Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Michael Elizabeth Chastain]

Hi Orjan,

I tested this on my system (native i686-pc-linux-gnu, red hat
linux 7.1).  I tried it before and after with both expect 5.30
and /usr/bin/expect (which is expect 5.31).

The bug is sensitive to the version of "expect" and "dejagnu" that
I use when I run the tests.  The fix works in all cases that
I tried.

(Please check your mailer: tabs got expanded and long lines showed up as
wrapped by the time I got the patch.  I edited the patch, no problem.)

Michael Elizabeth Chastain
<mec@shout.net>
"love without fear"

Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Orjan Friberg]

Michael Elizabeth Chastain wrote:
> 
> Hi Orjan,
> 
> I tested this on my system (native i686-pc-linux-gnu, red hat
> linux 7.1).  I tried it before and after with both expect 5.30
> and /usr/bin/expect (which is expect 5.31).
> 
> The bug is sensitive to the version of "expect" and "dejagnu" that
> I use when I run the tests.  The fix works in all cases that
> I tried.

Michael,

Thanks for testing it.  (I'll await Fernando's approval since I
submitted it as an RFA, although I'm thinking now it could have gone
under the obvious bug fix rule.)

> (Please check your mailer: tabs got expanded and long lines showed up as
> wrapped by the time I got the patch.  I edited the patch, no problem.)

Thanks for pointing that out.  I was wrapping outgoing messages at 72
characters, which obviously doesn't work well with patches.

-- 
Orjan Friberg              E-mail: orjan.friberg@axis.com
Axis Communications AB     Phone:  +46 46 272 17 68

Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Michael Elizabeth Chastain]

I see the patch on the mainline, but not on gdb_5_1_2001-07-29-branch.

What is the policy for the 5.1 branch these days?

In the fullness of time I could setup my tests to do both mainline
and branch.  But right now, I can handle just one configuration, and I
chose 5.1 branch.  I can switch to mainline if that's more useful.

Michael Elizabeth Chastain
<mec@shout.net>
"love without fear"

[RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Orjan Friberg]

Michael Elizabeth Chastain wrote:
> 
> gdb.base/setvar.exp:
> 
>   This script produces 2 ERRORs on native i686-pc-linux-gnu (red hat linux 7.1):
> 
>     set variable *(v_signed_short_pointer+1)=-456
>     (gdb) PASS: gdb.base/setvar.exp: set variable *(v_signed_short_pointer+1)=-456
>     ERROR: Process no longer exists
>     UNRESOLVED: gdb.base/setvar.exp: set variable signed short pointer
>     print v_signed_short_array
>     $83 = {123, -456}
> 
>     set variable *(v_unsigned_short_pointer+1)=-456
>     (gdb) PASS: gdb.base/setvar.exp: set variable *(v_unsigned_short_pointer+1)=-456
>     ERROR: Process no longer exists
>     UNRESOLVED: gdb.base/setvar.exp: set variable unsigned short pointer
>     print v_unsigned_short_array
>     $85 = {123, 65080}
>     (gdb) print *(v_unsigned_short_pointer+1)
>     $86 = 65080
> 
>   I have not analyzed these errors.
> 
> gdb.base/weird.exp:
> 
>   This script produces 2 ERRORs on native i686-pc-linux-gn (red hat linux 7.1):
> 
>     ptype bad_type1
>     type = <unknown type>
>     (gdb) PASS: gdb.stabs/weird.exp: print bad_type1
>     ERROR: Process no longer exists
>     UNRESOLVED: gdb.stabs/weird.exp: array0 with strange index
>     ERROR: Process no longer exists
>     UNRESOLVED: gdb.stabs/weird.exp: array1 with strange index
> 
>   I have not analyzed these errors.


I'm just grabbing some low-hanging fruit: these unresolved test cases
are due to unescaped left curly braces.  There have been a couple of
threads on the subject in the past.  (In printcmds.exp it's solved by a
wrapper function that does the escaping.)  Ok to commit?


2001-10-29  Orjan Friberg  <orjanf@axis.com>

        * testsuite/gdb.base/setvar.exp: Escape curly braces.
        * testsuite/gdb.stabs/weird.exp: Ditto.


Index: setvar.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.base/setvar.exp,v
retrieving revision 1.3
diff -u -r1.3 setvar.exp
--- setvar.exp  2001/03/06 08:21:51     1.3
+++ setvar.exp  2001/10/29 10:25:13
@@ -320,7 +320,7 @@
 gdb_test "set v_signed_short_pointer=v_signed_short_array" ""
 gdb_test "set variable *(v_signed_short_pointer)=123" ""
 gdb_test "set variable *(v_signed_short_pointer+1)=-456" ""
-gdb_test "print v_signed_short_array" ".\[0-9\]* =.*\{123,.*-456\}" \
+gdb_test "print v_signed_short_array" ".\[0-9\]* =.*\\{123,.*-456\\}" \
     "set variable signed short pointer"
 gdb_test "print *(v_signed_short_pointer+1)" ".\[0-9\]*.*=.*-456"
 #
@@ -331,7 +331,7 @@
 gdb_test "set variable *(v_unsigned_short_pointer+1)=-456" ""
 # DTS 10060CLLbs - bad type info from cc
 if {$hp_cc_compiler} {setup_xfail hppa*-*-*11* 10060CLLbs}
-gdb_test "print v_unsigned_short_array" ".\[0-9\]* =.*\{123,.*65080\}"
\
+gdb_test "print v_unsigned_short_array" ".\[0-9\]*
=.*\\{123,.*65080\\}" \
     "set variable unsigned short pointer"
 # DTS 10060CLLbs - bad type info from cc
 if {$hp_cc_compiler} {setup_xfail hppa*-*-*11* 10060CLLbs}


Index: weird.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.stabs/weird.exp,v
retrieving revision 1.3
diff -u -r1.3 weird.exp
--- weird.exp   2001/10/26 04:17:45     1.3
+++ weird.exp   2001/10/29 10:25:41
@@ -170,9 +170,9 @@
 
        # GDB does not yet support arrays indexed by anything at all
unusual
        setup_xfail "*-*-*"
-       gdb_test "p array0" " = \{42, 43, 44, 45, 46, 47\}" "array0 with
strange index"
+       gdb_test "p array0" " = \\{42, 43, 44, 45, 46, 47\\}" "array0
with strange index"
        setup_xfail "*-*-*"
-       gdb_test "p array1" " = \{42, 43, 44\}" "array1 with strange
index"
+       gdb_test "p array1" " = \\{42, 43, 44\\}" "array1 with strange
index"
 
        # GDB does not yet support this feature
        gdb_test "whatis one_var" "type = inttype_one" \


-- 
Orjan Friberg              E-mail: orjan.friberg@axis.com
Axis Communications AB     Phone:  +46 46 272 17 68

Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Fernando Nasser]

Orjan Friberg wrote:
> 
> I'm just grabbing some low-hanging fruit: these unresolved test cases
> are due to unescaped left curly braces.  There have been a couple of
> threads on the subject in the past.  (In printcmds.exp it's solved by a
> wrapper function that does the escaping.)  Ok to commit?
> 

Yes, thanks!

And thanks Michael for running the tests and posting the results.

Fernando


> 2001-10-29  Orjan Friberg  <orjanf@axis.com>
> 
>         * testsuite/gdb.base/setvar.exp: Escape curly braces.
>         * testsuite/gdb.stabs/weird.exp: Ditto.
> 
> Index: setvar.exp
> ===================================================================
> RCS file: /cvs/src/src/gdb/testsuite/gdb.base/setvar.exp,v
> retrieving revision 1.3
> diff -u -r1.3 setvar.exp
> --- setvar.exp  2001/03/06 08:21:51     1.3
> +++ setvar.exp  2001/10/29 10:25:13
> @@ -320,7 +320,7 @@
>  gdb_test "set v_signed_short_pointer=v_signed_short_array" ""
>  gdb_test "set variable *(v_signed_short_pointer)=123" ""
>  gdb_test "set variable *(v_signed_short_pointer+1)=-456" ""
> -gdb_test "print v_signed_short_array" ".\[0-9\]* =.*\{123,.*-456\}" \
> +gdb_test "print v_signed_short_array" ".\[0-9\]* =.*\\{123,.*-456\\}" \
>      "set variable signed short pointer"
>  gdb_test "print *(v_signed_short_pointer+1)" ".\[0-9\]*.*=.*-456"
>  #
> @@ -331,7 +331,7 @@
>  gdb_test "set variable *(v_unsigned_short_pointer+1)=-456" ""
>  # DTS 10060CLLbs - bad type info from cc
>  if {$hp_cc_compiler} {setup_xfail hppa*-*-*11* 10060CLLbs}
> -gdb_test "print v_unsigned_short_array" ".\[0-9\]* =.*\{123,.*65080\}"
> \
> +gdb_test "print v_unsigned_short_array" ".\[0-9\]*
> =.*\\{123,.*65080\\}" \
>      "set variable unsigned short pointer"
>  # DTS 10060CLLbs - bad type info from cc
>  if {$hp_cc_compiler} {setup_xfail hppa*-*-*11* 10060CLLbs}
> 
> Index: weird.exp
> ===================================================================
> RCS file: /cvs/src/src/gdb/testsuite/gdb.stabs/weird.exp,v
> retrieving revision 1.3
> diff -u -r1.3 weird.exp
> --- weird.exp   2001/10/26 04:17:45     1.3
> +++ weird.exp   2001/10/29 10:25:41
> @@ -170,9 +170,9 @@
> 
>         # GDB does not yet support arrays indexed by anything at all
> unusual
>         setup_xfail "*-*-*"
> -       gdb_test "p array0" " = \{42, 43, 44, 45, 46, 47\}" "array0 with
> strange index"
> +       gdb_test "p array0" " = \\{42, 43, 44, 45, 46, 47\\}" "array0
> with strange index"
>         setup_xfail "*-*-*"
> -       gdb_test "p array1" " = \{42, 43, 44\}" "array1 with strange
> index"
> +       gdb_test "p array1" " = \\{42, 43, 44\\}" "array1 with strange
> index"
> 
>         # GDB does not yet support this feature
>         gdb_test "whatis one_var" "type = inttype_one" \
> 
> --
> Orjan Friberg              E-mail: orjan.friberg@axis.com
> Axis Communications AB     Phone:  +46 46 272 17 68

-- 
Fernando Nasser
Red Hat Canada Ltd.                     E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Orjan Friberg]

Fernando Nasser wrote:
> 
> Orjan Friberg wrote:
> >
> > I'm just grabbing some low-hanging fruit: these unresolved test cases
> > are due to unescaped left curly braces.  There have been a couple of
> > threads on the subject in the past.  (In printcmds.exp it's solved by a
> > wrapper function that does the escaping.)  Ok to commit?
> >
> 
> Yes, thanks!

Committed; thanks for the quick approval.

-- 
Orjan Friberg              E-mail: orjan.friberg@axis.com
Axis Communications AB     Phone:  +46 46 272 17 68

Re: [RFA] Escape curly braces (Re: Sunday Project gdb test results, 2001-10-28)

[Orjan Friberg]

Orjan Friberg wrote:
> 
> Fernando Nasser wrote:
> >
> > Orjan Friberg wrote:
> > >
> > > I'm just grabbing some low-hanging fruit: these unresolved test cases
> > > are due to unescaped left curly braces.  There have been a couple of
> > > threads on the subject in the past.  (In printcmds.exp it's solved by a
> > > wrapper function that does the escaping.)  Ok to commit?
> > >
> >
> > Yes, thanks!
> 
> Committed; thanks for the quick approval.

Fernando,

I noticed from Michael's latest Sunday Project gdb test results that
this patch (originally posted in
http://sources.redhat.com/ml/gdb-cvs/2001-10/msg00156.html) is not on
the 5.1 branch yet.  With your permission, I'd like to commit it there
also.

-- 
Orjan Friberg
Axis Communications AB