Menu
▾
▴
RE: [LTP] [PATCH] new testcase setegid01 (small modification)
|
From: Aniruddha M M. <ani...@wi...> - 2003-04-09 13:25:09
|
Please consider this file instead of the previous one.=20
Thanks
Aniruddha.
|-----Original Message-----
|From: Aniruddha M Marathe=20
|Sent: Wednesday, April 09, 2003 5:31 PM
|To: Dan Kegel
|Cc: Robert Williamson; ltp...@li...
|Subject: RE: [LTP] [PATCH] new testcase setegid01
|
|
|Right now, Madhu is busy with some other work. However, I had=20
|discussion with him on this issue.
|He studied the issue and agreed with Dan's comment.
|We should fix getresgid02.
|I have attached the modified file. Please check the same.
|
|Thanks,
|Aniruddha
|---------------------------------------------------------------
|Mr. Aniruddha Marathe
|Systems Engineer,
|WIPRO technologies,
|53/1, Hosur road,
|Madivala,
|Bangalore - 560068
|Karnataka, India
|Phone: +91-80-5502001 extension 5092
|E-mail: ani...@wi...
|---------------------------------------------------------------
|
|
||-----Original Message-----
||From: Dan Kegel [mailto:da...@ke...]=20
||Sent: Tuesday, April 08, 2003 10:32 PM
||To: Robert Williamson
||Cc: 'ltp...@li...'; mck...@co...;=20
||sf...@ca...; madhu tarikere lakshminaryan
||Subject: Re: [LTP] [PATCH] new testcase setegid01
||
||
||Robert Williamson wrote:
||> I added this new testcase to the LTP and the runtest/syscall=20
||script. I was
||> actually made aware of this bug by someone working on ppc=20
||testing here, but
||> forgot about it.....thanks for reminding me. How do you=20
||think we should go
||> about fixing setresgid02?
||
||First off, let's ask Madhu, he wrote the test.
||But here's my guess:
||
||Notice the comment in setresgid02.setup():
||
|| /* Set effective group id by setregid, since root, sets=20
||saved gid also */
|| if (setregid(-1,nobody_gid) =3D=3D -1) {
||
||My first impression is that line should change to be
||more like the one in setresgid03.setup(), i.e.
||
|| /* Set real/effective/saved gid to root/nobody/nobody */
|| if (setresgid(root_gid, nobody_gid, nobody_gid) =3D=3D -1) {
||
||but I really want to know whether Madhu still believes that
||commment about "since root, sets saved gid also" after
||reading the Posix spec for setregid().
||- Dan
||
||>=20
||> - Robbie
||>=20
||> Robert V. Williamson <ro...@us...>
||> Linux Test Project
||> IBM Linux Technology Center
||> Phone: (512) 838-9295 T/L: 678-9295
||> Fax: (512) 838-4603
||> Web: http://ltp.sourceforge.net
||> IRC: #ltp on freenode.irc.net
||> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
||> "Only two things are infinite, the universe and human=20
||stupidity, and I'm
||> not sure about the former." -Albert Einstein
||>=20
||>=20
||> =20
|| =20
|| =20
||> Dan Kegel =20
|| =20
|| =20
||> <da...@ke...> To: =20
|| "'ltp...@li...'"=20
||<ltp...@li...> =20
||> Sent by: cc: =20
|| sf...@ca..., mck...@co... =20
|| =20
||> ltp...@li...ur =20
||Subject: [LTP] [PATCH] new testcase setegid01 (was:=20
||Bug#182386: libc6: setegid on =20
||> ceforge.net i386=20
||uses setregid and therefore sets saved gid) =20
|| =20
||> =20
|| =20
|| =20
||> =20
|| =20
|| =20
||> 04/06/2003 10:33 PM =20
|| =20
|| =20
||> =20
|| =20
|| =20
||>=20
||>=20
||>=20
||>=20
||> On 24 Feb 2003, Stephen Rothwell (sf...@ca...)=20
||wrote in newsgroup
||> linux.debian.maint.glibc:
||>=20
||>>One of the LTP tests (setresgid02) was failing on PPC64 and=20
||succeeding on
||>>I386 and it turns out that the test makes the assumption=20
||that setegid()
||>>sets the saved gid as well as the effective gid. According to the
||>=20
||> POSIX.1
||>=20
||>>2001 (SUSv3) setegid should not change the saved gid. On=20
||i386 setegid()
||>>is currently using setregid32() at the system call level=20
||while on ppc64
||>>it uses setresgid(). In the glibc sources, it appears that=20
|setegid()
||>>should be using setresgid() on i386, but strace/ltrace shows=20
||otherwise.
||>>
||>>The setegid manual page actually says:
||>>
||>> Under libc4, libc5 and glibc2.0 seteuid(euid) =20
||is equivalent
||>> to setreuid(-1, euid) and hence may change the=20
||saved user
||>> ID. Under glibc2.1 it is equivalent to=20
|setresuid(-1, euid,-1)
||>> and hence does not change the saved user ID. =20
|Similar remarks
||>> hold for setegid.
||>>
||>>I am not sure if this is just a bug in the way glibc is=20
||built for i386,
||>>or some strange interaction with the headre files.
||>=20
||>=20
||> I am also seeing setresgid02 failing on ppc32 and sh4 with glibc.
||> Here's the output on ppc32 with a couple added prints:
||>=20
||> # ./setresgid02
||> Calling setegid(99)
||> calling setresgid(-1, -1, -1)
||> getresgid reports rgid 0, egid 99, sgid 0
||> Expected rgid 0, egid 99, sgid 99
||> setresgid02 1 FAIL : Functionality test for=20
||setresgid(-1, -1, -1)
||> failed
||>=20
||> And (if we can trust it) here's what strace says:
||> write(1, "Calling setegid(99)\n", 20) =3D 20
||> setresgid(rgid 4294967295, egid 99, sgid 4294967295) =3D 0
||> write(1, "calling setresgid(-1, -1, -1)\n", 30) =3D 30
||> setresgid(rgid 4294967295, egid 4294967295, sgid 4294967295) =3D 0
||> getresgid(rgid ..., egid ..., sgid ...) =3D 0
||> write(1, "getresgid reports rgid 0, egid 9"..., 42) =3D 42
||> write(1, "Expected rgid 0, egid 99, sgid 9"..., 34) =3D 34
||> write(1, "setresgid02 1 FAIL : Funct"..., 79) =3D 79
||>=20
||> The same thing on x86 outputs:
||>=20
||> Calling setegid(99)
||> calling setresgid(-1, -1, -1)
||> getresgid reports rgid 0, egid 99, sgid 99
||> Expected rgid 0, egid 99, sgid 99
||> setresgid02 1 PASS : Test for setresgid(-1, -1, -1) successful
||>=20
||> And strace says:
||> write(1, "Calling setegid(99)\n", 20) =3D 20
||> setregid32(0xffffffff, 0x63) =3D 0
||> write(1, "calling setresgid(-1, -1, -1)\n", 30) =3D 30
||> setresgid32(0xffffffff, 0xffffffff, 0xffffffff) =3D 0
||> getsetgid32(0xbffffa44, 0xbffffa40, 0xbffffa3c) =3D 0
||> write(1, "getresgid reports rgid 0, egid 9"..., 43) =3D 43
||> write(1, "Expected rgid 0, egid 99, sgid 9"..., 34) =3D 34
||> write(1, "setresgid02 1 PASS : Test "..., 69) =3D 69
||>=20
||> It looks rather as if glibc has a bug in it in x86, just as
||> Stephen Rothwell asserted. I've written a new LTP testcase
||> for this, and it seems to confirm his theory.
||>=20
||> This means that the test setresgid02 is broken; it was
||> written as if the behavior on x86 is correct.
||>=20
||> Or so it seems to me. Anyone see a hole in this?
||> If people agree, I guess I should send a bug report in to glibc.
||>=20
||> Thanks,
||> Dan
||>=20
||> --
||> Dan Kegel
||> http://www.kegel.com
||>=20
|http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=3D78045
||> --- /dev/null 2002-08-30 16:31:37.000000000 -0700
||> +++ ltp-full-20030404/testcases/kernel/syscalls/setegid/setegid01.c
||> 2003-04-06 20:28:25.000000000 -0700
||> @@ -0,0 +1,115 @@
||> +/*
||> + * Copyright (c) Dan Kegel 2003
||> + *
||> + * This program is free software; you can redistribute=20
||it and/or modify
||> + * it under the terms of the GNU General Public License=20
||as published by
||> + * the Free Software Foundation; either version 2 of the=20
||License, or
||> + * (at your option) any later version.
||> + *
||> + * This program is distributed in the hope that it will=20
|be useful,
||> + * but WITHOUT ANY WARRANTY; without even the implied=20
|warranty of
||> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
||> + * the GNU General Public License for more details.
||> + *
||> + * You should have received a copy of the GNU General=20
||Public License
||> + * along with this program; if not, write to the Free Software
||> + * Foundation, Inc., 59 Temple Place, Suite 330, Boston,=20
||MA 02111-1307
||> USA
||> + */
||> +
||> +/*
||> + * Test Name: setegid01
||> + *
||> + * Test Description:
||> + * Verify that setegid does not modify the saved gid or real gid.
||> + */
||> +#include <pwd.h>
||> +#include <sys/types.h>
||> +#include <stdio.h>
||> +#include <unistd.h>
||> +#include <errno.h>
||> +#include "test.h"
||> +#include "usctest.h"
||> +
||> +char *TCID =3D "setegid01";
||> +int TST_TOTAL =3D 1; /* is this number right? */
||> +int verbose =3D 0;
||> +
||> +option_t options[] =3D {
||> + { "v", &verbose, NULL }, /* No argument */
||> + { NULL, NULL, NULL } /* NULL required to end array */
||> +};
||> +
||> +void help()
||> +{
||> + printf(" -v verbose\n");
||> +}
||> +
||> +int main(int argc, char **argv)
||> +{
||> + struct passwd nobody;
||> + gid_t nobody_gid;
||> + gid_t cur_rgid, cur_egid, cur_sgid;
||> + gid_t orig_rgid, orig_egid, orig_sgid;
||> + char *msg;
||> +
||> + if ((msg =3D parse_opts(argc, argv, options, help)) !=3D=20
||(char *)NULL) {
||> + tst_brkm(TBROK, NULL, "Option parsing error - %s", msg);
||> + tst_exit();
||> + }
||> +
||> + if (geteuid() !=3D 0) {
||> + tst_brkm(TBROK, NULL, "Must be super/root for this test!");
||> + tst_exit();
||> + }
||> +
||> + nobody =3D *getpwnam("nobody");
||> + nobody_gid =3D nobody.pw_gid;
||> +
||> + TEST(getresgid(&orig_rgid, &orig_egid, &orig_sgid));
||> + if (TEST_RETURN =3D=3D -1) {
||> + tst_resm(TBROK, "getresgid() Failed, errno=3D%d :=20
||%s", TEST_ERRNO,
||> strerror(TEST_ERRNO));
||> + tst_exit();
||> + }
||> + if (verbose) {
||> + printf("getresgid reports rgid %d, egid %d, sgid=20
||%d\n", orig_rgid,
||> + orig_egid, orig_sgid);
||> + printf("calling setegid(nobody_gid %d)\n", nobody_gid);
||> + }
||> + TEST(setegid(nobody_gid));
||> + if (TEST_RETURN =3D=3D -1) {
||> + tst_resm(TFAIL, "setegid() Failed, errno=3D%d : %s",=20
||TEST_ERRNO,
||> strerror(TEST_ERRNO));
||> + tst_exit();
||> + }
||> +
||> + TEST(getresgid(&cur_rgid, &cur_egid, &cur_sgid));
||> + if (TEST_RETURN =3D=3D -1) {
||> + tst_resm(TBROK, "setegid() Failed, errno=3D%d : %s",=20
||TEST_ERRNO,
||> strerror(TEST_ERRNO));
||> + tst_exit();
||> + }
||> + if (verbose) {
||> + printf("getresgid reports rgid %d, egid %d, sgid=20
||%d\n", cur_rgid,
||> + cur_egid, cur_sgid);
||> + }
||> +
||> + /* make sure it at least does what its name says */
||> + if (nobody_gid !=3D cur_egid) {
||> + tst_resm(TFAIL, "setegid() Failed functional test:=20
||it failed to
||> change the effective gid");
||> + tst_exit();
||> + }
||> +
||> + /* SUSv3 says the real group ID and saved set-gid must
||> + * remain unchanged by setgid. See
||> + *
||> http://www.opengroup.org/onlinepubs/007904975/functions/setegid.html
||> + */
||> + if (orig_sgid !=3D cur_sgid) {
||> + tst_resm(TFAIL, "setegid() Failed functional test:=20
||it changed the
||> saved set-gid");
||> + tst_exit();
||> + }
||> + if (orig_rgid !=3D cur_rgid) {
||> + tst_resm(TFAIL, "setegid() Failed functional test:=20
||it changed the
||> real gid");
||> + tst_exit();
||> + }
||> + tst_resm(TPASS, "setegid() passed functional test");
||> + tst_exit();
||> + return 0;
||> +}
||> --- /dev/null 2002-08-30 16:31:37.000000000 -0700
||> +++ ltp-full-20030404/testcases/kernel/syscalls/setegid/Makefile
||> 2003-04-06 20:07:12.000000000 -0700
||> @@ -0,0 +1,36 @@
||> +#
||> +# Copyright (c) International Business Machines Corp., 2001
||> +#
||> +# This program is free software; you can redistribute it=20
||and/or modify
||> +# it under the terms of the GNU General Public License as=20
||published by
||> +# the Free Software Foundation; either version 2 of the=20
|License, or
||> +# (at your option) any later version.
||> +#
||> +# This program is distributed in the hope that it will be useful,
||> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
||> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
||> +# the GNU General Public License for more details.
||> +#
||> +# You should have received a copy of the GNU General=20
|Public License
||> +# along with this program; if not, write to the Free Software
||> +# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA=20
||02111-1307 USA
||> +#
||> +
||>=20
||+##############################################################
||#############
||>=20
||> +# name of file : Makefile
||> #
||>=20
||+##############################################################
||#############
||>=20
||> +CFLAGS+=3D -I../../../../include -Wall -g
||> +LOADLIBES+=3D -L../../../../lib -lltp
||> +
||> +SRCS=3D$(wildcard *.c)
||> +TARGETS=3D$(patsubst %.c,%,$(SRCS))
||> +
||> +all: $(TARGETS)
||> +
||> +install:
||> + @set -e; for i in $(TARGETS); do ln -f $$i=20
||../../../bin/$$i ;
||> done
||> +
||> +clean:
||> + rm -f $(TARGETS)
||> +
||> +
||>=20
||>=20
||>=20
||>=20
||>=20
||
||
||--=20
||Dan Kegel
||http://www.kegel.com
||http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=3D78045=
||
||
||
||-------------------------------------------------------
||This SF.net email is sponsored by: ValueWeb:=20
||Dedicated Hosting for just $79/mo with 500 GB of bandwidth!=20
||No other company gives more support or power for your dedicated server
||http://click.atdmt.com/AFF/go/sdnxxaff00300020aff/direct/01/
||_______________________________________________
||Ltp-list mailing list
||Ltp...@li...
||https://lists.sourceforge.net/lists/listinfo/ltp-list
||
|
|