[PATCH] powerpc: Eliminate "exceeds stub group size" linker warning
authorPaul Mackerras <paulus@samba.org>
Wed, 1 Nov 2006 22:44:37 +0000 (09:44 +1100)
committerLinus Torvalds <torvalds@g5.osdl.org>
Wed, 1 Nov 2006 22:56:59 +0000 (14:56 -0800)
It turns out that the linker warnings on 64-bit powerpc about "section
blah exceeds stub group size" were being triggered by conditional
branches in head_64.S branching to global symbols, whether in
head_64.S or in other files.  This eliminates the warnings by making
some global symbols in head_64.S no longer global, and by rearranging
some branches.

Signed-off-by: Paul Mackerras <paulus@samba.org>
[ Yee-haa. Maybe I'll notice newly introduced real warnings now - Linus ]
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
arch/powerpc/kernel/head_64.S

index 291e362..e720729 100644 (file)
@@ -487,7 +487,7 @@ BEGIN_FTR_SECTION
        rlwimi  r13,r12,16,0x20
        mfcr    r12
        cmpwi   r13,0x2c
-       beq     .do_stab_bolted_pSeries
+       beq     do_stab_bolted_pSeries
        mtcrf   0x80,r12
        mfspr   r12,SPRN_SPRG2
 END_FTR_SECTION_IFCLR(CPU_FTR_SLB)
@@ -600,7 +600,7 @@ system_call_pSeries:
        STD_EXCEPTION_PSERIES(., performance_monitor)
 
        .align  7
-_GLOBAL(do_stab_bolted_pSeries)
+do_stab_bolted_pSeries:
        mtcrf   0x80,r12
        mfspr   r12,SPRN_SPRG2
        EXCEPTION_PROLOG_PSERIES(PACA_EXSLB, .do_stab_bolted)
@@ -1046,7 +1046,7 @@ slb_miss_fault:
        li      r5,0
        std     r4,_DAR(r1)
        std     r5,_DSISR(r1)
-       b       .handle_page_fault
+       b       handle_page_fault
 
 unrecov_user_slb:
        EXCEPTION_PROLOG_COMMON(0x4200, PACA_EXGEN)
@@ -1174,12 +1174,13 @@ program_check_common:
        .globl fp_unavailable_common
 fp_unavailable_common:
        EXCEPTION_PROLOG_COMMON(0x800, PACA_EXGEN)
-       bne     .load_up_fpu            /* if from user, just load it up */
+       bne     1f                      /* if from user, just load it up */
        bl      .save_nvgprs
        addi    r3,r1,STACK_FRAME_OVERHEAD
        ENABLE_INTS
        bl      .kernel_fp_unavailable_exception
        BUG_OPCODE
+1:     b       .load_up_fpu
 
        .align  7
        .globl altivec_unavailable_common
@@ -1279,10 +1280,10 @@ _GLOBAL(do_hash_page)
        std     r4,_DSISR(r1)
 
        andis.  r0,r4,0xa450            /* weird error? */
-       bne-    .handle_page_fault      /* if not, try to insert a HPTE */
+       bne-    handle_page_fault       /* if not, try to insert a HPTE */
 BEGIN_FTR_SECTION
        andis.  r0,r4,0x0020            /* Is it a segment table fault? */
-       bne-    .do_ste_alloc           /* If so handle it */
+       bne-    do_ste_alloc            /* If so handle it */
 END_FTR_SECTION_IFCLR(CPU_FTR_SLB)
 
        /*
@@ -1324,7 +1325,7 @@ BEGIN_FW_FTR_SECTION
         * because ret_from_except_lite will check for and handle pending
         * interrupts if necessary.
         */
-       beq     .ret_from_except_lite
+       beq     13f
        /* For a hash failure, we don't bother re-enabling interrupts */
        ble-    12f
 
@@ -1346,14 +1347,14 @@ BEGIN_FW_FTR_SECTION
 END_FW_FTR_SECTION_IFCLR(FW_FEATURE_ISERIES)
 
 /* Here we have a page fault that hash_page can't handle. */
-_GLOBAL(handle_page_fault)
+handle_page_fault:
        ENABLE_INTS
 11:    ld      r4,_DAR(r1)
        ld      r5,_DSISR(r1)
        addi    r3,r1,STACK_FRAME_OVERHEAD
        bl      .do_page_fault
        cmpdi   r3,0
-       beq+    .ret_from_except_lite
+       beq+    13f
        bl      .save_nvgprs
        mr      r5,r3
        addi    r3,r1,STACK_FRAME_OVERHEAD
@@ -1370,12 +1371,14 @@ _GLOBAL(handle_page_fault)
        bl      .low_hash_fault
        b       .ret_from_except
 
+13:    b       .ret_from_except_lite
+
        /* here we have a segment miss */
-_GLOBAL(do_ste_alloc)
+do_ste_alloc:
        bl      .ste_allocate           /* try to insert stab entry */
        cmpdi   r3,0
-       beq+    fast_exception_return
-       b       .handle_page_fault
+       bne-    handle_page_fault
+       b       fast_exception_return
 
 /*
  * r13 points to the PACA, r9 contains the saved CR,