riscv: extable: fix err reg writing in dedicated uaccess handler
PATCH 1/1 riscv: extable: fix err reg writing in dedicated uaccess handler
Mayuresh reported commit 20802d8d477d ("riscv: extable: add a dedicated uaccess handler") breaks the writev02 test case in LTP. This is due to the err reg isn't correctly set with the errno(-EFAULT in writev02 case). First of all, the err and zero regs are reg numbers rather than reg offsets in struct pt_regs; Secondly, regs_set_gpr() should write the regs when offset isn't zero(zero means epc)
Fix it by correcting regs_set_gpr() logic and passing the correct reg offset to it.
--- arch/riscv/mm/extable.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
arch/riscv/mm/extable.c +3 -3
@@ -33,7 +33,7 @@ static inline void regs_set_gpr(struct pt_regs *regs, unsigned int offset,if (unlikely (offset >MAX_REG_OFFSET ))return ; -if (!offset) +if (offset) *(unsigned long *)((unsigned long )regs + offset) = val; }
@@ -43,8 +43,8 @@ static bool ex_handler_uaccess_err_zero(const struct exception_table_entry *ex,int reg_err =FIELD_GET (EX_DATA_REG_ERR , ex->data);int reg_zero =FIELD_GET (EX_DATA_REG_ZERO , ex->data); -regs_set_gpr (regs, reg_err, -EFAULT ); -regs_set_gpr (regs, reg_zero,0 ); +regs_set_gpr (regs, reg_err *sizeof (unsigned long ), -EFAULT ); +regs_set_gpr (regs, reg_zero *sizeof (unsigned long ),0 ); regs->epc =get_ex_fixup (ex);return true;
Signature · 1 line
2.34.1
Palmer Dabbelt
maintainerQuoted 31 lines · On Sun, 23 Jan 2022 01:04:17 PST (-0800), [email protected] wrote:
Mayuresh reported commit
20802d8d477d("riscv: extable: add a dedicated uaccess handler") breaks the writev02 test case in LTP. This is due to the err reg isn't correctly set with the errno(-EFAULT in writev02 case). First of all, the err and zero regs are reg numbers rather than reg offsets in structpt_regs; Secondly,regs_set_gpr()should write the regs when offset isn't zero(zero means epc)Fix it by correcting
regs_set_gpr()logic and passing the correct reg offset to it.Reported-by Mayuresh Chitale <[email protected]>Fixes 20802d8d477d ("riscv: extable: add a dedicated uaccess handler")Signed-off-by Jisheng Zhang <[email protected]>--- arch/riscv/mm/extable.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)arch/riscv/mm/extable.c +1
-1@@ -33,7 +33,7 @@ static inline void regs_set_gpr(struct pt_regs *regs, unsigned int offset,if (unlikely (offset >MAX_REG_OFFSET ))return ; -if (!offset) +if (offset) *(unsigned long *)((unsigned long )regs + offset) = val; }
arch/riscv/mm/extable.c static bool ex_handler_uaccess_err_zero(const struct exception_table_entry *ex, · line 50
Quoted 10 lines
arch/riscv/mm/extable.c +2
-2@@ -43,8 +43,8 @@ static bool ex_handler_uaccess_err_zero(const struct exception_table_entry *ex,int reg_err =FIELD_GET (EX_DATA_REG_ERR , ex->data);int reg_zero =FIELD_GET (EX_DATA_REG_ZERO , ex->data); -regs_set_gpr (regs, reg_err, -EFAULT ); -regs_set_gpr (regs, reg_zero,0 ); +regs_set_gpr (regs, reg_err *sizeof (unsigned long ), -EFAULT ); +regs_set_gpr (regs, reg_zero *sizeof (unsigned long ),0 ); regs->epc =get_ex_fixup (ex);return true;
Thanks, this is on fixes. I'm not sure if it's saner to have these be register numbers rather than offsets, but regs_get_register is this way so it's probably better to have them match.
_______________________________________________
linux-riscv mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-riscv