Commit15e46fd

committed

Fix bug in record_image_ops on big endian machines.

The buildfarm pointed out the problem.Fix based on suggestion by Robert Haas.

1 parent4d212ba commit15e46fdCopy full SHA for 15e46fd

File tree

-6

lines changed

-6

lines changed

Lines changed: 3 additions & 6 deletions

Original file line number	Diff line number	Diff line change
`@@ -1464,9 +1464,8 @@ record_image_cmp(PG_FUNCTION_ARGS)`
`1464`	`1464`	`}`
`1465`	`1465`	`elseif (tupdesc1->attrs[i1]->attbyval)`
`1466`	`1466`	`{`
`1467`		`-cmpresult=memcmp(&(values1[i1]),`
`1468`		`-&(values2[i2]),`
`1469`		`-tupdesc1->attrs[i1]->attlen);`
	`1467`	`+if (values1[i1]!=values2[i2])`
	`1468`	`+cmpresult= (values1[i1]<values2[i2]) ?-1 :1;`
`1470`	`1469`	`}`
`1471`	`1470`	`else`
`1472`	`1471`	`{`
`@@ -1695,9 +1694,7 @@ record_image_eq(PG_FUNCTION_ARGS)`
`1695`	`1694`	`}`
`1696`	`1695`	`elseif (tupdesc1->attrs[i1]->attbyval)`
`1697`	`1696`	`{`
`1698`		`-result= (memcmp(&(values1[i1]),`
`1699`		`-&(values2[i2]),`
`1700`		`-tupdesc1->attrs[i1]->attlen)==0);`
	`1697`	`+result= (values1[i1]==values2[i2]);`
`1701`	`1698`	`}`
`1702`	`1699`	`else`
`1703`	`1700`	`{`

Comments

(0)