Skip to contentSkip to conversation

tools/nolibc: add pipe() and its testcase

PATCH Superseded Yuan Tan 4 participants 13 messages Jul 25, 2023 – Jul 30, 2023

Subsystems (2)

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Yuan Tan

author
#0 lore raw

cover tools/nolibc: add pipe() and its testcase

Hi,

Zhangjin and I are working on a tiny shell with nolibc. This patch enables the missing pipe() with its testcase.

Thanks.

Yuan Tan (2):
tools/nolibc: add pipe() support
selftests/nolibc: add testcase for pipe.

 tools/include/nolibc/sys.h                   | 17 ++++++++++
 tools/testing/selftests/nolibc/nolibc-test.c | 34 ++++++++++++++++++++
 2 files changed, 51 insertions(+)
Signature · 1 line
2.39.2
author
#1 lore raw

PATCH 1/2 tools/nolibc: add pipe() support

pipe is crucial for shell.

---
 tools/include/nolibc/sys.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
tools/include/nolibc/sys.h +17 -0
index 8bfe7db20b80..09841fc266fe 100644
@@ -752,6 +752,23 @@ int open(const char *path, int flags, ...)
 }
 
 
+/*
+ * int pipe(int pipefd[2]);
+ */
+
+static __attribute__((unused))
+int sys_pipe(int pipefd[2])
+{
+	return my_syscall1(__NR_pipe, pipefd);
+}
+
+static __attribute__((unused))
+int pipe(int pipefd[2])
+{
+	return __sysret(sys_pipe(pipefd));
+}
+
+
 /*
  * int prctl(int option, unsigned long arg2, unsigned long arg3,
  *                       unsigned long arg4, unsigned long arg5);
Signature · 1 line
2.39.2
#2 lore raw

PATCH 1/2 tools/nolibc: add pipe() support

Hi, Yuan

Quoted 2 lines

pipe is crucial for shell.

As the syscall manpage[1] shows, pipe() is just one of the exceptions how may require two return registers in some platforms, e.g.:
arch/mips/kernel/syscall.c

    * For historic reasons the pipe(2) syscall on MIPS has an unusual calling
     * convention.  It returns results in registers $v0 / $v1 which means there
     * is no need for it to do verify the validity of a userspace pointer
     * argument.  Historically that used to be expensive in Linux.  These days
     * the performance advantage is negligible.
     */
    asmlinkage int sysm_pipe(void)
    {
            int fd[2];
            int error = do_pipe_flags(fd, 0);
            if (error)
                    return error;
            current_pt_regs()->regs[3] = fd[1];
            return fd[0];
    }

The other exceptions are getxpid(2), getxuid(2), and getxgid(2) on Alpha.

Since we are able to use pipe2() for pipe() (introduced from early Linux 2.6.27, glibc 2.9) and use getpid+getppid for getxpid, getuid+geteuid for getxuid and getgid+getegit for getxgid.

So, it is possible provide such pipe() as a wraper of pipe2() and getxpid, getxuid and getxgid as wrappers of their corresponding syscall pairs, then, no need to provide a second return value for all of the existing architectures, for example:

    #ifndef pipe
    int pipe(int pipefd[2])
    {
        pipe2(pipefd, 0);
    }
    #endif

And for mips:

    // may be in tools/include/nolibc/types.h ?
    struct fd_pair {
           long fd[2];
    };

    // tools/include/nolibc/arch-mips.h
    struct fd_pair pipe(void)
    {
            struct fd_pair fds;
            int pipefds[2];
    
            pipe2(pipefds, 0);
    
            fds.fd[0] = pipefds[0];
            fds.fd[1] = pipefds[1];
    
            return fds;
    }

To use such method, the test case should be changed too, perhaps an easier method is even only provide pipe2() for all and let users define their own pipe() if really required, we also need to change the test case.

Best regards,
Zhangjin
[1]: https://man7.org/linux/man-pages/man2/syscall.2.html

Quoted 35 lines
---
 tools/include/nolibc/sys.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
tools/include/nolibc/sys.h +17 -0
index 8bfe7db20b80..09841fc266fe 100644
@@ -752,6 +752,23 @@ int open(const char *path, int flags, ...)
 }
 
 
+/*
+ * int pipe(int pipefd[2]);
+ */
+
+static __attribute__((unused))
+int sys_pipe(int pipefd[2])
+{
+	return my_syscall1(__NR_pipe, pipefd);
+}
+
+static __attribute__((unused))
+int pipe(int pipefd[2])
+{
+	return __sysret(sys_pipe(pipefd));
+}
+
+
 /*
  * int prctl(int option, unsigned long arg2, unsigned long arg3,
  *                       unsigned long arg4, unsigned long arg5);
Signature · 1 line
2.39.2
maintainer
#3 lore raw

Re: [PATCH 1/2] tools/nolibc: add pipe() support

Hi Zhangjin, hi Yuan,

Quoted 9 lines · Zhangjin Wu

Hi, Yuan

Quoted 2 lines

pipe is crucial for shell.

As the syscall manpage[1] shows, pipe() is just one of the exceptions how may require two return registers in some platforms, e.g.:
arch/mips/kernel/syscall.c

Quoted 6 lines

* For historic reasons the pipe(2) syscall on MIPS has an unusual calling
* convention. It returns results in registers $v0 / $v1 which means there
* is no need for it to do verify the validity of a userspace pointer
* argument. Historically that used to be expensive in Linux. These days
* the performance advantage is negligible.
*/

(...)

Ah indeed! I vaguely remembered that I had left that one aside for some time but did not remember why. Now I remember that I couldn't handle the MIPS implementation by then while it used to be my primary target.

Quoted 5 lines

Since we are able to use pipe2() for pipe() (introduced from early Linux 2.6.27, glibc 2.9) and use getpid+getppid for getxpid, getuid+geteuid for getxuid and getgid+getegit for getxgid.

So, it is possible provide such pipe() as a wraper of pipe2() and

Indeed.

Quoted 2 lines

getxpid, getxuid and getxgid as wrappers of their corresponding syscall pairs,

I doubt anyone needs these ones, I didn't know them and do not even have their man page. Let's keep the focus on what developers really use.

Quoted 25 lines

then, no need to provide a second return value for all of the existing architectures, for example:

    #ifndef pipe
    int pipe(int pipefd[2])
    {
        pipe2(pipefd, 0);
    }
    #endif

And for mips:

    // may be in tools/include/nolibc/types.h ?
    struct fd_pair {
           long fd[2];
    };

    // tools/include/nolibc/arch-mips.h
    struct fd_pair pipe(void)
    {
            struct fd_pair fds;
            int pipefds[2];
Quoted 7 lines
            pipe2(pipefds, 0);
    
            fds.fd[0] = pipefds[0];
            fds.fd[1] = pipefds[1];
    
            return fds;
    }

This one does not have the correct prototype for the function exposed to the user, pipe really is "int pipe(int pipefd[2])". Maybe you were thinking about sys_pipe() instead ? But since MIPS also has pipe2() now, there's no reason to make an exception.

Quoted 4 lines

To use such method, the test case should be changed too, perhaps an easier method is even only provide pipe2() for all and let users define their own pipe() if really required, we also need to change the test case.

No, we need to provide users with what they need to compile standard code. If we rely on pipe2() to deliver pipe(), that's fine. We can even do it per-arch if there are constraints but it seems to me that pipe2() is OK.

Thanks,
Willy

#4 lore raw

Re: [PATCH 1/2] tools/nolibc: add pipe() support

Quoted 18 lines

Hi Zhangjin, hi Yuan,

Quoted 9 lines · Zhangjin Wu

Hi, Yuan

Quoted 2 lines

pipe is crucial for shell.

As the syscall manpage[1] shows, pipe() is just one of the exceptions how may require two return registers in some platforms, e.g.:
arch/mips/kernel/syscall.c

Quoted 6 lines

* For historic reasons the pipe(2) syscall on MIPS has an unusual calling
* convention. It returns results in registers $v0 / $v1 which means there
* is no need for it to do verify the validity of a userspace pointer
* argument. Historically that used to be expensive in Linux. These days
* the performance advantage is negligible.
*/

Quoted 6 lines

(...)

Ah indeed! I vaguely remembered that I had left that one aside for some time but did not remember why. Now I remember that I couldn't handle the MIPS implementation by then while it used to be my primary target.

Seems pipe() is the **only** one some architectures (except Alpha) return two values, and now we have pipe2(), so, it is ok for us to simply use pipe2() instead ;-)

Quoted 6 lines
Quoted 5 lines

Since we are able to use pipe2() for pipe() (introduced from early Linux 2.6.27, glibc 2.9) and use getpid+getppid for getxpid, getuid+geteuid for getxuid and getgid+getegit for getxgid.

So, it is possible provide such pipe() as a wraper of pipe2() and

Quoted 8 lines

Indeed.

Quoted 2 lines

getxpid, getxuid and getxgid as wrappers of their corresponding syscall pairs,

I doubt anyone needs these ones, I didn't know them and do not even have their man page. Let's keep the focus on what developers really use.

Yeah.

Quoted 22 lines
Quoted 22 lines

then, no need to provide a second return value for all of the existing architectures, for example:

    #ifndef pipe
    int pipe(int pipefd[2])
    {
        pipe2(pipefd, 0);
    }
    #endif

And for mips:

    // may be in tools/include/nolibc/types.h ?
    struct fd_pair {
           long fd[2];
    };

    // tools/include/nolibc/arch-mips.h
    struct fd_pair pipe(void)
    {

[...]

Quoted 2 lines
Quoted 2 lines
    
            pipe2(pipefds, 0);

[...]

Quoted 7 lines
Quoted 1 line
    }

This one does not have the correct prototype for the function exposed to the user, pipe really is "int pipe(int pipefd[2])". Maybe you were thinking about sys_pipe() instead ? But since MIPS also has pipe2() now, there's no reason to make an exception.

Yes, pipe2() should be a better choice, but I have seen this sentence in syscall manpage [1]:

       /* On Alpha, IA-64, MIPS, SuperH, and SPARC/SPARC64, pipe() has the
          following prototype; see NOTES */

       #include <unistd.h>

       struct fd_pair {
           long fd[2];
       };
       struct fd_pair pipe(void);

If it is about syscall, then we are ok to align all of the architectures together to use "int pipe(int pipefd[2])", otherwise, it will be required to define them in their own arch-<ARCH>.h, just like some defined for arch-s390.h.

[1]: https://www.man7.org/linux/man-pages/man2/pipe.2.html

Quoted 10 lines
Quoted 4 lines

To use such method, the test case should be changed too, perhaps an easier method is even only provide pipe2() for all and let users define their own pipe() if really required, we also need to change the test case.

No, we need to provide users with what they need to compile standard code. If we rely on pipe2() to deliver pipe(), that's fine. We can even do it per-arch if there are constraints but it seems to me that pipe2() is OK.

Ok.

Thanks,
Zhangjin

Quoted 2 lines

Thanks,
Willy

maintainer
#5 lore raw

Re: [PATCH 1/2] tools/nolibc: add pipe() support

Quoted 19 lines · Zhangjin Wu
Quoted 5 lines

This one does not have the correct prototype for the function exposed to the user, pipe really is "int pipe(int pipefd[2])". Maybe you were thinking about sys_pipe() instead ? But since MIPS also has pipe2() now, there's no reason to make an exception.

Yes, pipe2() should be a better choice, but I have seen this sentence in syscall manpage [1]:

       /* On Alpha, IA-64, MIPS, SuperH, and SPARC/SPARC64, pipe() has the
          following prototype; see NOTES */

       #include <unistd.h>

       struct fd_pair {
           long fd[2];
       };
       struct fd_pair pipe(void);
Quoted 2 lines

If it is about syscall, then we are ok to align all of the architectures together to use "int pipe(int pipefd[2])"

Yes it's OK, that's how applications expect it to be used:

https://pubs.opengroup.org/onlinepubs/9699919799/functions/pipe.html

For the archs you mention above, it's the libc that wraps the call, exactly what we ought to do as well (using pipe2() since it will be easier).

Willy

author
#6 lore raw

PATCH 2/2 selftests/nolibc: add testcase for pipe.

In reply to Yuan Tan

Add a testcase of pipe that child process sends message to parent process.

---
 tools/testing/selftests/nolibc/nolibc-test.c | 34 ++++++++++++++++++++
 1 file changed, 34 insertions(+)
tools/testing/selftests/nolibc/nolibc-test.c +34 -0
index 03b1d30f5507..43ba2884fd1e 100644
@@ -767,6 +767,39 @@ int test_mmap_munmap(void)
 	return ret;
 }
 
+int test_pipe(void)
+{
+	int pipefd[2];
+	char buf[32];
+	pid_t pid;
+	char *msg = "hello, nolibc";
+
+	if (pipe(pipefd) == -1)
+		return 1;
+
+	pid = fork();
+
+	switch (pid) {
+	case -1:
+		return 1;
+
+	case 0:
+		close(pipefd[0]);
+		write(pipefd[1], msg, strlen(msg));
+		close(pipefd[1]);
+		exit(EXIT_SUCCESS);
+
+	default:
+		close(pipefd[1]);
+		read(pipefd[0], buf, 32);
+		close(pipefd[0]);
+		wait(NULL);
+
+		if (strcmp(buf, msg))
+			return 1;
+		return 0;
+	}
+}
 
 /* Run syscall tests between IDs <min> and <max>.
  * Return 0 on success, non-zero on failure.
@@ -851,6 +884,7 @@ int run_syscall(int min, int max)
 		CASE_TEST(mmap_munmap_good);  EXPECT_SYSZR(1, test_mmap_munmap()); break;
 		CASE_TEST(open_tty);          EXPECT_SYSNE(1, tmp = open("/dev/null", 0), -1); if (tmp != -1) close(tmp); break;
 		CASE_TEST(open_blah);         EXPECT_SYSER(1, tmp = open("/proc/self/blah", 0), -1, ENOENT); if (tmp != -1) close(tmp); break;
+		CASE_TEST(pipe);              EXPECT_SYSZR(1, test_pipe()); break;
 		CASE_TEST(poll_null);         EXPECT_SYSZR(1, poll(NULL, 0, 0)); break;
 		CASE_TEST(poll_stdout);       EXPECT_SYSNE(1, ({ struct pollfd fds = { 1, POLLOUT, 0}; poll(&fds, 1, 0); }), -1); break;
 		CASE_TEST(poll_fault);        EXPECT_SYSER(1, poll((void *)1, 1, 0), -1, EFAULT); break;
Signature · 1 line
2.39.2
#7 lore raw

Re: [PATCH 2/2] selftests/nolibc: add testcase for pipe.

Quoted 12 lines · Yuan Tan

Add a testcase of pipe that child process sends message to parent process.

---
 tools/testing/selftests/nolibc/nolibc-test.c | 34 ++++++++++++++++++++
 1 file changed, 34 insertions(+)
tools/testing/selftests/nolibc/nolibc-test.c +0 -0
index 03b1d30f5507..43ba2884fd1e 100644

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 775

Quoted 9 lines
tools/testing/selftests/nolibc/nolibc-test.c +6 -0
@@ -767,6 +767,39 @@ int test_mmap_munmap(void)
 	return ret;
 }
 
+int test_pipe(void)
+{
+	int pipefd[2];
+	char buf[32];
+	pid_t pid;
+	char *msg = "hello, nolibc";

const char * const

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 788

Quoted 13 lines
tools/testing/selftests/nolibc/nolibc-test.c +13 -0
+
+	if (pipe(pipefd) == -1)
+		return 1;
+
+	pid = fork();
+
+	switch (pid) {
+	case -1:
+		return 1;
+
+	case 0:
+		close(pipefd[0]);
+		write(pipefd[1], msg, strlen(msg));

Isn't this missing to write trailing the 0 byte?
Also check the return value.

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 789

Quoted 1 line
tools/testing/selftests/nolibc/nolibc-test.c +1 -0
+		close(pipefd[1]);

Do we need to close the pipefds? The process is exiting anyways.

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 794

Quoted 5 lines
tools/testing/selftests/nolibc/nolibc-test.c +5 -0
+		exit(EXIT_SUCCESS);
+
+	default:
+		close(pipefd[1]);
+		read(pipefd[0], buf, 32);

Use sizeof(buf). Check return value == strlen(msg).

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 796

Quoted 2 lines
tools/testing/selftests/nolibc/nolibc-test.c +2 -0
+		close(pipefd[0]);
+		wait(NULL);

waitpid(pid, NULL, 0);

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 800

Quoted 4 lines
tools/testing/selftests/nolibc/nolibc-test.c +4 -0
+
+		if (strcmp(buf, msg))
+			return 1;
+		return 0;

return !!strcmp(buf, msg);

Quoted 16 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	}
+}
 
 /* Run syscall tests between IDs <min> and <max>.
  * Return 0 on success, non-zero on failure.
@@ -851,6 +884,7 @@ int run_syscall(int min, int max)
 		CASE_TEST(mmap_munmap_good);  EXPECT_SYSZR(1, test_mmap_munmap()); break;
 		CASE_TEST(open_tty);          EXPECT_SYSNE(1, tmp = open("/dev/null", 0), -1); if (tmp != -1) close(tmp); break;
 		CASE_TEST(open_blah);         EXPECT_SYSER(1, tmp = open("/proc/self/blah", 0), -1, ENOENT); if (tmp != -1) close(tmp); break;
+		CASE_TEST(pipe);              EXPECT_SYSZR(1, test_pipe()); break;
 		CASE_TEST(poll_null);         EXPECT_SYSZR(1, poll(NULL, 0, 0)); break;
 		CASE_TEST(poll_stdout);       EXPECT_SYSNE(1, ({ struct pollfd fds = { 1, POLLOUT, 0}; poll(&fds, 1, 0); }), -1); break;
 		CASE_TEST(poll_fault);        EXPECT_SYSER(1, poll((void *)1, 1, 0), -1, EFAULT); break;
Signature · 1 line
2.39.2
maintainer
#8 lore raw

Re: [PATCH 2/2] selftests/nolibc: add testcase for pipe.

Quoted 5 lines · Thomas Weißschuh

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 788

Quoted 3 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	case 0:
+		close(pipefd[0]);
+		write(pipefd[1], msg, strlen(msg));

Isn't this missing to write trailing the 0 byte?

It depends if the other side expects to get the trailing 0. In general it's better to avoid sending it since it's only used for internal representation, and the other side must be prepared to receive anything anyway.

Quoted 1 line

Also check the return value.

Indeed!

Quoted 3 lines

Review

Quoted 1 line
(quoted diff) +1 -0
+		close(pipefd[1]);

Do we need to close the pipefds? The process is exiting anyways.

It's better to, because we could imagine looping over the tests for example. Thus each test shoulld have as little impact as possible on other tests.

Quoted 13 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 794

Quoted 5 lines
tools/testing/selftests/nolibc/nolibc-test.c +5 -0
+		exit(EXIT_SUCCESS);
+
+	default:
+		close(pipefd[1]);
+		read(pipefd[0], buf, 32);

Use sizeof(buf). Check return value == strlen(msg).

Review

Quoted 2 lines
(quoted diff) +2 -0
+		close(pipefd[0]);
+		wait(NULL);

waitpid(pid, NULL, 0);

Quoted 6 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 800

Quoted 4 lines
tools/testing/selftests/nolibc/nolibc-test.c +4 -0
+
+		if (strcmp(buf, msg))
+			return 1;
+		return 0;

return !!strcmp(buf, msg);

In fact before that we need to terminate the output buffer. If for any reason the transfer fails (e.g. the syscall fails or transfers data at another location or of another length, we could end up comparing past the end of the buffer. Thus I suggest adding this immediately after the read():

		buf[sizeof(buf) - 1] = 0;

Willy

#9 lore raw

Re: [PATCH 2/2] selftests/nolibc: add testcase for pipe.

Quoted 16 lines · Willy Tarreau
Quoted 5 lines · Thomas Weißschuh

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 788

Quoted 3 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	case 0:
+		close(pipefd[0]);
+		write(pipefd[1], msg, strlen(msg));

Isn't this missing to write trailing the 0 byte?

It depends if the other side expects to get the trailing 0. In general it's better to avoid sending it since it's only used for internal representation, and the other side must be prepared to receive anything anyway.

Quoted 1 line

Also check the return value.

Indeed!

Quoted 7 lines
Quoted 3 lines

Review

Quoted 1 line
(quoted diff) +1 -0
+		close(pipefd[1]);

Do we need to close the pipefds? The process is exiting anyways.

It's better to, because we could imagine looping over the tests for example. Thus each test shoulld have as little impact as possible on other tests.

I meant the newly forked child exiting, not nolibc-test in general. The exit is just below, so the fds in the child are close here anyways.

                |
		|
		v
Quoted 20 lines
Quoted 13 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 794

Quoted 5 lines
tools/testing/selftests/nolibc/nolibc-test.c +5 -0
+		exit(EXIT_SUCCESS);
+
+	default:
+		close(pipefd[1]);
+		read(pipefd[0], buf, 32);

Use sizeof(buf). Check return value == strlen(msg).

Review

Quoted 2 lines
(quoted diff) +2 -0
+		close(pipefd[0]);
+		wait(NULL);

waitpid(pid, NULL, 0);

Quoted 6 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 800

Quoted 4 lines
tools/testing/selftests/nolibc/nolibc-test.c +4 -0
+
+		if (strcmp(buf, msg))
+			return 1;
+		return 0;

return !!strcmp(buf, msg);

Quoted 7 lines

In fact before that we need to terminate the output buffer. If for any reason the transfer fails (e.g. the syscall fails or transfers data at another location or of another length, we could end up comparing past the end of the buffer. Thus I suggest adding this immediately after the read():

		buf[sizeof(buf) - 1] = 0;

This would still access uninitialized memory and lead to UB in strcmp as not all bytes in buf were written to by read().

If we want to be really sure we should use memcmp() instead of strcmp(). For memcmp() I would prefer to transfer and check without the '\0', so my review comments from before need to be adapted a bit.

maintainer
#10 lore raw

Re: [PATCH 2/2] selftests/nolibc: add testcase for pipe.

Quoted 21 lines · Thomas Weißschuh
Quoted 20 lines · Willy Tarreau
Quoted 5 lines · Thomas Weißschuh

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 788

Quoted 3 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	case 0:
+		close(pipefd[0]);
+		write(pipefd[1], msg, strlen(msg));

Isn't this missing to write trailing the 0 byte?

It depends if the other side expects to get the trailing 0. In general it's better to avoid sending it since it's only used for internal representation, and the other side must be prepared to receive anything anyway.

Quoted 1 line

Also check the return value.

Indeed!

Quoted 3 lines

Review

Quoted 1 line
(quoted diff) +1 -0
+		close(pipefd[1]);

Do we need to close the pipefds? The process is exiting anyways.

Quoted 6 lines
Quoted 3 lines

It's better to, because we could imagine looping over the tests for example. Thus each test shoulld have as little impact as possible on other tests.

I meant the newly forked child exiting, not nolibc-test in general. The exit is just below, so the fds in the child are close here anyways.

Ah OK, but still it remains cleaner with it IMHO (i.e. better rely on explicit things in tests, that's less doubts when they fail).

Quoted 24 lines
Quoted 24 lines
Quoted 11 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 794

Quoted 3 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	default:
+		close(pipefd[1]);
+		read(pipefd[0], buf, 32);

Use sizeof(buf). Check return value == strlen(msg).

Review

Quoted 2 lines
(quoted diff) +2 -0
+		close(pipefd[0]);
+		wait(NULL);

waitpid(pid, NULL, 0);

Quoted 6 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 800

Quoted 4 lines
tools/testing/selftests/nolibc/nolibc-test.c +4 -0
+
+		if (strcmp(buf, msg))
+			return 1;
+		return 0;

return !!strcmp(buf, msg);

In fact before that we need to terminate the output buffer. If for any reason the transfer fails (e.g. the syscall fails or transfers data at another location or of another length, we could end up comparing past the end of the buffer. Thus I suggest adding this immediately after the read():

Quoted 8 lines
Quoted 1 line
		buf[sizeof(buf) - 1] = 0;

This would still access uninitialized memory and lead to UB in strcmp as not all bytes in buf were written to by read().

If we want to be really sure we should use memcmp() instead of strcmp(). For memcmp() I would prefer to transfer and check without the '\0', so my review comments from before need to be adapted a bit.

In fact you make a good point regarding the fact that the test doesn't use read()'s return value. This problem totally goes away if the return value is used, e.g.:

      len = read(pipefd[0], buf, sizeof(buf));
      close(pipefd[0]);
      waitpid(pid, NULL, 0);
      return len < 0 || len > sizeof(buf) || len > strlen(msg) || memcmp(buf, msg, len) != 0;

Willy

#11 lore raw

Re: [PATCH 2/2] selftests/nolibc: add testcase for pipe.

Quoted 26 lines · Willy Tarreau
Quoted 25 lines · Thomas Weißschuh
Quoted 23 lines · Willy Tarreau
Quoted 5 lines · Thomas Weißschuh

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 788

Quoted 3 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	case 0:
+		close(pipefd[0]);
+		write(pipefd[1], msg, strlen(msg));

Isn't this missing to write trailing the 0 byte?

It depends if the other side expects to get the trailing 0. In general it's better to avoid sending it since it's only used for internal representation, and the other side must be prepared to receive anything anyway.

Quoted 1 line

Also check the return value.

Indeed!

Quoted 3 lines

Review

Quoted 1 line
(quoted diff) +1 -0
+		close(pipefd[1]);

Do we need to close the pipefds? The process is exiting anyways.

It's better to, because we could imagine looping over the tests for example. Thus each test shoulld have as little impact as possible on other tests.

Quoted 5 lines
Quoted 2 lines

I meant the newly forked child exiting, not nolibc-test in general. The exit is just below, so the fds in the child are close here anyways.

Ah OK, but still it remains cleaner with it IMHO (i.e. better rely on explicit things in tests, that's less doubts when they fail).

Accepted :-)

Quoted 33 lines
Quoted 24 lines
Quoted 24 lines
Quoted 11 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 794

Quoted 3 lines
tools/testing/selftests/nolibc/nolibc-test.c +3 -0
+	default:
+		close(pipefd[1]);
+		read(pipefd[0], buf, 32);

Use sizeof(buf). Check return value == strlen(msg).

Review

Quoted 2 lines
(quoted diff) +2 -0
+		close(pipefd[0]);
+		wait(NULL);

waitpid(pid, NULL, 0);

Quoted 6 lines

tools/testing/selftests/nolibc/nolibc-test.c int test_mmap_munmap(void) · line 800

Quoted 4 lines
tools/testing/selftests/nolibc/nolibc-test.c +4 -0
+
+		if (strcmp(buf, msg))
+			return 1;
+		return 0;

return !!strcmp(buf, msg);

In fact before that we need to terminate the output buffer. If for any reason the transfer fails (e.g. the syscall fails or transfers data at another location or of another length, we could end up comparing past the end of the buffer. Thus I suggest adding this immediately after the read():

Quoted 8 lines
Quoted 1 line
		buf[sizeof(buf) - 1] = 0;

This would still access uninitialized memory and lead to UB in strcmp as not all bytes in buf were written to by read().

If we want to be really sure we should use memcmp() instead of strcmp(). For memcmp() I would prefer to transfer and check without the '\0', so my review comments from before need to be adapted a bit.

Quoted 8 lines

In fact you make a good point regarding the fact that the test doesn't use read()'s return value. This problem totally goes away if the return value is used, e.g.:

      len = read(pipefd[0], buf, sizeof(buf));
      close(pipefd[0]);
      waitpid(pid, NULL, 0);
      return len < 0 || len > sizeof(buf) || len > strlen(msg) || memcmp(buf, msg, len) != 0;

Wouldn't this happily accept len == 0?

Why not just:

if (len != strlen(msg))
return 1;
return !!memcmp(buf, msg, len);

Also so far we have assumed that one call one call to read() is enough. But looking at pipe(7) this is not guaranteed by the spec.
If we want to be really sure, a loop around read() seems to be necessary.

maintainer
#12 lore raw

Re: [PATCH 2/2] selftests/nolibc: add testcase for pipe.

Quoted 9 lines · Thomas Weißschuh
Quoted 8 lines

In fact you make a good point regarding the fact that the test doesn't use read()'s return value. This problem totally goes away if the return value is used, e.g.:

      len = read(pipefd[0], buf, sizeof(buf));
      close(pipefd[0]);
      waitpid(pid, NULL, 0);
      return len < 0 || len > sizeof(buf) || len > strlen(msg) || memcmp(buf, msg, len) != 0;
Quoted 7 lines

Wouldn't this happily accept len == 0?

Why not just:

if (len != strlen(msg))
return 1;
return !!memcmp(buf, msg, len);

Indeed, works for me.

Quoted 3 lines

Also so far we have assumed that one call one call to read() is enough. But looking at pipe(7) this is not guaranteed by the spec.
If we want to be really sure, a loop around read() seems to be necessary.

In practice it will be OK as the message is small and sent in one syscall, so let's not care too much about this for now.

Willy