RT-Thread/rt-thread

Bug in linker script for all stm32 targets

Open

#7,527 opened on May 18, 2023

View on GitHub
 (0 comments) (3 reactions) (0 assignees)C (4,755 forks)batch import
good first issue

Repository metrics

Stars
 (9,223 stars)
PR merge metrics
 (Avg merge 7d 22h) (50 merged PRs in 30d)

Description

The following prints -2 but must print 1234:

void variadic_test(int n, ...) {
 va_list vl;
 va_start(vl, n);
 double val = va_arg(vl, double);
 rt_kprintf("%d", (int)val);
 va_end(vl);
}

void main() {
 double v = 1234.5678;
 variadic_test(1, v);
 while(1);
}

Seems like stack is aligned by 4 for all stm32 targets but double size is 8 bytes. The following patch fixes the issue:

--- link.lds.o  2023-05-18 11:11:50.831727100 +0300
+++ link.lds    2023-05-18 11:43:42.305975225 +0300
@@ -93,10 +93,10 @@
 
     .stack :
     {
-        . = ALIGN(4);
+        . = ALIGN(8);
         _sstack = .;
         . = . + _system_stack_size;
-        . = ALIGN(4);
+        . = ALIGN(8);
         _estack = .;
     } >RAM

Contributor guide

Research direction
Examine the linker script for STM32 targets, specifically the .stack section. The current alignment is 4 bytes, but variadic functions expecting 8 byte alignment for double cause incorrect values. Change ALIGN(4) to ALIGN(8) as shown in the provided patch.
Tech stack
c
Domain
embedded iot
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
1
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Under 1 hour
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Linker scripts
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
90