Hi @all,
if i try to read a character from a sequential file, using
s dev="TestFile.dat" o dev:readonly
u dev r *a
c dev
and the character code 'a' is higher than 127, i got values from
-128 (for $c(128)) up to -1 (for $c(255)).
What is the reason for? Is this my fault or a "bug" in GT.M?
If i write these character (0-255) by
s dev="TestFile.dat" o dev:readonly
f i=0:1:255 u dev w $c(i)
c dev
in this file, the result (written character) is ok,
Does anybody have an idea? Thanx,
Stefan
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Bhaskar,
i say thank you for fixing and your very fast and friendly support on a Sunday(!), but,
sorry, another/new problem or Bug since GT.M 5.2-000A:
compiling/linking a routine with a syntax-error like
if $$os^MINISYSkrn()'="GTM" do
.set $zy=$p(%z,"[",2)-1
will crash GT.M:
%GTM-E-INVSVN, Invalid special variable name
%GTM-F-GTMASSERT, GT.M V5.2-000B Linux x86 - Assert failed /usr/library/V52000B/src/emit_code.c line 1110
Ok, in GT.M it's not allowed to store values in a variable named "$zy" - but this code is not valid for GT.M and other GT.M releases ignore/bypass this syntax error.
Stefan
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for reporting this issue. It is an issue in GT.M that was introduced in V5.2-000A. This will be fixed in the next GT.M release. The tracking # for this issue is C9H04-002856.
Please try out GT.M V5.2-000B and let us know if you encounter any other issues.
Thanks,
Narayanan.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Narayanan,
that is a major problem for my colleagues and me. If we can't compile the internal minisys QM test-suite, we can't test anything. Is it possible to get a SW-patch in advance (Binary x86)?
Regards,
Stefan
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi @all,
if i try to read a character from a sequential file, using
s dev="TestFile.dat" o dev:readonly
u dev r *a
c dev
and the character code 'a' is higher than 127, i got values from
-128 (for $c(128)) up to -1 (for $c(255)).
What is the reason for? Is this my fault or a "bug" in GT.M?
If i write these character (0-255) by
s dev="TestFile.dat" o dev:readonly
f i=0:1:255 u dev w $c(i)
c dev
in this file, the result (written character) is ok,
Does anybody have an idea? Thanx,
Stefan
Sorry, one bug while copy/paste:
...
If i write these character (0-255) by
.s dev="TestFile.dat" o dev:newversion
.f i=0:1:255 u dev w $c(i)
.c dev
in this file, the result (written character) is ok,
...
Stefan
Hi Stefan,
Thank you for reporting this. It is a bug in GT.M introduced in V5.2-000. The following tracking # has been created for this issue.
C9H04-002843 READ * returns incorrect values for ch > 127 and < 256 in SD and SOCKET devices
This will be fixed in the next GT.M release.
Thanks,
Narayanan.
This is fixed in GT.M V5.2-000B, which is available for download from Source Forge. Thanks to Stefan for bringing this regression to our attention.
Regards
-- Bhaskar
Bhaskar,
i say thank you for fixing and your very fast and friendly support on a Sunday(!), but,
sorry, another/new problem or Bug since GT.M 5.2-000A:
compiling/linking a routine with a syntax-error like
if $$os^MINISYSkrn()'="GTM" do
.set $zy=$p(%z,"[",2)-1
will crash GT.M:
%GTM-E-INVSVN, Invalid special variable name
%GTM-F-GTMASSERT, GT.M V5.2-000B Linux x86 - Assert failed /usr/library/V52000B/src/emit_code.c line 1110
Ok, in GT.M it's not allowed to store values in a variable named "$zy" - but this code is not valid for GT.M and other GT.M releases ignore/bypass this syntax error.
Stefan
Stefan,
Thanks for reporting this issue. It is an issue in GT.M that was introduced in V5.2-000A. This will be fixed in the next GT.M release. The tracking # for this issue is C9H04-002856.
Please try out GT.M V5.2-000B and let us know if you encounter any other issues.
Thanks,
Narayanan.
Narayanan,
that is a major problem for my colleagues and me. If we can't compile the internal minisys QM test-suite, we can't test anything. Is it possible to get a SW-patch in advance (Binary x86)?
Regards,
Stefan