Skip to contentSkip to conversation

ltp: fix warnings in "fsstress.c"

PATCH Superseded Alex Elder 3 participants 3 messages Apr 11, 2011 – Apr 13, 2011

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Alex Elder

author
#0 lore raw

PATCH v2 3/3 ltp: fix warnings in "fsstress.c"

Fix remaining warnings in "testcases/kernel/fs/fsstress/fsstress.c". Three of them are due to mismatches between fprintf() conversion specifications and the variables being formatted. Fix these by casting the variables to the right type.

Two others are due to the use of a __uint64_t local variable to represent a value of type __u64 (which evidently can be defined differently in some environments). Fix these by using the exact matching type in the cases that produce the warning messages.

Updated to address the concern raised by Garrett Cooper about the use of types that are not identical to what's used in the XFS ioctl() interface in the kernel. Also added memset() calls so we're sure no garbage resides in data structures passed between user space and the kernel.

---
 ltp/fsstress.c |   83
+++++++++++++++++++++++++++++++++------------------------
 1 file changed, 49 insertions(+), 34 deletions(-)
ltp/fsstress.c +1 -0
Index: b/ltp/fsstress.c
===================================================================
@@ -419,6 +419,7 @@ int main(int argc, char **argv)
                }
 #ifndef NO_XFS
                if (!no_xfs) {
+               memset(&geom, 0, sizeof geom);
                i = ioctl(fd, XFS_IOC_FSGEOMETRY, &geom);
                if (i >= 0 && geom.rtblocks)
                        rtpct = MIN(MAX(geom.rtblocks * 100 /
@@ -447,6 +448,7 @@ int main(int argc, char **argv)
                                errtag += (random() % (XFS_ERRTAG_MAX -

errtag));
}
printf("Injecting failure on tag #%d\n",
errtag);

+                       memset(&err_inj, 0, sizeof err_inj);
                        err_inj.errtag = errtag;
                        err_inj.fd = fd;
                        srval = ioctl(fd, XFS_IOC_ERROR_INJECTION,
&err_inj);
@@ -476,6 +478,7 @@ int main(int argc, char **argv)
                }
 #ifndef NO_XFS
                if (errtag != 0) {
+                       memset(&err_inj, 0, sizeof err_inj);
                        err_inj.errtag = 0;
                        err_inj.fd = fd;
                        if ((srval = ioctl(fd, XFS_IOC_ERROR_CLEARALL,
&err_inj)) != 0) {
@@ -1396,9 +1399,9 @@ allocsp_f(int opno, long r)
        int             e;
        pathname_t      f;
        int             fd;
-       struct flock64  fl;
-       __int64_t       lr;
-       off64_t         off;
+       struct xfs_flock64 fl;
+       __s64           lr;
+       __s64           off;
        struct stat64   stb;
        int             v;

@@ -1427,9 +1430,10 @@ allocsp_f(int opno, long r)

                close(fd);
                return;
        }

- lr = ((__int64_t)random() << 32) + random();
- off = (off64_t)(lr % MIN(stb.st_size + (1024 * 1024),
MAXFSIZE));
+ lr = ((__s64) random() << 32) + random();

+       off = lr % MIN(stb.st_size + (1024 * 1024), MAXFSIZE);
        off %= maxfsize;
+       memset(&fl, 0, sizeof fl);
        fl.l_whence = SEEK_SET;
        fl.l_start = off;
        fl.l_len = 0;

@@ -1547,10 +1551,10 @@ attr_set_f(int opno, long r)
void
bulkstat_f(int opno, long r)
{

-       int             count;
+       __s32           count;
        int             fd;
-       __uint64_t      last;
-       int             nent;
+       __u64           last;
+       __s32           nent;
        xfs_bstat_t     *t;
        __int64_t       total;
        xfs_fsop_bulkreq_t bsr;
@@ -1561,6 +1565,7 @@ bulkstat_f(int opno, long r)
        fd = open(".", O_RDONLY);
        total = 0;
+       memset(&bsr, 0, sizeof bsr);
        bsr.lastip=&last;
        bsr.icount=nent;
        bsr.ubuffer=t;

@@ -1571,7 +1576,7 @@ bulkstat_f(int opno, long r)
free(t);
if (verbose)
printf("%d/%d: bulkstat nent %d total %lld\n",
- procid, opno, nent, (long long)total);
+ procid, opno, (int) nent, (long long) total);

        close(fd);

}

@@ -1582,7 +1587,7 @@ bulkstat1_f(int opno, long r)

        pathname_t      f;
        int             fd;
        int             good;
-       __uint64_t      ino;
+       __u64           ino;
        struct stat64   s;
        xfs_bstat_t     t;
        int             v;
@@ -1617,6 +1622,7 @@ bulkstat1_f(int opno, long r)
        }
        fd = open(".", O_RDONLY);
+       memset(&bsr, 0, sizeof bsr);
        bsr.lastip=&ino;
        bsr.icount=1;
        bsr.ubuffer=&t;
@@ -1625,7 +1631,8 @@ bulkstat1_f(int opno, long r)
        e = ioctl(fd, XFS_IOC_FSBULKSTAT_SINGLE, &bsr) < 0 ? errno : 0;
        if (v)
                printf("%d/%d: bulkstat1 %s ino %lld %d\n",
-                       procid, opno, good?"real":"random",
(int64_t)ino, e);
+                       procid, opno, good ? "real" : "random",
+                       (long long) ino, e);
        close(fd);

}

@@ -1698,6 +1705,7 @@ creat_f(int opno, long r)

        if (fd >= 0) {
 #ifndef NO_XFS
                struct fsxattr  a;
+               memset(&a, 0, sizeof a);
                if (extsize && ioctl(fd, XFS_IOC_FSGETXATTR, &a) >= 0) {
                        a.fsx_xflags |= XFS_XFLAG_REALTIME;
                        a.fsx_extsize =
@@ -1799,6 +1807,7 @@ dread_f(int opno, long r)
                return;
        }
+       memset(&diob, 0, sizeof diob);
        if (no_xfs) {
                diob.d_miniosz = stb.st_blksize;
                diob.d_maxiosz = stb.st_blksize * 256;  /* good number ?
*/
@@ -1889,6 +1898,7 @@ dwrite_f(int opno, long r)
                close(fd);
                return;
        }
+       memset(&diob, 0, sizeof diob);
        if (no_xfs) {
                diob.d_miniosz = stb.st_blksize;
                diob.d_maxiosz = stb.st_blksize * 256;  /* good number ?
*/
@@ -1976,9 +1986,9 @@ freesp_f(int opno, long r)
        int             e;
        pathname_t      f;
        int             fd;
-       struct flock64  fl;
-       __int64_t       lr;
-       off64_t         off;
+       struct xfs_flock64 fl;
+       __s64           lr;
+       __s64           off;
        struct stat64   stb;
        int             v;

@@ -2007,16 +2017,17 @@ freesp_f(int opno, long r)

                close(fd);
                return;
        }

- lr = ((__int64_t)random() << 32) + random();
- off = (off64_t)(lr % MIN(stb.st_size + (1024 * 1024),
MAXFSIZE));
+ lr = ((__s64) random() << 32) + random();

+       off = lr % MIN(stb.st_size + (1024 * 1024), MAXFSIZE);
        off %= maxfsize;
+       memset(&fl, 0, sizeof fl);
        fl.l_whence = SEEK_SET;
        fl.l_start = off;
        fl.l_len = 0;
        e = ioctl(fd, XFS_IOC_FREESP64, &fl) < 0 ? errno : 0;
        if (v)
                printf("%d/%d: ioctl(XFS_IOC_FREESP64) %s %lld 0 %d\n",
-                       procid, opno, f.path, (long long)off, e);
+                       procid, opno, f.path, (long long) off, e);
        free_pathname(&f);
        close(fd);
 }
@@ -2349,9 +2360,9 @@ resvsp_f(int opno, long r)
        int             e;
        pathname_t      f;
        int             fd;
-       struct flock64  fl;
-       __int64_t       lr;
-       off64_t         off;
+       struct xfs_flock64 fl;
+       __s64           lr;
+       __s64           off;
        struct stat64   stb;
        int             v;

@@ -2380,16 +2391,18 @@ resvsp_f(int opno, long r)

                close(fd);
                return;
        }

- lr = ((__int64_t)random() << 32) + random();
- off = (off64_t)(lr % MIN(stb.st_size + (1024 * 1024),
MAXFSIZE));
+ lr = ((__s64) random() << 32) + random();

+       off = lr % MIN(stb.st_size + (1024 * 1024), MAXFSIZE);
        off %= maxfsize;
+       memset(&fl, 0, sizeof fl);
        fl.l_whence = SEEK_SET;
        fl.l_start = off;
-       fl.l_len = (off64_t)(random() % (1024 * 1024));
+       fl.l_len = (__s64) (random() % (1024 * 1024));
        e = ioctl(fd, XFS_IOC_RESVSP64, &fl) < 0 ? errno : 0;
        if (v)

- printf("%d/%d: ioctl(XFS_IOC_RESVSP64) %s %lld %ld %d
\n",
- procid, opno, f.path, (long long)off, fl.l_len,
e);
+ printf("%d/%d: ioctl(XFS_IOC_RESVSP64) %s %lld %lld %d
\n",
+ procid, opno, f.path, (long long) off,

+                       (long long) fl.l_len, e);
        free_pathname(&f);
        close(fd);
 }
@@ -2524,7 +2537,7 @@ truncate_f(int opno, long r)
                return;
        }
        lr = ((__int64_t)random() << 32) + random();
-       off = (off64_t)(lr % MIN(stb.st_size + (1024 * 1024),
MAXFSIZE));
+       off = lr % MIN(stb.st_size + (1024 * 1024), MAXFSIZE);
        off %= maxfsize;
        e = truncate64_path(&f, off) < 0 ? errno : 0;
        check_cwd();
@@ -2566,9 +2579,9 @@ unresvsp_f(int opno, long r)
        int             e;
        pathname_t      f;
        int             fd;
-       struct flock64  fl;
-       __int64_t       lr;
-       off64_t         off;
+       struct xfs_flock64 fl;
+       __s64           lr;
+       __s64           off;
        struct stat64   stb;
        int             v;

@@ -2597,16 +2610,18 @@ unresvsp_f(int opno, long r)

                close(fd);
                return;
        }

- lr = ((__int64_t)random() << 32) + random();
- off = (off64_t)(lr % MIN(stb.st_size + (1024 * 1024),
MAXFSIZE));
+ lr = ((__s64) random() << 32) + random();

+       off = lr % MIN(stb.st_size + (1024 * 1024), MAXFSIZE);
        off %= maxfsize;
+       memset(&fl, 0, sizeof fl);
        fl.l_whence = SEEK_SET;
        fl.l_start = off;
-       fl.l_len = (off64_t)(random() % (1 << 20));
+       fl.l_len = (__s64) (random() % (1 << 20));
        e = ioctl(fd, XFS_IOC_UNRESVSP64, &fl) < 0 ? errno : 0;
        if (v)

- printf("%d/%d: ioctl(XFS_IOC_UNRESVSP64) %s %lld %ld %d
\n",
- procid, opno, f.path, (long long)off, fl.l_len,
e);
+ printf("%d/%d: ioctl(XFS_IOC_UNRESVSP64) %s %lld %lld %d
\n",
+ procid, opno, f.path, (long long) off,

+                       (long long) fl.l_len, e);
        free_pathname(&f);
        close(fd);

}

------------------------------------------------------------------------------
Forrester Wave Report
Recovery time is now measured in hours and minutes not days. key insights are discussed in the 2010 Forrester Wave Report as part of an in-depth evaluation of disaster recovery service providers.
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

#1 lore raw
Quoted 10 lines · Alex Elder

Fix remaining warnings in "testcases/kernel/fs/fsstress/fsstress.c". Three of them are due to mismatches between fprintf() conversion specifications and the variables being formatted.  Fix these by casting the variables to the right type.

Two others are due to the use of a __uint64_t local variable to represent a value of type __u64 (which evidently can be defined differently in some environments).  Fix these by using the exact matching type in the cases that produce the warning messages.

Quoted 5 lines

Updated to address the concern raised by Garrett Cooper about the use of types that are not identical to what's used in the XFS ioctl() interface in the kernel.  Also added memset() calls so we're sure no garbage resides in data structures passed between user space and the kernel.

1. Please use sizeof() instead of bareword sizeof.
2. I'm not sure if this is necessarily LKML style (the code I've sampled in the mainline kernel do this however), but casts don't require spurious spaces in FreeBSD for instance, i.e.

    foo = (void*)bar;

instead of

    foo = (void*) bar;

Cheers,
-Garrett

------------------------------------------------------------------------------
Forrester Wave Report - Recovery time is now measured in hours and minutes not days. Key insights are discussed in the 2010 Forrester Wave Report as part of an in-depth evaluation of disaster recovery service providers. Forrester found the best-in-class provider in terms of services and vision. Read this report now! http://p.sf.net/sfu/ibm-webcastpromo
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

#2 lore raw

Hi!

Quoted 10 lines

1. Please use sizeof() instead of bareword sizeof.
2. I'm not sure if this is necessarily LKML style (the code I've sampled in the mainline kernel do this however), but casts don't require spurious spaces in FreeBSD for instance, i.e.

    foo = (void*)bar;

instead of

    foo = (void*) bar;

Agreed, while it's not written explicitly it seems to be common sense.

Signature · 13 lines
Cyril Hrubis
[email protected]

------------------------------------------------------------------------------
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list