support oid columns. That already was painful for the existing, but upcoming work aiming to make table storage pluggable, would have required expanding and duplicating that "specialness" significantly. WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0). Remove it. Removing includes: - CREATE TABLE and ALTER TABLE syntax for declaring the table to be WITH OIDS has been removed (WITH (oids[ = true]) will error out) - pg_dump does not support dumping tables declared WITH OIDS and will issue a warning when dumping one (and ignore the oid column). - restoring an pg_dump archive with pg_restore will warn when restoring a table with oid contents (and ignore the oid column) - COPY will refuse to load binary dump that includes oids. - pg_upgrade will error out when encountering tables declared WITH OIDS, they have to be altered to remove the oid column first. - Functionality to access the oid of the last inserted row (like plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed. The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false) for CREATE TABLE) is still supported. While that requires a bit of support code, it seems unnecessary to break applications / dumps that do not use oids, and are explicit about not using them. The biggest user of WITH OID columns was postgres' catalog. This commit changes all 'magic' oid columns to be columns that are normally declared and stored. To reduce unnecessary query breakage all the newly added columns are still named 'oid', even if a table's column naming scheme would indicate 'reloid' or such. This obviously requires adapting a lot code, mostly replacing oid access via HeapTupleGetOid() with access to the underlying Form_pg_*->oid column. The bootstrap process now assigns oids for all oid columns in genbki.pl that do not have an explicit value (starting at the largest oid previously used), only oids assigned later by oids will be above FirstBootstrapObjectId. As the oid column now is a normal column the special bootstrap syntax for oids has been removed. Oids are not automatically assigned during insertion anymore, all backend code explicitly assigns oids with GetNewOidWithIndex(). For the rare case that insertions into the catalog via SQL are called for the new pg_nextoid() function can be used (which only works on catalog tables). The fact that oid columns on system tables are now normal columns means that they will be included in the set of columns expanded by * (i.e. SELECT * FROM pg_class will now include the table's oid, previously it did not). It'd not technically be hard to hide oid column by default, but that'd mean confusing behavior would either have to be carried forward forever, or it'd cause breakage down the line. While it's not unlikely that further adjustments are needed, the scope/invasiveness of the patch makes it worthwhile to get merge this now. It's painful to maintain externally, too complicated to commit after the code code freeze, and a dependency of a number of other patches. Catversion bump, for obvious reasons. Author: Andres Freund, with contributions by John Naylor Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
object.objectId = nspForm->oid;
object.objectSubId = 0;
break;
case RelationRelationId:
relForm = (Form_pg_class) GETSTRUCT(tuple);
if (relForm->relkind == RELKIND_RELATION ||
relForm->relkind == RELKIND_PARTITIONED_TABLE)
objtype = SELABEL_DB_TABLE;
else if (relForm->relkind == RELKIND_SEQUENCE)
objtype = SELABEL_DB_SEQUENCE;
else if (relForm->relkind == RELKIND_VIEW)
objtype = SELABEL_DB_VIEW;
else
continue; /* no need to assign security label */
namespace_name = get_namespace_name(relForm->relnamespace);
objname = quote_object_name(database_name,
namespace_name,
NameStr(relForm->relname),
NULL);
pfree(namespace_name);
object.classId = RelationRelationId;
Remove WITH OIDS support, change oid catalog column visibility. Previously tables declared WITH OIDS, including a significant fraction of the catalog tables, stored the oid column not as a normal column, but as part of the tuple header. This special column was not shown by default, which was somewhat odd, as it's often (consider e.g. pg_class.oid) one of the more important parts of a row. Neither pg_dump nor COPY included the contents of the oid column by default. The fact that the oid column was not an ordinary column necessitated a significant amount of special case code to support oid columns. That already was painful for the existing, but upcoming work aiming to make table storage pluggable, would have required expanding and duplicating that "specialness" significantly. WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0). Remove it. Removing includes: - CREATE TABLE and ALTER TABLE syntax for declaring the table to be WITH OIDS has been removed (WITH (oids[ = true]) will error out) - pg_dump does not support dumping tables declared WITH OIDS and will issue a warning when dumping one (and ignore the oid column). - restoring an pg_dump archive with pg_restore will warn when restoring a table with oid contents (and ignore the oid column) - COPY will refuse to load binary dump that includes oids. - pg_upgrade will error out when encountering tables declared WITH OIDS, they have to be altered to remove the oid column first. - Functionality to access the oid of the last inserted row (like plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed. The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false) for CREATE TABLE) is still supported. While that requires a bit of support code, it seems unnecessary to break applications / dumps that do not use oids, and are explicit about not using them. The biggest user of WITH OID columns was postgres' catalog. This commit changes all 'magic' oid columns to be columns that are normally declared and stored. To reduce unnecessary query breakage all the newly added columns are still named 'oid', even if a table's column naming scheme would indicate 'reloid' or such. This obviously requires adapting a lot code, mostly replacing oid access via HeapTupleGetOid() with access to the underlying Form_pg_*->oid column. The bootstrap process now assigns oids for all oid columns in genbki.pl that do not have an explicit value (starting at the largest oid previously used), only oids assigned later by oids will be above FirstBootstrapObjectId. As the oid column now is a normal column the special bootstrap syntax for oids has been removed. Oids are not automatically assigned during insertion anymore, all backend code explicitly assigns oids with GetNewOidWithIndex(). For the rare case that insertions into the catalog via SQL are called for the new pg_nextoid() function can be used (which only works on catalog tables). The fact that oid columns on system tables are now normal columns means that they will be included in the set of columns expanded by * (i.e. SELECT * FROM pg_class will now include the table's oid, previously it did not). It'd not technically be hard to hide oid column by default, but that'd mean confusing behavior would either have to be carried forward forever, or it'd cause breakage down the line. While it's not unlikely that further adjustments are needed, the scope/invasiveness of the patch makes it worthwhile to get merge this now. It's painful to maintain externally, too complicated to commit after the code code freeze, and a dependency of a number of other patches. Catversion bump, for obvious reasons. Author: Andres Freund, with contributions by John Naylor Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
object.objectId = relForm->oid;
object.objectSubId = 0;
break;
case AttributeRelationId:
attForm = (Form_pg_attribute) GETSTRUCT(tuple);
if (get_rel_relkind(attForm->attrelid) != RELKIND_RELATION &&
get_rel_relkind(attForm->attrelid) != RELKIND_PARTITIONED_TABLE)
continue; /* no need to assign security label */
objtype = SELABEL_DB_COLUMN;
namespace_id = get_rel_namespace(attForm->attrelid);
namespace_name = get_namespace_name(namespace_id);
relation_name = get_rel_name(attForm->attrelid);
objname = quote_object_name(database_name,
namespace_name,
relation_name,
NameStr(attForm->attname));
pfree(namespace_name);
pfree(relation_name);
object.classId = RelationRelationId;
object.objectId = attForm->attrelid;
object.objectSubId = attForm->attnum;
break;
case ProcedureRelationId:
proForm = (Form_pg_proc) GETSTRUCT(tuple);
objtype = SELABEL_DB_PROCEDURE;
namespace_name = get_namespace_name(proForm->pronamespace);
objname = quote_object_name(database_name,
namespace_name,
NameStr(proForm->proname),
NULL);
pfree(namespace_name);
object.classId = ProcedureRelationId;
Remove WITH OIDS support, change oid catalog column visibility. Previously tables declared WITH OIDS, including a significant fraction of the catalog tables, stored the oid column not as a normal column, but as part of the tuple header. This special column was not shown by default, which was somewhat odd, as it's often (consider e.g. pg_class.oid) one of the more important parts of a row. Neither pg_dump nor COPY included the contents of the oid column by default. The fact that the oid column was not an ordinary column necessitated a significant amount of special case code to support oid columns. That already was painful for the existing, but upcoming work aiming to make table storage pluggable, would have required expanding and duplicating that "specialness" significantly. WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0). Remove it. Removing includes: - CREATE TABLE and ALTER TABLE syntax for declaring the table to be WITH OIDS has been removed (WITH (oids[ = true]) will error out) - pg_dump does not support dumping tables declared WITH OIDS and will issue a warning when dumping one (and ignore the oid column). - restoring an pg_dump archive with pg_restore will warn when restoring a table with oid contents (and ignore the oid column) - COPY will refuse to load binary dump that includes oids. - pg_upgrade will error out when encountering tables declared WITH OIDS, they have to be altered to remove the oid column first. - Functionality to access the oid of the last inserted row (like plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed. The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false) for CREATE TABLE) is still supported. While that requires a bit of support code, it seems unnecessary to break applications / dumps that do not use oids, and are explicit about not using them. The biggest user of WITH OID columns was postgres' catalog. This commit changes all 'magic' oid columns to be columns that are normally declared and stored. To reduce unnecessary query breakage all the newly added columns are still named 'oid', even if a table's column naming scheme would indicate 'reloid' or such. This obviously requires adapting a lot code, mostly replacing oid access via HeapTupleGetOid() with access to the underlying Form_pg_*->oid column. The bootstrap process now assigns oids for all oid columns in genbki.pl that do not have an explicit value (starting at the largest oid previously used), only oids assigned later by oids will be above FirstBootstrapObjectId. As the oid column now is a normal column the special bootstrap syntax for oids has been removed. Oids are not automatically assigned during insertion anymore, all backend code explicitly assigns oids with GetNewOidWithIndex(). For the rare case that insertions into the catalog via SQL are called for the new pg_nextoid() function can be used (which only works on catalog tables). The fact that oid columns on system tables are now normal columns means that they will be included in the set of columns expanded by * (i.e. SELECT * FROM pg_class will now include the table's oid, previously it did not). It'd not technically be hard to hide oid column by default, but that'd mean confusing behavior would either have to be carried forward forever, or it'd cause breakage down the line. While it's not unlikely that further adjustments are needed, the scope/invasiveness of the patch makes it worthwhile to get merge this now. It's painful to maintain externally, too complicated to commit after the code code freeze, and a dependency of a number of other patches. Catversion bump, for obvious reasons. Author: Andres Freund, with contributions by John Naylor Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
object.objectId = proForm->oid;
object.objectSubId = 0;
break;
default:
elog(ERROR, "unexpected catalog id: %u", catalogId);
2011-04-10 17:42:00 +02:00
objname = NULL; /* for compiler quiet */
break;
}
if (selabel_lookup_raw(sehnd, &context, objname, objtype) == 0)
{
PG_TRY();
{
/*
* Check SELinux permission to relabel the fetched object,
* then do the actual relabeling.
*/
sepgsql_object_relabel(&object, context);
SetSecurityLabel(&object, SEPGSQL_LABEL_TAG, context);
}
PG_CATCH();
{
freecon(context);
PG_RE_THROW();
}
PG_END_TRY();
freecon(context);
}
else if (errno == ENOENT)
ereport(WARNING,
(errmsg("SELinux: no initial label assigned for %s (type=%d), skipping",
objname, objtype)));
else
ereport(ERROR,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: could not determine initial security label for %s (type=%d): %m", objname, objtype)));
pfree(objname);
}
systable_endscan(sscan);
table_close(rel, NoLock);
}
/*
* BOOL sepgsql_restorecon(TEXT specfile)
*
* This function tries to assign initial security labels on all the object
* within the current database, according to the system setting.
* It is typically invoked by sepgsql-install script just after initdb, to
* assign initial security labels.
*
* If @specfile is not NULL, it uses explicitly specified specfile, instead
* of the system default.
*/
PG_FUNCTION_INFO_V1(sepgsql_restorecon);
Datum
sepgsql_restorecon(PG_FUNCTION_ARGS)
{
2011-04-10 17:42:00 +02:00
struct selabel_handle *sehnd;
struct selinux_opt seopts;
/*
* SELinux has to be enabled on the running platform.
*/
if (!sepgsql_is_enabled())
ereport(ERROR,
(errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
errmsg("sepgsql is not currently enabled")));
2011-04-10 17:42:00 +02:00
/*
2011-04-10 17:42:00 +02:00
* Check DAC permission. Only superuser can set up initial security
* labels, like root-user in filesystems
*/
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
errmsg("SELinux: must be superuser to restore initial contexts")));
/*
2011-04-10 17:42:00 +02:00
* Open selabel_lookup(3) stuff. It provides a set of mapping between an
* initial security label and object class/name due to the system setting.
*/
if (PG_ARGISNULL(0))
{
seopts.type = SELABEL_OPT_UNUSED;
seopts.value = NULL;
}
else
{
seopts.type = SELABEL_OPT_PATH;
seopts.value = TextDatumGetCString(PG_GETARG_DATUM(0));
}
sehnd = selabel_open(SELABEL_CTX_DB, &seopts, 1);
if (!sehnd)
ereport(ERROR,
(errcode(ERRCODE_INTERNAL_ERROR),
errmsg("SELinux: failed to initialize labeling handle: %m")));
PG_TRY();
{
exec_object_restorecon(sehnd, DatabaseRelationId);
exec_object_restorecon(sehnd, NamespaceRelationId);
exec_object_restorecon(sehnd, RelationRelationId);
exec_object_restorecon(sehnd, AttributeRelationId);
exec_object_restorecon(sehnd, ProcedureRelationId);
}
PG_CATCH();
{
selabel_close(sehnd);
PG_RE_THROW();
}
2011-04-10 17:42:00 +02:00
PG_END_TRY();
selabel_close(sehnd);
PG_RETURN_BOOL(true);
}