Re: [patch emutls]:Fix PR 65566

2015-03-31 Thread Jakub Jelinek
On Tue, Mar 31, 2015 at 02:21:23PM +0200, Kai Tietz wrote: > addressable used static QI file thread-local-var-1-lbv.c line 28 > col 5 align 8 context > attributes initial > result > ignored SI file thread-local-var-1-lbv.c line 28 col 5 size > unit size > align 32 con

Re: [patch emutls]:Fix PR 65566

2015-03-31 Thread Kai Tietz
2015-03-31 13:42 GMT+02:00 Jakub Jelinek : > On Tue, Mar 31, 2015 at 01:35:56PM +0200, Kai Tietz wrote: >> Hi, >> >> This patch avoids that we try to operate on function-decl's cfun equal >> to NULL within lower_emutls_function_body. > > If DECL_STRUCT_FUNCTION (node->decl) is already NULL (for whi

Re: [patch emutls]:Fix PR 65566

2015-03-31 Thread Jakub Jelinek
On Tue, Mar 31, 2015 at 01:35:56PM +0200, Kai Tietz wrote: > Hi, > > This patch avoids that we try to operate on function-decl's cfun equal > to NULL within lower_emutls_function_body. If DECL_STRUCT_FUNCTION (node->decl) is already NULL (for which functions?), then what is the point doing push_c

[patch emutls]:Fix PR 65566

2015-03-31 Thread Kai Tietz
Hi, This patch avoids that we try to operate on function-decl's cfun equal to NULL within lower_emutls_function_body. ChangeLog 2015-03-31 Kai Tietz PR target/65566 * tree-emutls.c (lower_emutls_function_body): Don't try to operate on node's decl function is NULL. Ok for apply?