Bug 55258

Summary:	od --width seems to assume sizeof(double)==12
Product:	[Retired] Red Hat Linux	Reporter:	Robert Dodier <dodier>
Component:	textutils	Assignee:	Bernhard Rosenkraenzer <bero>
Status:	CLOSED RAWHIDE	QA Contact:	Ben Levenson <benl>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	7.1
Target Milestone:	---
Target Release:	---
Hardware:	i686
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2001-10-29 01:08:12 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Robert Dodier 2001-10-28 15:46:21 UTC

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux 2.4.2-2 i686; en-US; 0.7) Gecko/20010316

Description of problem:
od --width seems to be confused about the size of a double float. With
--width a multiple of 8, od prints a warning message and sets --width to be
a multiple of 12 instead. However, od seems to be correctly reading 8 bytes
for each double, so if --width is a multiple of 8 and 12 (e.g., 24) then
the output is correct.

Also, od is confused about sizeof(float) -- od seems to think sizeof(float)
== 8 for purposes of --width.

Version-Release number of selected component (if applicable):


How reproducible:
Always

Steps to Reproduce:
1. Create data file t.data containing 6 doubles, 1, 2, 3, 4, 5, 6. (See
additional info.)

2. Run "od -t f8 --width=8 t.data".

	

Actual Results:  [robert@chromium /tmp]$ od -t f8 --width=8 t.data
od: warning: invalid width 8; using 12 instead
0000000   1.000000000000000e+00
0000014  5.304989477413181e-315
0000030   4.000000000000000e+00
0000044  5.311465294646351e-315
0000060


Expected Results:  Output should be:
0000000   1.000000000000000e+00
0000010   2.000000000000000e+00
0000020   3.000000000000000e+00
0000030   4.000000000000000e+00
0000040   5.000000000000000e+00
0000050   6.000000000000000e+00


Additional info:

1. "od -t f8 --width=24 t.data" yields correct output, i.e., 3 doubles per
line, 2 lines.

2. "od -t f8 --width=12 t.data" yields same incorrect output as "Actual
Results" above, and there is no warning message (as there should be since
width is not a multiple of sizeof(double)).

3. Bytes in t.data are the following:
$ od -b t.data
0000000 000 000 000 000 000 000 360 077 000 000 000 000 000 000 000 100
0000020 000 000 000 000 000 000 010 100 000 000 000 000 000 000 020 100
0000040 000 000 000 000 000 000 024 100 000 000 000 000 000 000 030 100
0000060

4. "od -t f4 --width=4 t.data" yields "od: warning: invalid width 4; using
8 instead". (Output is goofy since t.data contains doubles instead of
singles, but that's beside the point.)

5. I am using od built from textutils-2.0.11-7.src.rpm, according to rpm.

Comment 1 Robert Dodier 2001-10-29 01:08:06 UTC

OK, I think I found it. There are 2 parallel lists containing type info, 
size_spec (a list of enums), and width_bytes (a list of sizeofs). width_bytes 
was missing sizeof(long long), so the placement of the float, double, and long 
double items was off by 1. The patch shown below fixes the --width problem.

BTW I am working with od.c from textutils-2.0.11-7.src.rpm.

[root@chromium src]# diff -c od.original od.c
*** od.original Sun Oct 28 17:30:11 2001
--- od.c        Sun Oct 28 17:46:27 2001
***************
*** 135,140 ****
--- 135,141 ----
    sizeof (short int),
    sizeof (int),
    sizeof (long int),
+   sizeof (long long),
    sizeof (float),
    sizeof (double),
    sizeof (LONG_DOUBLE)

Comment 2 Bernhard Rosenkraenzer 2001-10-30 12:44:27 UTC

You're right, fixed in 2.0.16-1.