[4.3 Regression] Wrong code when -O3 or -O2 -fstack-protector used
Bug #316019 reported by
Matthias Klose
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
gcc |
Fix Released
|
Medium
|
|||
gcc-4.3 (Ubuntu) |
Fix Released
|
High
|
Unassigned | ||
Intrepid |
Fix Released
|
High
|
Unassigned | ||
Jaunty |
Fix Released
|
High
|
Unassigned |
Bug Description
Binary package hint: gcc-4.3
PR middle-end/38616
Changed in gcc: | |
status: | Unknown → Confirmed |
Changed in gcc-4.3: | |
importance: | Undecided → Medium |
status: | New → In Progress |
status: | New → In Progress |
Changed in gcc: | |
status: | Confirmed → Fix Released |
Changed in gcc: | |
status: | Unknown → New |
Changed in gcc: | |
status: | New → Confirmed |
Changed in gcc: | |
status: | Confirmed → Fix Released |
Changed in gcc: | |
importance: | Unknown → Medium |
To post a comment you must log in.
On 64bit, the truncation happens at byte 24 (32bit happens at 8). Does not happen on Gutsy or Hardy (or Dapper, which didn't use -fstack-protector).
I have no idea how to even search for stack variables that are assigned an initial string value. That's rather semantically sensitive to search for.