| Message ID | 20200311171210.GF2156@tucnak |
|---|---|
| State | New |
| Headers | show |
| Series | pdp11: Fix handling of common (local and global) vars [PR94134] | expand |
Ok, thanks. paul > On Mar 11, 2020, at 1:12 PM, Jakub Jelinek <jakub@redhat.com> wrote: > > Hi! > > As mentioned in the PR, the generic code decides to put the a variable into > lcomm_section, which is a NOSWITCH section and thus the generic code doesn't > switch into a particular section before using > ASM_OUTPUT{_ALIGNED{,_DECL}_}_LOCAL, on many targets that results just in > .lcomm (or for non-local .comm) directives which don't need a switch to some > section, other targets put switch_to_section (bss_section) at the start of > that macro. > pdp11 doesn't do that (and doesn't have bss_section), and so emits the > lcomm/comm variables in whatever section is current (it has only .text/.data > and for DEC assembler rodata). > > The following patch fixes that by putting it always into data section, and > additionally avoids emitting an empty line in the assembly for the lcomm > vars. > > Tested on the testcase, I'm afraid I have no other way to test this target. > Ok for trunk? > > 2020-03-11 Jakub Jelinek <jakub@redhat.com> > > PR target/94134 > * config/pdp11/pdp11.c (pdp11_asm_output_var): Call switch_to_section > at the start to switch to data section. Don't print extra newline if > .globl directive has not been emitted. > > * gcc.c-torture/execute/pr94134.c: New test. > > --- gcc/config/pdp11/pdp11.c.jj 2020-01-12 11:54:36.382413876 +0100 > +++ gcc/config/pdp11/pdp11.c 2020-03-11 15:44:43.373970000 +0100 > @@ -743,6 +743,7 @@ void > pdp11_asm_output_var (FILE *file, const char *name, int size, > int align, bool global) > { > + switch_to_section (data_section); > if (align > 8) > fprintf (file, "\t.even\n"); > if (TARGET_DEC_ASM) > @@ -763,8 +764,8 @@ pdp11_asm_output_var (FILE *file, const > { > fprintf (file, ".globl "); > assemble_name (file, name); > + fprintf (file, "\n"); > } > - fprintf (file, "\n"); > assemble_name (file, name); > fputs (":", file); > ASM_OUTPUT_SKIP (file, size); > --- gcc/testsuite/gcc.c-torture/execute/pr94134.c.jj 2020-03-11 15:46:09.540710642 +0100 > +++ gcc/testsuite/gcc.c-torture/execute/pr94134.c 2020-03-11 15:40:18.538840663 +0100 > @@ -0,0 +1,14 @@ > +/* PR target/94134 */ > + > +static volatile int a = 0; > +static volatile int b = 1; > + > +int > +main () > +{ > + a++; > + b++; > + if (a != 1 || b != 2) > + __builtin_abort (); > + return 0; > +} > > Jakub >
--- gcc/config/pdp11/pdp11.c.jj 2020-01-12 11:54:36.382413876 +0100 +++ gcc/config/pdp11/pdp11.c 2020-03-11 15:44:43.373970000 +0100 @@ -743,6 +743,7 @@ void pdp11_asm_output_var (FILE *file, const char *name, int size, int align, bool global) { + switch_to_section (data_section); if (align > 8) fprintf (file, "\t.even\n"); if (TARGET_DEC_ASM) @@ -763,8 +764,8 @@ pdp11_asm_output_var (FILE *file, const { fprintf (file, ".globl "); assemble_name (file, name); + fprintf (file, "\n"); } - fprintf (file, "\n"); assemble_name (file, name); fputs (":", file); ASM_OUTPUT_SKIP (file, size); --- gcc/testsuite/gcc.c-torture/execute/pr94134.c.jj 2020-03-11 15:46:09.540710642 +0100 +++ gcc/testsuite/gcc.c-torture/execute/pr94134.c 2020-03-11 15:40:18.538840663 +0100 @@ -0,0 +1,14 @@ +/* PR target/94134 */ + +static volatile int a = 0; +static volatile int b = 1; + +int +main () +{ + a++; + b++; + if (a != 1 || b != 2) + __builtin_abort (); + return 0; +}
Hi! As mentioned in the PR, the generic code decides to put the a variable into lcomm_section, which is a NOSWITCH section and thus the generic code doesn't switch into a particular section before using ASM_OUTPUT{_ALIGNED{,_DECL}_}_LOCAL, on many targets that results just in .lcomm (or for non-local .comm) directives which don't need a switch to some section, other targets put switch_to_section (bss_section) at the start of that macro. pdp11 doesn't do that (and doesn't have bss_section), and so emits the lcomm/comm variables in whatever section is current (it has only .text/.data and for DEC assembler rodata). The following patch fixes that by putting it always into data section, and additionally avoids emitting an empty line in the assembly for the lcomm vars. Tested on the testcase, I'm afraid I have no other way to test this target. Ok for trunk? 2020-03-11 Jakub Jelinek <jakub@redhat.com> PR target/94134 * config/pdp11/pdp11.c (pdp11_asm_output_var): Call switch_to_section at the start to switch to data section. Don't print extra newline if .globl directive has not been emitted. * gcc.c-torture/execute/pr94134.c: New test. Jakub