We can have SAVE_MULTIPLE while we do not have REST_MULTIPLE. If the
inline restore does not restore all registers, the CFI for the save
and restore can conflict if things are shrink-wrapped.
We could restore all registers that are saved (not ideal), or emit
the CFI notes to say we did (which will work fine, but is also not
so great); instead, let's not save the registers that are unused.
PR target/80938
* config/rs6000/rs6000.c (rs6000_savres_strategy): Don't use
SAVE_MULTIPLE if not all the registers that saves, should be saved.
From-SVN: r251005
+2017-08-09 Segher Boessenkool <segher@kernel.crashing.org>
+
+ PR target/80938
+ * config/rs6000/rs6000.c (rs6000_savres_strategy): Don't use
+ SAVE_MULTIPLE if not all the registers that saves, should be saved.
+
2017-08-09 Jim Wilson <jim.wilson@linaro.org>
* config/aarch64/aarch64-cores.def (falkor): Use falkor pipeline.
else if (!lr_save_p && info->first_gp_reg_save > 29)
strategy |= SAVE_INLINE_GPRS | REST_INLINE_GPRS;
+ /* We can only use save multiple if we need to save all the registers from
+ first_gp_reg_save. Otherwise, the CFI gets messed up (we save some
+ register we do not restore). */
+ if (strategy & SAVE_MULTIPLE)
+ {
+ int i;
+
+ for (i = info->first_gp_reg_save; i < 32; i++)
+ if (fixed_reg_p (i) || !save_reg_p (i))
+ {
+ strategy &= ~SAVE_MULTIPLE;
+ break;
+ }
+ }
+
/* We can only use load multiple or the out-of-line routines to
restore gprs if we've saved all the registers from
first_gp_reg_save. Otherwise, we risk loading garbage.