Thread: [HACKERS] [PATCH] Removes uninitialized variable compiler warning

[HACKERS] [PATCH] Removes uninitialized variable compiler warning

From
Todd Sedano
Date:
This patch removes a compiler warning. 

warning: variable 'lenlemm' is uninitialized when used here [-Wuninitialized]

This is my first commit to postgres. I've read through http://wiki.postgresql.org/wiki/Submitting_a_Patch, but I may have missed something. 


diff --git a/src/backend/tsearch/ts_parse.c b/src/backend/tsearch/ts_parse.c
index b612fb0e2c..3d66b2babd 100644
--- a/src/backend/tsearch/ts_parse.c
+++ b/src/backend/tsearch/ts_parse.c
@@ -357,8 +357,8 @@ LexizeExec(LexizeData *ld, ParsedLex **correspondLexem)
 void
 parsetext(Oid cfgId, ParsedText *prs, char *buf, int buflen)
 {
-       int                     type,
-                               lenlemm;
+       int                     type;
+       int                     lenlemm = 0;
        char       *lemm = NULL;
        LexizeData      ldata;
        TSLexeme   *norms;
Attachment

Re: [HACKERS] [PATCH] Removes uninitialized variable compiler warning

From
Tom Lane
Date:
Todd Sedano <tsedano@pivotal.io> writes:
> This patch removes a compiler warning.
> warning: variable 'lenlemm' is uninitialized when used here
> [-Wuninitialized]

Hm, on what compiler?  AFAICS, that parsetext() function hasn't
changed meaningfully since 2007, and nobody complained of
uninitialized-variable warnings in it before.

We're generally willing to try to silence such warnings on mainstream
compilers, but not on weird ones ...
        regards, tom lane



Re: [HACKERS] [PATCH] Removes uninitialized variable compiler warning

From
Todd Sedano
Date:
We can disregard this patch. It appears that PointerGetDatum initializes lenlemm.


On Mon, Mar 20, 2017 at 10:04 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Todd Sedano <tsedano@pivotal.io> writes:
> This patch removes a compiler warning.
> warning: variable 'lenlemm' is uninitialized when used here
> [-Wuninitialized]

Hm, on what compiler?  AFAICS, that parsetext() function hasn't
changed meaningfully since 2007, and nobody complained of
uninitialized-variable warnings in it before.

We're generally willing to try to silence such warnings on mainstream
compilers, but not on weird ones ...

                        regards, tom lane