This change prevents scattering of ll-related hacks and simplifies pread/pwrite
syscalls parsers' logic a bit.
* util.c (getllval): Add fixup for arg_no for SuperH when argument number
is equal to 3 (see comment).
* io.c (PREAD_OFFSET_ARG): Macro removed.
(SYS_FUNC(pread)): Always use argument
e updates tests
and second one fixes the aforementioned problem.
Eugene Syromyatnikov (2):
Fill old_value argument in timer{,fd}_xettime tests
Fix old_value argument retrieval in timerfd_settime parser
tests/timer_xettime.c |4
tests/timerfd_xettime.c |4
time.c
It is needed in order to differentiate it from the value returned by the
call. As a consequence, it enables revealing possible bugs in syscall
parsers, for example, when the value read on syscall entering and not on
syscall exiting (as it is the case with timerfd_settime parser).
* tests/timer_xet
This is done similar to timer_settime syscall parser.
* time.c (SYS_FUNC(timerfd_settime)): retrieve old_value argument on
exiting and not on entering. Return 0 instead of RVAL_DECODED
(since call hasn't been decoded in full on entering).
---
time.c | 20 +++-
1 file changed
Hello.
This patchset contains some initial proposal for the futex syscall test
along with some syscall parser formatting tweaks that appeared sensible
to me during the test's realisation.
Eugene Syromyatnikov (8):
Fix FUTEX_WAKE_OP compare function mask
futex: Add handling of FUT
* futex.c (SYS_FUNC(futex)) : compare function does not
have FUTEX_OP_OPARG_SHIFT flag and occupies 4 bits from bit 24 to bit
27, as per both initial and current (v4.7) kernel implementations, so
bring futex parser in accordance with this.
---
futex.c |4 +---
1 file changed, 1 insertion
Since FUTEX_FD is known (and most likely reserved for the future use), albeit
obsolete, it has some expected argument format, so it's probably wise to
employ it.
* futex.c (SYS_FUNC(futex)): add handling of FUTEX_FD command.
---
futex.c |1 +
1 file changed, 1 insertion(+)
diff --git a/futex
Several changes have been performed:
* alternate form is used for timeout value since it is pointer half
of the time. And half of the time it is not, which can be confusing
without number base prefix.
* uaddr is pointer across all existing operations, so it's rather expected
that it woul
The support for these has been added in commit g337f130, before Linux 4.5.
* xlat/futexops.in: add FUTEX_WAIT|FUTEX_CLOCK_REALTIME and
FUTEX_WAIT_PRIVATE|FUTEX_CLOCK_REALTIME records.
---
xlat/futexops.in |2 ++
1 file changed, 2 insertions(+)
diff --git a/xlat/futexops.in b/xlat/futexops.
In order to distinguish it from the cases where it is printed in
decimal.
* futex.c (SYS_FUNC(futex)) : use
alternate form for the val3 argument.
---
futex.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/futex.c b/futex.c
index 6bd739f..1418c89 100644
--- a/futex.c
Analogous to timeout argument omitting in FUTEX_WAKE_BITSET command.
* futex.c (SYS_FUNC(futex)) : Remove val2 argument printing.
---
futex.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/futex.c b/futex.c
index 1418c89..a49677a 100644
--- a/futex.c
+++ b/futex.c
@@ -
Analogous to timeout argument omitting in FUTEX_WAKE_BITSET command.
* futex.c (SYS_FUNC(futex)): Remove common val argument print call, add
it into specific command cases (all except FUTEX_LOCK_PI,
FUTEX_UNLOCK_PI, FUTEX_TRYLOCK_PI).
---
futex.c | 14 +-
1 file changed, 13 inse
* tests/futex.c: New file.
* tests/futex.test: Likewise.
* tests/Makefile.am (check_PROGRAMS): Add futex binary to the auxiliary
programs list.
(DECODER_TESTS): Add futex.test to the tests list.
---
tests/Makefile.am |2 +
tests/futex.c | 655 ++
On Mon, Aug 29, 2016 at 5:21 PM, Dmitry V. Levin wrote:
> futex_atomic_op_inuser
Can you please refer to the specific file where it is used for cmp
operation? I've checked include/asm-generic/futex.h and
arch/x86/include/asm/futex.h and haven't found any notion of usage of
this flag for cmp operat
On Mon, Aug 29, 2016 at 5:45 PM, Dmitry V. Levin wrote:
> On Mon, Aug 29, 2016 at 06:12:55PM +0300, Eugene Syromyatnikov wrote:
>> The support for these has been added in commit g337f130, before Linux 4.5.
>
> commit g337f130? Is it a septendecimal number?
Sorry, I've overloo
on used in comments.
* Futex test: instances of direct usage of magic constants replaced
with appropriate macro usage (VAL/VAL_PR)
* Futex test: some typo fixes, etc.
* Other patches haven't been changed.
Eugene Syromyatnikov (7):
Fix FUTEX_WAKE_OP compare function mask
futex: Add handl
* futex.c (SYS_FUNC(futex)) : compare function does not
have FUTEX_OP_OPARG_SHIFT flag and occupies 4 bits from bit 24 to bit
27, as per both initial and current (v4.7) kernel implementations, so
bring futex parser in accordance with this.
---
futex.c |4 +---
1 file changed, 1 insertion
Since FUTEX_FD is known (and most likely reserved for the future use), albeit
obsolete, it has some expected argument format, so it's probably wise to
employ it.
* futex.c (SYS_FUNC(futex)): add handling of FUTEX_FD command.
---
futex.c |1 +
1 file changed, 1 insertion(+)
diff --git a/futex
Several changes have been performed:
* alternate form is used for timeout value since it is pointer half
of the time. And half of the time it is not, which can be confusing
without number base prefix.
* uaddr is pointer across all existing operations, so it's rather expected
that it woul
The support for these has been added in commit 337f130, before Linux 4.5.
* xlat/futexops.in: add FUTEX_WAIT|FUTEX_CLOCK_REALTIME and
FUTEX_WAIT_PRIVATE|FUTEX_CLOCK_REALTIME records.
---
xlat/futexops.in |2 ++
1 file changed, 2 insertions(+)
diff --git a/xlat/futexops.in b/xlat/futexops.i
In order to distinguish it from the cases where it is printed in
decimal.
* futex.c (SYS_FUNC(futex)) : use
alternate form for the val3 argument.
---
futex.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/futex.c b/futex.c
index 6bd739f..1418c89 100644
--- a/futex.c
Analogous to timeout argument omitting in FUTEX_WAKE_BITSET command.
* futex.c (SYS_FUNC(futex)): Remove common val argument print call, add
it into specific command cases (all except FUTEX_LOCK_PI,
FUTEX_UNLOCK_PI, FUTEX_TRYLOCK_PI).
---
futex.c | 14 +-
1 file changed, 13 inse
* tests/futex.c: New file.
* tests/futex.test: Likewise.
* tests/Makefile.am (check_PROGRAMS): Add futex binary to the auxiliary
programs list.
(DECODER_TESTS): Add futex.test to the decoder tests list.
---
tests/Makefile.am |2 +
tests/futex.c | 663 ++
configure script doesn't perform check for linux/futex.h presence, so
HAVE_LINUX_FUTEX_H is always undefined and inclusion of system header is
never performed. Moreover, this header had an incorrect definition of
FUTEX_WAIT_BITSET_PRIVATE/FUTEX_WAKE_BITSET_PRIVATE in the past: from
commit v2.6.24-6
* tests/futex.c: New file.
* tests/futex.test: Likewise.
* tests/Makefile.am (check_PROGRAMS): Add futex.
(DECODER_TESTS): Add futex.test.
---
Changes since v2:
* Commands not available on the old kernels (2.6.18 onwards) are wrapped in
CHECK_FUTEX_ENOSYS which prevents failure in case errno
* xlat/sa_handler_values.in: New file.
* signal.c: Include "xlat/sa_handler_values.h".
(get_sa_handler_str, print_sa_handler): New functions.
(SYS_FUNC(sigsetmask), SYS_FUNC(signal), decode_new_sigaction): Use them.
---
signal.c | 64 -
Hello.
There is some initial implementation of readahead syscall decoder test.
While it is implemented, it has been noticed that incorrect format is
used for printing "count" argument.
Eugene Syromyatnikov (2):
readahead: Fix print format for the "count" argument
test
It is size_t (according to documentation and syscalls.h), so "%lu" should
be used instead of "%ld".
* readahead.c (SYS_FUNC(readahead)): fix conversion specifier for the
"count" argument.
---
readahead.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/readahead.c b/readah
* tests/readahead.c: New file.
* tests/readahead.test: New test.
* tests/.gitignore: Add readahead.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add readahead.test.
---
Simple test for simple decoder.
tests/.gitignore |1 +
tests/Makefile.am|2 ++
tests/reada
Since realpath utility is fairly new in GNU coreutils, there could be
possibility that it is not present in the system. Instead of checking
its presence in configure script it probably makes sense to resort to
calling readlink since it is (probably) more widespread.
---
Makefile.am |3 ++-
1 f
On Thu, Sep 1, 2016 at 9:35 AM, Dmitry V. Levin wrote:
> On Thu, Sep 01, 2016 at 02:47:34AM +0300, Eugene Syromyatnikov wrote:
>> Since realpath utility is fairly new in GNU coreutils, there could be
>> possibility that it is not present in the system. Instead of checking
>
On Thu, Sep 1, 2016 at 9:39 AM, Dmitry V. Levin wrote:
> On Thu, Sep 01, 2016 at 02:43:30AM +0300, Eugene Syromyatnikov wrote:
>> * tests/readahead.c: New file.
>> * tests/readahead.test: New test.
>> * tests/.gitignore: Add readahead.
>> * tests/Makefile.am
pe of
this patch set, since looks like it is fine as is.
Eugene Syromyatnikov (5):
tests/futex: Rename retstr to sprintrc
tests/futex: Increase static sprintrc buffer size
tests/futex: Add support for return codes other than 0 and -1 to
sprintrc
tests: Move return code printing into a sep
* tests/futex.c (sprintrc) : Value increased from 256 to
4095.
---
tests/futex.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/futex.c b/tests/futex.c
index dca191e..edb5c9d 100644
--- a/tests/futex.c
+++ b/tests/futex.c
@@ -145,7 +145,7 @@ void invalid_op(int *val
* tests/futex.c (retstr): Rename to sprintrc,
(main): Update all retstr calls to sprintrc.
---
tests/futex.c | 88 +
1 file changed, 45 insertions(+), 43 deletions(-)
diff --git a/tests/futex.c b/tests/futex.c
index dfdf68d..dca191e 1006
* tests/futex.c (sprintrc): Print the actual return code provided, not
just "0".
---
tests/futex.c | 13 -
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/tests/futex.c b/tests/futex.c
index edb5c9d..fb7833c 100644
--- a/tests/futex.c
+++ b/tests/futex.c
@@ -147,11 +1
* tests/tests.h: Add sprintrc declaration.
* tests/futex.c (sprintrc): Remove.
* tests/sprintrc.c: New file.
* tests/Makefile.am (libtests_a_SOURCES): Add sprintrc.c.
---
tests/Makefile.am |1 +
tests/futex.c | 24
tests/sprintrc.c | 36 +++
* tests/readahead.c: New file.
* tests/readahead.test: New test.
* tests/.gitignore: Add readahead.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add readahead.test.
---
tests/.gitignore |1 +
tests/Makefile.am|2 ++
tests/readahead.c| 62
Hello.
Changes from v1:
* sprintrc now takes long as rc argument since syscalls return long.
* Changes in various rc values support as suggested by Dmitry.
Eugene Syromyatnikov (5):
tests/futex: Rename retstr to sprintrc
tests/futex: Increase static sprintrc buffer size
tests/futex: Add
* tests/futex.c (retstr): Rename to sprintstr,
(main): Convert all retstr calls to sprintrc.
---
tests/futex.c | 88 +
1 file changed, 45 insertions(+), 43 deletions(-)
diff --git a/tests/futex.c b/tests/futex.c
index dfdf68d..fe3741a 10
* tests/futex.c (sprintrc) : Value increased from 256 to
4096.
---
tests/futex.c |3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/tests/futex.c b/tests/futex.c
index fe3741a..28d9df8 100644
--- a/tests/futex.c
+++ b/tests/futex.c
@@ -145,8 +145,7 @@ void invalid_op(int *v
* tests/futex.c (sprintrc): Print the actual return code provided, not
just "0"; checks for snprintf return code added.
---
tests/futex.c | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/tests/futex.c b/tests/futex.c
index 28d9df8..09a6c25 100644
--- a/tests/futex.
* tests/readahead.c: New file.
* tests/readahead.test: New test.
* tests/.gitignore: Add readahead.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add readahead.test.
---
tests/.gitignore |1 +
tests/Makefile.am|2 ++
tests/readahead.c| 62
* tests/tests.h: Add sprintrc declaration.
* tests/futex.c (sprintrc): Remove.
* tests/sprintrc.c: New file.
* tests/Makefile.am (libtests_a_SOURCES): Add sprintrc.c.
---
tests/Makefile.am |1 +
tests/futex.c | 20
tests/sprintrc.c | 33 +++
* tests/readahead.c: New file.
* tests/readahead.test: New test.
* tests/.gitignore: Add readahead.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add readahead.test.
---
Changes since v2:
* Added workaround for an old glibc bug
(https://sourceware.org/bugzilla/show_bug.cgi?
* tests/readahead.c: New file.
* tests/readahead.test: New test.
* tests/.gitignore: Add readahead.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add readahead.test.
---
Changes from v3:
* Scraping incorrect attempt to work around glibc bug, now just skipping
the test.
te
Hello.
This is quite inconsistent collection of patches for various decoder
test in attempt to make them more thorough.
Eugene Syromyatnikov (9):
tests: check decoding of perf_event_open syscall
tests: Use sprintrc for return code output in sched_xetscheduler test
tests: Additional decoder
* tests/perf_event_open.c: New file.
* tests/perf_event_open.test: New test.
* tests/.gitignore: Add perf_event_open.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add perf_event_open.test.
---
tests/.gitignore |1 +
tests/Makefile.am |2 ++
tests/pe
Some additional syscall parser tests:
* Incorrect (negative) PID in sched_getscheduler, sched_setscheduler
* Incorrect address of sched_param structure.
* Incorrect policy value.
* tests/sched_xetscheduler.c: additional tests.
---
tests/sched_xetscheduler.c | 19 +++
1 file c
* tests/sched_rr_get_interval.c: Use sprintrc for return code output.
---
tests/sched_rr_get_interval.c |3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/tests/sched_rr_get_interval.c b/tests/sched_rr_get_interval.c
index 8501346..73e2d39 100644
--- a/tests/sched_rr_get_inte
Some additional syscall parser tests:
* Incorrect timespec pointer.
* Successful call.
* tests/sched_rr_get_interval.c: Additional tests.
---
tests/sched_rr_get_interval.c | 21 +++--
1 file changed, 19 insertions(+), 2 deletions(-)
diff --git a/tests/sched_rr_get_interval.c
struct iocb contains two fields with fd semantics: aio_fildes and
aio_resfd. It is quite reasonable to use appropriate function for
printing them (apart from just "%d").
* aio.c (print_common_flags): Add struct tcb pointer to parameter list;
use printfd for printing aio_resfd field.
(print_ioc
* aio.c: Additional checks added.
---
tests/aio.c | 108 +++
1 file changed, 108 insertions(+)
diff --git a/tests/aio.c b/tests/aio.c
index 8c5bfa7..00de5ae 100644
--- a/tests/aio.c
+++ b/tests/aio.c
@@ -111,6 +111,33 @@ main(void)
* tests/aio.c: Use sprintrc for return code output.
---
tests/aio.c |8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/tests/aio.c b/tests/aio.c
index a6a3b9c..5a023cf 100644
--- a/tests/aio.c
+++ b/tests/aio.c
@@ -147,8 +147,8 @@ main(void)
printf("io_setup(%u,
* tests/sched_xetscheduler.c: Use sprintrc for return code output.
---
tests/sched_xetscheduler.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/sched_xetscheduler.c b/tests/sched_xetscheduler.c
index d3b7759..ffb3307 100644
--- a/tests/sched_xetscheduler.c
+++ b/
* aio.c (main): change output format for min_nr and nr arguments in
io_getevents check to %ld, cast arguments to long.
---
tests/aio.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/aio.c b/tests/aio.c
index 5a023cf..8c5bfa7 100644
--- a/tests/aio.c
+++ b/tests/
Since realpath utility is fairly new in GNU coreutils, there could be
possibility that it is not present in the system. Instead of checking
its presence in configure script it probably makes sense to fall back to
dirname call since it is more widespread.
Initial implementation utilized readlink in
* tests/perf_event_open.c: New file.
* tests/perf_event_open.test: New test.
* tests/.gitignore: Add perf_event_open.
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add perf_event_open.test.
---
Changes since v1:
* HAVE_LINUX_PERF_EVENT_H check added for strange cases where
* aio.c: Additional checks added.
---
Changes since v1:
* All io_submit checks now use long as second argument, last io_submit
check now always passes positive nr value.
* Additional decoder checks reworked a bit in order to increase
coverage.
tests/aio.c | 137 ++
sprintrc_grep function is sprintrc function equivalent suitable for
tests where grep-base pattern matching is employed.
* tests/sprintrc.c (enum sprintrc_fmt): New sprintrc format enumeration.
(sprintrc_ex): New function, renamed from sprintrc with updates
regarding support of different format
---
tests/sched_xetattr.c | 10 --
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/tests/sched_xetattr.c b/tests/sched_xetattr.c
index b437b04..b982cca 100644
--- a/tests/sched_xetattr.c
+++ b/tests/sched_xetattr.c
@@ -54,7 +54,10 @@ main(void)
if (syscall(__NR_sche
tail_alloc with precise size of the structure is used now.
* tests/sched_xetattr.c: Eliminate usage of anonymous union type; rename
sched to sched_attr; change type of sched_attr to struct pointer; use
tail_alloc for sched_attr allocation; update printf statements
accrodingly.
---
tests/sch
Hello.
This time, mostly sched_xetattr tweaks. Another notable changes are new
sprintrc_grep function and additional shmget check.
Eugene Syromyatnikov (7):
tests: Split long lines in sched_xetattr test
tests: Perform more strict structure allocation in sched_xetattr test
tests: Add
* tests/ipc_shm.c: Additional checks for shmget syscall decoding.
---
tests/ipc_shm.c | 20
1 file changed, 20 insertions(+)
diff --git a/tests/ipc_shm.c b/tests/ipc_shm.c
index 54723e2..80a8f0f 100644
--- a/tests/ipc_shm.c
+++ b/tests/ipc_shm.c
@@ -27,11 +27,15 @@
*/
* tests/sched_xetattr.c: Additional checs for sched_getattr and
sched_setattr decoding.
---
tests/sched_xetattr.c | 43 +++
1 file changed, 43 insertions(+)
diff --git a/tests/sched_xetattr.c b/tests/sched_xetattr.c
index 9ff8c72..ce5c417 100644
--- a/t
Since it is how it is declared.
* tests/sched_xetattr.c: Cast size of struct sched_attr to unsigned,
update format specifier accordingly.
---
tests/sched_xetattr.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/sched_xetattr.c b/tests/sched_xetattr.c
index 2e0c
Kernel headers declare this field as s32, and strace prints it with %d
specifier.
* tests/sched_xetattr.c: Change type of sched_nice field of struct
sched_attr to int32_t, update format specifiers accordingly.
---
tests/sched_xetattr.c |6 +++---
1 file changed, 3 insertions(+), 3 deletions
Hello.
I've noticed that IPC tests do not check first parameter (IPC ID) enough
which allowed incorrect format specifier usage. These patches augment
tests with regard to aforementioned issue and correct illicit behaviour.
Eugene Syromyatnikov (2):
tests: Additional IPC checks
Change ty
* tests/ipc_msg.c: Additional msgget (parameter format), msgctl
(parameter format, decoding of struct msqid_ds in IPC_SET/IPC_STAT
commands) checks.
* tests/ipc_sem.c: Additional semget, semctl checks.
* tests/ipc_shm.c: Additional shmctl checks.
* tests/semop.c: Additional semop checks. Added
* ipc_msgctl.c (SYS_FUNC(msgctl)): Convert format specifier of the first
argument from "%lu" to "%d", cast argument to int.
* ipc_sem.c (SYS_FUNC(semop)): Likewise.
(SYS_FUNC(semtimedop)): Likewise.
(SYS_FUNC(semget)): Likewise.
* ipc_shm.c (SYS_FUNC(shmat)): Likewise.
* ipc_shmctl.c (SYS_FUN
* ipc_msgctl.c (SYS_FUNC(msgctl)): Convert format specifier of the first
argument from "%lu" to "%d", cast argument to int.
* ipc_sem.c (SYS_FUNC(semop)): Likewise.
(SYS_FUNC(semtimedop)): Likewise.
(SYS_FUNC(semget)): Likewise.
* ipc_shm.c (SYS_FUNC(shmat)): Likewise.
* ipc_shmctl.c (SYS_FUN
key_t is actually int.
* ipc_msg.c (SYS_FUNC(msgget)): Change format specifier of the first
argument to "%#x".
* ipc_sem.c (SYS_FUNC(semget)): Likewise.
* ipc_shm (SYS_FUNC(shmget)): Likewise.
---
ipc_msg.c |5 +++--
ipc_sem.c |5 +++--
ipc_shm.c |5 +++--
3 files changed, 9 inserti
f decoding
checks.
* Invalid addresses are now made more invalid in order to avoid accidental
decoding.
Eugene Syromyatnikov (9):
Change type of the format specifier from "%lu" to "%d" for IPC ID
Fix key_t argument printing in IPC *get calls
tests: Split long lines in
sprintrc_grep function is sprintrc function equivalent suitable for
tests where grep-base pattern matching is employed.
* tests/sprintrc.c (enum sprintrc_fmt): New sprintrc format enumeration.
(sprintrc_ex): New function, renamed from sprintrc with updates
regarding support of different format
Since it is how it is declared.
* tests/sched_xetattr.c: Cast size of struct sched_attr to unsigned,
update format specifier accordingly.
---
tests/sched_xetattr.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/sched_xetattr.c b/tests/sched_xetattr.c
index 2e0c
tail_alloc with precise size of the structure is used now.
* tests/sched_xetattr.c: Eliminate usage of anonymous union type; rename
sched to sched_attr; change type of sched_attr to struct pointer; use
tail_alloc for sched_attr allocation; update printf statements
accrodingly.
---
tests/sch
---
tests/sched_xetattr.c | 10 --
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/tests/sched_xetattr.c b/tests/sched_xetattr.c
index b437b04..b982cca 100644
--- a/tests/sched_xetattr.c
+++ b/tests/sched_xetattr.c
@@ -54,7 +54,10 @@ main(void)
if (syscall(__NR_sche
Kernel headers declare this field as s32, and strace prints it with %d
specifier.
* tests/sched_xetattr.c: Change type of sched_nice field of struct
sched_attr to int32_t, update format specifiers accordingly.
---
tests/sched_xetattr.c |6 +++---
1 file changed, 3 insertions(+), 3 deletions
* tests/sched_xetattr.c: Additional checs for sched_getattr and
sched_setattr decoding.
---
tests/sched_xetattr.c | 43 +++
1 file changed, 43 insertions(+)
diff --git a/tests/sched_xetattr.c b/tests/sched_xetattr.c
index 9ff8c72..ce5c417 100644
--- a/t
* tests/ipc_msg.c: Additional msgget (parameter format), msgctl
(parameter format, decoding of struct msqid_ds in IPC_SET/IPC_STAT
commands) checks.
* tests/ipc_sem.c: Additional semget, semctl checks.
* tests/ipc_shm.c: Additional shmget, shmctl checks.
* tests/semop.c: Additional semop checks
Changes since v2:
* Workaround for musl which sets size argument in shmget call to SIZE_MAX
if it is greater than PTRDIFF_MAX.
* Fix pattern for indirect semctl call.
* Added workaround for broken msgctl call with faulty pointer in third
argument on PPC64.
Eugene Syromyatnikov (2
* tests/ipc_msg.c: Additional msgget (parameter format), msgctl
(parameter format, decoding of struct msqid_ds in IPC_SET/IPC_STAT
commands) checks.
* tests/ipc_sem.c: Additional semget, semctl checks.
* tests/ipc_shm.c: Additional shmget, shmctl checks.
* tests/semop.c: Additional semop checks
glibc before 2.23 (glibc-2.22-122-gbe48165? to be more precise) tried to access
data pointed by third argument of msgctl call on 64-bit PowerPC which led to
segmentataion fault (instead of EFAULT from kernel).
* tests/ipc_msg.c (BROKEN_GLIBC): new macro.
(main): test for incorrect third argument
On Tue, Sep 13, 2016 at 7:32 AM, Andreas Schwab wrote:
> On Sep 13 2016, Eugene Syromyatnikov
> wrote:
>
>> glibc before 2.23 (glibc-2.22-122-gbe48165? to be more precise) tried to
>> access
>> data pointed by third argument of msgctl call on 64-bit PowerPC which le
* tests/ipc_msg.c: Additional msgget (parameter format), msgctl
(parameter format, decoding of struct msqid_ds in IPC_SET/IPC_STAT
commands) checks.
* tests/ipc_sem.c: Additional semget, semctl checks.
* tests/ipc_shm.c: Additional shmget, shmctl checks.
* tests/semop.c: Additional semop checks
exiting and it isn't possible to reliably enforce syscall to
return success (file system with quota enabled and CAP_SYS_ADMIN are among
requirements for this to work).
Eugene Syromyatnikov (12):
quota: Display quota command as a macro
quota: Use printuid for id parameter printing
quota:
* quota.c (SYS_FUNC(quotactl)): Replace disjunction of two xlat values
(which also lacks shift of left part) with a macro call.
---
quota.c |5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/quota.c b/quota.c
index 0a388b2..9f18c0d 100644
--- a/quota.c
+++ b/quota.c
@@ -
---
quota.c | 638 +++
1 file changed, 319 insertions(+), 319 deletions(-)
diff --git a/quota.c b/quota.c
index 495bd5f..68f8a79 100644
--- a/quota.c
+++ b/quota.c
@@ -215,380 +215,380 @@ static int
decode_cmd_data(struct tcb *tcp, uin
* quota.c (SYS_FUNC(quotactl)): since id call argument is used for
user/group/project ID which all have special semantics for -1, it
looks reasonable to print it the same way as UIDs are printed.
---
quota.c |3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/quota.c b/quo
id argument is not used for Q_QUOTAOFF, Q_GETFMT, Q_V2_GETINFO,
Q_GETINFO, Q_V2_SEETINFO, Q_SETINFO, Q_SYNC, Q_XQUOTAON, Q_XQUOTAOFF,
Q_XQUOTARM, Q_XGETQSTAT, Q_XGETQSTATV, Q_XQUOTASYNC, Q_V1_GETSTATS,
Q_V2_GETSTATS subcommands, so it makes sense to avoid printing it.
* quota.c (decode_cmd_data):
Q_XQUOTASYNC is no-op, but it does not require id/addr arguments anyway.
* quota.c (decode_cmd_data): Add Q_SYNC and Q_XQUOTASYNC to subcommand
dispatch switch since id/addr arguments for these subcommands are
known to be ignored.
---
quota.c |2 ++
1 file changed, 2 insertions(+)
diff -
* quota.c (decode_cmd_data): Print field names in accordance with their
definition.
---
quota.c | 191 ---
1 file changed, 98 insertions(+), 93 deletions(-)
diff --git a/quota.c b/quota.c
index 68f8a79..1e1597f 100644
--- a/quota.c
++
Support for these commands has been dropped long ago (before 2.6.12 and
around 2.4.22 as stated by related man page), so it is decided to drop
it in order to minimize amount of code which should be subjected to
testing and structured conversion.
* quota.c: Drop support for decoding of obsolete Q_V
Since the only difference between 32-bit and 64-bit environments
regarding this structure is its padding, it has been decided to just add
packed attribute to it instead of going full mpers.
* quota.c: Add ATTRIBUTE_PACKED to struct if_dqblk definition.
---
quota.c |6 +-
1 file changed, 5
* quota.c (decode_cmd_data): Add Q_XQUOTARM handling to subcommand
switch.
---
quota.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/quota.c b/quota.c
index 53730d1..905fece 100644
--- a/quota.c
+++ b/quota.c
@@ -591,6 +591,19 @@ decode_cmd_data(struct tcb *tcp, uint32_t i
* tests/init.sh (run_strace_match_diff): Add support for providing
arguments to program via the PROG_ARGS environment variable.
---
tests/init.sh |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/tests/init.sh b/tests/init.sh
index 46a08f9..e37fb20 100644
--- a/tests/init
* quota.c (decode_cmd_data): Add printing of d_rtb_hardlimit and
d_rtb_softlimit fields to struct xfs_dqblk printing code.
---
quota.c |4
1 file changed, 4 insertions(+)
diff --git a/quota.c b/quota.c
index bc55638..24370d4 100644
--- a/quota.c
+++ b/quota.c
@@ -269,6 +269,10 @@ decod
* configure.ac (AC_CHECK_HEADERS): Add linux/dqblk_xfs.h, sys/quota.h.
* tests/.gitignore: Add quotactl, quotactl-xfs
* tests/Makefile.am (check_PROGRAMS): Likewise.
(DECODER_TESTS): Add quotactl.test, quotactl-v.test,
quotactl-xfs.test, quotactl-xfs-v.test.
(EXTRA_DIST): quotactl.h
* quotact
* tests/ioctl.block (init_magic): Add iterator value to fill magic value
in order to enable detection of possible 4-byte aligned shifts.
---
tests/ioctl_block.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/ioctl_block.c b/tests/ioctl_block.c
index 3dc030d..bac9f68
* tests/getcwd.c: Additional checks for getcwd arguments decoding.
---
tests/getcwd.c |8
1 file changed, 8 insertions(+)
diff --git a/tests/getcwd.c b/tests/getcwd.c
index 704a315..e4a791f 100644
--- a/tests/getcwd.c
+++ b/tests/getcwd.c
@@ -13,6 +13,14 @@ main(void)
{
long
1 - 100 of 310 matches
Mail list logo
