This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
| Other format: | [Raw text] | |
In my GCC 3.4 status report, I mentioned that there is an open regression in GCC 3.4 relating to -fwritable-strings. We have made a conscious decision to remove features relating to "traditional", and this option is one such feature. Therefore, I think that we should probably just remove this option in GCC 3.4. Kelley Cook has a patch for GCC 3.3.3 to deprecate this option, and Zack Weinberg has a (somewhat out of date) patch to remove -fwritable-strings from the mainline. Richard Earnshaw suggested that we could deprecate the feature in GCC 3.4, and just not fix the bug. That's a plausible alternative, but I'd prefer just to completely eliminate the feature. The implementation of this feature is particularly unfortunate; rather than being handled entirely in the front ends (as it could be), there is actually logic in several of the back ends to deal with it. It was quite surprising to me to see how much code Zack could remove by getting rid of this option. Zack raised the idea of removing -fwritable-strings back in April, and nobody objected -- but we never followed through. So, here is a final chance: Does anyone object to the plan of (a) deprecating -fwritable-strings in GCC 3.3.3 and (b) removing it in GCC 3.4? If so, why? -- Mark Mitchell <mark@codesourcery.com> CodeSourcery, LLC
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |