drivers/dma/mediatek/mtk-hsdma.c:451:17: sparse: sparse: incorrect type in assignment (different base types)

From: kernel test robot
Date: Mon Jul 06 2020 - 22:56:28 EST

Next message: Xing Zhengjun: "Re: [LKP] [x86, sched] 1567c3e346: vm-scalability.median -15.8% regression"
Previous message: Daniel Axtens: "Re: [PATCH] powerpc/pseries: detect secure and trusted boot state of the system."
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

tree: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
head: bfe91da29bfad9941d5d703d45e29f0812a20724
commit: a5460b5e5fb82656807840d40d3deaecad094044 READ_ONCE: Simplify implementations of {READ,WRITE}_ONCE()
date: 3 months ago
config: i386-randconfig-s002-20200707 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-14) 9.3.0
reproduce:
# apt-get install sparse
# sparse version: v0.6.2-31-gabbfd661-dirty
git checkout a5460b5e5fb82656807840d40d3deaecad094044
# save the attached .config to linux build tree
make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=i386

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@xxxxxxxxx>

sparse warnings: (new ones prefixed by >>)

>> drivers/dma/mediatek/mtk-hsdma.c:451:17: sparse: sparse: incorrect type in assignment (different base types) @@ expected restricted __le32 volatile @@ got unsigned int [usertype] src @@
drivers/dma/mediatek/mtk-hsdma.c:451:17: sparse: expected restricted __le32 volatile
>> drivers/dma/mediatek/mtk-hsdma.c:451:17: sparse: got unsigned int [usertype] src
drivers/dma/mediatek/mtk-hsdma.c:452:17: sparse: sparse: restricted __le32 degrades to integer
drivers/dma/mediatek/mtk-hsdma.c:452:17: sparse: sparse: incorrect type in assignment (different base types) @@ expected restricted __le32 volatile @@ got unsigned int @@
drivers/dma/mediatek/mtk-hsdma.c:452:17: sparse: expected restricted __le32 volatile
drivers/dma/mediatek/mtk-hsdma.c:452:17: sparse: got unsigned int
>> drivers/dma/mediatek/mtk-hsdma.c:456:17: sparse: sparse: incorrect type in assignment (different base types) @@ expected restricted __le32 volatile @@ got unsigned int [usertype] dest @@
drivers/dma/mediatek/mtk-hsdma.c:456:17: sparse: expected restricted __le32 volatile
>> drivers/dma/mediatek/mtk-hsdma.c:456:17: sparse: got unsigned int [usertype] dest
drivers/dma/mediatek/mtk-hsdma.c:457:17: sparse: sparse: incorrect type in assignment (different base types) @@ expected restricted __le32 volatile @@ got int @@
drivers/dma/mediatek/mtk-hsdma.c:457:17: sparse: expected restricted __le32 volatile
drivers/dma/mediatek/mtk-hsdma.c:457:17: sparse: got int
drivers/dma/mediatek/mtk-hsdma.c:580:33: sparse: sparse: restricted __le32 degrades to integer
drivers/dma/mediatek/mtk-hsdma.c:604:26: sparse: sparse: Using plain integer as NULL pointer
drivers/dma/mediatek/mtk-hsdma.c:878:18: sparse: sparse: incorrect type in initializer (different base types) @@ expected restricted __le32 [usertype] ddone @@ got unsigned long @@
drivers/dma/mediatek/mtk-hsdma.c:878:18: sparse: expected restricted __le32 [usertype] ddone
drivers/dma/mediatek/mtk-hsdma.c:878:18: sparse: got unsigned long
drivers/dma/mediatek/mtk-hsdma.c:879:16: sparse: sparse: incorrect type in initializer (different base types) @@ expected restricted __le32 [usertype] ls0 @@ got unsigned long @@
drivers/dma/mediatek/mtk-hsdma.c:879:16: sparse: expected restricted __le32 [usertype] ls0
drivers/dma/mediatek/mtk-hsdma.c:879:16: sparse: got unsigned long
drivers/dma/mediatek/mtk-hsdma.c:883:18: sparse: sparse: incorrect type in initializer (different base types) @@ expected restricted __le32 [usertype] ddone @@ got unsigned long @@
drivers/dma/mediatek/mtk-hsdma.c:883:18: sparse: expected restricted __le32 [usertype] ddone
drivers/dma/mediatek/mtk-hsdma.c:883:18: sparse: got unsigned long
drivers/dma/mediatek/mtk-hsdma.c:884:16: sparse: sparse: incorrect type in initializer (different base types) @@ expected restricted __le32 [usertype] ls0 @@ got unsigned long @@
drivers/dma/mediatek/mtk-hsdma.c:884:16: sparse: expected restricted __le32 [usertype] ls0
drivers/dma/mediatek/mtk-hsdma.c:884:16: sparse: got unsigned long

vim +451 drivers/dma/mediatek/mtk-hsdma.c

548c4597e984b7 Sean Wang 2018-03-15 411
548c4597e984b7 Sean Wang 2018-03-15 412 static int mtk_hsdma_issue_pending_vdesc(struct mtk_hsdma_device *hsdma,
548c4597e984b7 Sean Wang 2018-03-15 413 struct mtk_hsdma_pchan *pc,
548c4597e984b7 Sean Wang 2018-03-15 414 struct mtk_hsdma_vdesc *hvd)
548c4597e984b7 Sean Wang 2018-03-15 415 {
548c4597e984b7 Sean Wang 2018-03-15 416 struct mtk_hsdma_ring *ring = &pc->ring;
548c4597e984b7 Sean Wang 2018-03-15 417 struct mtk_hsdma_pdesc *txd, *rxd;
548c4597e984b7 Sean Wang 2018-03-15 418 u16 reserved, prev, tlen, num_sgs;
548c4597e984b7 Sean Wang 2018-03-15 419 unsigned long flags;
548c4597e984b7 Sean Wang 2018-03-15 420
548c4597e984b7 Sean Wang 2018-03-15 421 /* Protect against PC is accessed by multiple VCs simultaneously */
548c4597e984b7 Sean Wang 2018-03-15 422 spin_lock_irqsave(&hsdma->lock, flags);
548c4597e984b7 Sean Wang 2018-03-15 423
548c4597e984b7 Sean Wang 2018-03-15 424 /*
548c4597e984b7 Sean Wang 2018-03-15 425 * Reserve rooms, where pc->nr_free is used to track how many free
548c4597e984b7 Sean Wang 2018-03-15 426 * rooms in the ring being updated in user and IRQ context.
548c4597e984b7 Sean Wang 2018-03-15 427 */
548c4597e984b7 Sean Wang 2018-03-15 428 num_sgs = DIV_ROUND_UP(hvd->len, MTK_HSDMA_MAX_LEN);
548c4597e984b7 Sean Wang 2018-03-15 429 reserved = min_t(u16, num_sgs, atomic_read(&pc->nr_free));
548c4597e984b7 Sean Wang 2018-03-15 430
548c4597e984b7 Sean Wang 2018-03-15 431 if (!reserved) {
548c4597e984b7 Sean Wang 2018-03-15 432 spin_unlock_irqrestore(&hsdma->lock, flags);
548c4597e984b7 Sean Wang 2018-03-15 433 return -ENOSPC;
548c4597e984b7 Sean Wang 2018-03-15 434 }
548c4597e984b7 Sean Wang 2018-03-15 435
548c4597e984b7 Sean Wang 2018-03-15 436 atomic_sub(reserved, &pc->nr_free);
548c4597e984b7 Sean Wang 2018-03-15 437
548c4597e984b7 Sean Wang 2018-03-15 438 while (reserved--) {
548c4597e984b7 Sean Wang 2018-03-15 439 /* Limit size by PD capability for valid data moving */
548c4597e984b7 Sean Wang 2018-03-15 440 tlen = (hvd->len > MTK_HSDMA_MAX_LEN) ?
548c4597e984b7 Sean Wang 2018-03-15 441 MTK_HSDMA_MAX_LEN : hvd->len;
548c4597e984b7 Sean Wang 2018-03-15 442
548c4597e984b7 Sean Wang 2018-03-15 443 /*
548c4597e984b7 Sean Wang 2018-03-15 444 * Setup PDs using the remaining VD info mapped on those
548c4597e984b7 Sean Wang 2018-03-15 445 * reserved rooms. And since RXD is shared memory between the
548c4597e984b7 Sean Wang 2018-03-15 446 * host and the device allocated by dma_alloc_coherent call,
548c4597e984b7 Sean Wang 2018-03-15 447 * the helper macro WRITE_ONCE can ensure the data written to
548c4597e984b7 Sean Wang 2018-03-15 448 * RAM would really happens.
548c4597e984b7 Sean Wang 2018-03-15 449 */
548c4597e984b7 Sean Wang 2018-03-15 450 txd = &ring->txd[ring->cur_tptr];
548c4597e984b7 Sean Wang 2018-03-15 @451 WRITE_ONCE(txd->desc1, hvd->src);
548c4597e984b7 Sean Wang 2018-03-15 452 WRITE_ONCE(txd->desc2,
548c4597e984b7 Sean Wang 2018-03-15 453 hsdma->soc->ls0 | MTK_HSDMA_DESC_PLEN(tlen));
548c4597e984b7 Sean Wang 2018-03-15 454
548c4597e984b7 Sean Wang 2018-03-15 455 rxd = &ring->rxd[ring->cur_tptr];
548c4597e984b7 Sean Wang 2018-03-15 @456 WRITE_ONCE(rxd->desc1, hvd->dest);
548c4597e984b7 Sean Wang 2018-03-15 457 WRITE_ONCE(rxd->desc2, MTK_HSDMA_DESC_PLEN(tlen));
548c4597e984b7 Sean Wang 2018-03-15 458
548c4597e984b7 Sean Wang 2018-03-15 459 /* Associate VD, the PD belonged to */
548c4597e984b7 Sean Wang 2018-03-15 460 ring->cb[ring->cur_tptr].vd = &hvd->vd;
548c4597e984b7 Sean Wang 2018-03-15 461
548c4597e984b7 Sean Wang 2018-03-15 462 /* Move forward the pointer of TX ring */
548c4597e984b7 Sean Wang 2018-03-15 463 ring->cur_tptr = MTK_HSDMA_NEXT_DESP_IDX(ring->cur_tptr,
548c4597e984b7 Sean Wang 2018-03-15 464 MTK_DMA_SIZE);
548c4597e984b7 Sean Wang 2018-03-15 465
548c4597e984b7 Sean Wang 2018-03-15 466 /* Update VD with remaining data */
548c4597e984b7 Sean Wang 2018-03-15 467 hvd->src += tlen;
548c4597e984b7 Sean Wang 2018-03-15 468 hvd->dest += tlen;
548c4597e984b7 Sean Wang 2018-03-15 469 hvd->len -= tlen;
548c4597e984b7 Sean Wang 2018-03-15 470 }
548c4597e984b7 Sean Wang 2018-03-15 471
548c4597e984b7 Sean Wang 2018-03-15 472 /*
548c4597e984b7 Sean Wang 2018-03-15 473 * Tagging flag for the last PD for VD will be responsible for
548c4597e984b7 Sean Wang 2018-03-15 474 * completing VD.
548c4597e984b7 Sean Wang 2018-03-15 475 */
548c4597e984b7 Sean Wang 2018-03-15 476 if (!hvd->len) {
548c4597e984b7 Sean Wang 2018-03-15 477 prev = MTK_HSDMA_LAST_DESP_IDX(ring->cur_tptr, MTK_DMA_SIZE);
548c4597e984b7 Sean Wang 2018-03-15 478 ring->cb[prev].flag = MTK_HSDMA_VDESC_FINISHED;
548c4597e984b7 Sean Wang 2018-03-15 479 }
548c4597e984b7 Sean Wang 2018-03-15 480
548c4597e984b7 Sean Wang 2018-03-15 481 /* Ensure all changes indeed done before we're going on */
548c4597e984b7 Sean Wang 2018-03-15 482 wmb();
548c4597e984b7 Sean Wang 2018-03-15 483
548c4597e984b7 Sean Wang 2018-03-15 484 /*
548c4597e984b7 Sean Wang 2018-03-15 485 * Updating into hardware the pointer of TX ring lets HSDMA to take
548c4597e984b7 Sean Wang 2018-03-15 486 * action for those pending PDs.
548c4597e984b7 Sean Wang 2018-03-15 487 */
548c4597e984b7 Sean Wang 2018-03-15 488 mtk_dma_write(hsdma, MTK_HSDMA_TX_CPU, ring->cur_tptr);
548c4597e984b7 Sean Wang 2018-03-15 489
548c4597e984b7 Sean Wang 2018-03-15 490 spin_unlock_irqrestore(&hsdma->lock, flags);
548c4597e984b7 Sean Wang 2018-03-15 491
548c4597e984b7 Sean Wang 2018-03-15 492 return 0;
548c4597e984b7 Sean Wang 2018-03-15 493 }
548c4597e984b7 Sean Wang 2018-03-15 494

:::::: The code at line 451 was first introduced by commit
:::::: 548c4597e984b79aad8190235d664f1c3a433f94 dmaengine: mediatek: Add MediaTek High-Speed DMA controller for MT7622 and MT7623 SoC

:::::: TO: Sean Wang <sean.wang@xxxxxxxxxxxx>
:::::: CC: Vinod Koul <vinod.koul@xxxxxxxxx>

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@xxxxxxxxxxxx

Attachment: .config.gz
Description: application/gzip

Next message: Xing Zhengjun: "Re: [LKP] [x86, sched] 1567c3e346: vm-scalability.median -15.8% regression"
Previous message: Daniel Axtens: "Re: [PATCH] powerpc/pseries: detect secure and trusted boot state of the system."
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]