[EXT_EP-10449] Optimizer drops part of a compound conditional expression controlling a loop Created: 12/Jul/21 Updated: 10/Feb/22 Resolved: 10/Feb/22 |
|
| Status: | Fixed |
| Project: | Embedded Software & Tools |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Medium |
| Reporter: | TI User | Assignee: | TI User |
| Resolution: | Fixed | Votes: | 0 |
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Product: | Code Generation Tools |
| Internal ID: | CODEGEN-9082 |
| Forum URL: | https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1014790/c6000-cgt-cgt-8-3-9---cl6x-not-considering-multiple-exit-conditions-present-in-for-loop-syntax |
| Found In Release: | C6000_8.3.9 |
| Fix In Release: | C7000_2.1.1.LTS MSP430_18.12.9.LTS* C6000_8.3.12* C7000_1.4.3.LTS* MSP430_20.2.6.LTS MSP430_21.6.1.LTS* |
| Affected Platform/Device: | default |
| Description |
|
The attached test case contains this code ... for (u32 i = 0; (i < arrSize && i < CByteAligner::bitsToOctetsRoundUp(tbSize)); i++) /* PROBLEM FOR LOOP */
localMem[i] = p_corruptedPdu[i];
The generated code accounts for the i < CByteAligner::bitsToOctetsRoundUp(tbSize)) part of the condition controlling the loop. But not the i < arrSize part. Build it ... % cl6x -@options.txt file.cpp Inspect the compiler generated comments for the problem loop. ;* 128211 ----------------------- memset(&localMem, 0, 16u); ;* 128212 ----------------------- K$13 = tbSize+7u; ;* 128212 ----------------------- if ( !(C$5 = K$13>>3) ) goto g4; ;** ----------------------- U$21 = p_corruptedPdu; ;* 128213 ----------------------- L$1 = C$5; ;** ----------------------- U$23 = &localMem[0]; ;** ----------------------- #pragma MUST_ITERATE(1, 16, 1) ;** ----------------------- #pragma LOOP_FLAGS(4096u) ;** -----------------------g3: ;* 128213 ----------------------- *U$23++ = *U$21++; ;* 128212 ----------------------- if ( L$1 = L$1-1 ) goto g3; ;** -----------------------g4: L$1 controls this loop. It down counts from CByteAligner::bitsToOctetsRoundUp(tbSize)) to 0. But this ignores the i < 16 part of the loop condition. |