| Message ID | 1581413771-18005-1-git-send-email-vbadigan@codeaurora.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [V1] mmc: mmc_test: Pass different sg lists for non-blocking requests | expand |
Attaching the full error log On 2/11/2020 3:06 PM, Veerabhadrarao Badiganti wrote: > Supply a separate sg list for each of the request in non-blocking > IO test cases where two requests will be issued at same time. > > Otherwise, sg memory may get unmapped when a request is done while > same memory is being accessed by controller from the other request, > and it leads to iommu errors with below call stack: > > __arm_lpae_unmap+0x2e0/0x478 > arm_lpae_unmap+0x54/0x70 > arm_smmu_unmap+0x64/0xa4 > __iommu_unmap+0xb8/0x1f0 > iommu_unmap_fast+0x38/0x48 > __iommu_dma_unmap+0x88/0x108 > iommu_dma_unmap_sg+0x90/0xa4 > sdhci_post_req+0x5c/0x78 > mmc_test_start_areq+0x10c/0x120 [mmc_test] > mmc_test_area_io_seq+0x150/0x264 [mmc_test] > mmc_test_rw_multiple+0x174/0x1c0 [mmc_test] > mmc_test_rw_multiple_sg_len+0x44/0x6c [mmc_test] > mmc_test_profile_sglen_wr_nonblock_perf+0x6c/0x94 [mmc_test] > mtf_test_write+0x238/0x3cc [mmc_test] > > Signed-off-by: Veerabhadrarao Badiganti <vbadigan@codeaurora.org> > --- > drivers/mmc/core/mmc_test.c | 38 ++++++++++++++++++++++++++++++++------ > 1 file changed, 32 insertions(+), 6 deletions(-) > > diff --git a/drivers/mmc/core/mmc_test.c b/drivers/mmc/core/mmc_test.c > index 492dd45..69bdf60 100644 > --- a/drivers/mmc/core/mmc_test.c > +++ b/drivers/mmc/core/mmc_test.c > @@ -71,6 +71,7 @@ struct mmc_test_mem { > * @sg_len: length of currently mapped scatterlist @sg > * @mem: allocated memory > * @sg: scatterlist > + * @sg_areq: scatterlist for non blocking request > */ > struct mmc_test_area { > unsigned long max_sz; > @@ -82,6 +83,7 @@ struct mmc_test_area { > unsigned int sg_len; > struct mmc_test_mem *mem; > struct scatterlist *sg; > + struct scatterlist *sg_areq; > }; > > /** > @@ -836,7 +838,9 @@ static int mmc_test_start_areq(struct mmc_test_card *test, > } > > static int mmc_test_nonblock_transfer(struct mmc_test_card *test, > - struct scatterlist *sg, unsigned sg_len, > + struct scatterlist *sg, > + struct scatterlist *sg_areq, > + unsigned int sg_len, > unsigned dev_addr, unsigned blocks, > unsigned blksz, int write, int count) > { > @@ -867,6 +871,7 @@ static int mmc_test_nonblock_transfer(struct mmc_test_card *test, > prev_mrq = &rq2->mrq; > > swap(mrq, prev_mrq); > + swap(sg, sg_areq); > dev_addr += blocks; > } > > @@ -1396,7 +1401,7 @@ static int mmc_test_no_highmem(struct mmc_test_card *test) > * Map sz bytes so that it can be transferred. > */ > static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz, > - int max_scatter, int min_sg_len) > + int max_scatter, int min_sg_len, bool nonblock) > { > struct mmc_test_area *t = &test->area; > int err; > @@ -1411,6 +1416,20 @@ static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz, > err = mmc_test_map_sg(t->mem, sz, t->sg, 1, t->max_segs, > t->max_seg_sz, &t->sg_len, min_sg_len); > } > + > + if (err || !nonblock) > + goto err; > + > + if (max_scatter) { > + err = mmc_test_map_sg_max_scatter(t->mem, sz, t->sg_areq, > + t->max_segs, t->max_seg_sz, > + &t->sg_len); > + } else { > + err = mmc_test_map_sg(t->mem, sz, t->sg_areq, 1, t->max_segs, > + t->max_seg_sz, &t->sg_len, min_sg_len); > + } > + > +err: > if (err) > pr_info("%s: Failed to map sg list\n", > mmc_hostname(test->card->host)); > @@ -1458,15 +1477,16 @@ static int mmc_test_area_io_seq(struct mmc_test_card *test, unsigned long sz, > sz = max_tfr; > } > > - ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len); > + ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len, nonblock); > if (ret) > return ret; > > if (timed) > ktime_get_ts64(&ts1); > if (nonblock) > - ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_len, > - dev_addr, t->blocks, 512, write, count); > + ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_areq, > + t->sg_len, dev_addr, t->blocks, 512, write, > + count); > else > for (i = 0; i < count && ret == 0; i++) { > ret = mmc_test_area_transfer(test, dev_addr, write); > @@ -1584,6 +1604,12 @@ static int mmc_test_area_init(struct mmc_test_card *test, int erase, int fill) > goto out_free; > } > > + t->sg_areq = kmalloc_array(t->max_segs, sizeof(*t->sg), GFP_KERNEL); > + if (!t->sg_areq) { > + ret = -ENOMEM; > + goto out_free; > + } > + > t->dev_addr = mmc_test_capacity(test->card) / 2; > t->dev_addr -= t->dev_addr % (t->max_sz >> 9); > > @@ -2468,7 +2494,7 @@ static int __mmc_test_cmds_during_tfr(struct mmc_test_card *test, > if (!(test->card->host->caps & MMC_CAP_CMD_DURING_TFR)) > return RESULT_UNSUP_HOST; > > - ret = mmc_test_area_map(test, sz, 0, 0); > + ret = mmc_test_area_map(test, sz, 0, 0, use_areq); > if (ret) > return ret; > [ 57.314539] mmc1: Test case 42. Write performance non-blocking req 1 to 512 sg elems... [ 57.343340] arm-smmu 15000000.iommu: Unhandled context fault: fsr=0x402, iova=0xfffe00600, fsynr=0x7d0003, cbfrsynra=0x80, cb=2 [ 57.355270] mmc1: ADMA error: 0x02000000 [ 57.359339] mmc1: sdhci: ============ SDHCI REGISTER DUMP =========== [ 57.365990] mmc1: sdhci: Sys addr: 0x00000008 | Version: 0x00007202 [ 57.372628] mmc1: sdhci: Blk size: 0x00000200 | Blk cnt: 0x00000280 [ 57.379268] mmc1: sdhci: Argument: 0x03b80400 | Trn mode: 0x00000023 [ 57.385904] mmc1: sdhci: Present: 0x03f801f6 | Host ctl: 0x0000001f [ 57.392551] mmc1: sdhci: Power: 0x00000001 | Blk gap: 0x00000000 [ 57.399198] mmc1: sdhci: Wake-up: 0x00000000 | Clock: 0x00000007 [ 57.405849] mmc1: sdhci: Timeout: 0x0000000e | Int stat: 0x00000000 [ 57.412501] mmc1: sdhci: Int enab: 0x03ff100b | Sig enab: 0x03ff100b [ 57.419154] mmc1: sdhci: ACmd stat: 0x00000000 | Slot int: 0x00000000 [ 57.425801] mmc1: sdhci: Caps: 0x322d32b2 | Caps_1: 0x0000808f [ 57.432463] mmc1: sdhci: Cmd: 0x0000193a | Max curr: 0x00000000 [ 57.439119] mmc1: sdhci: Resp[0]: 0x00000900 | Resp[1]: 0x00000000 [ 57.445766] mmc1: sdhci: Resp[2]: 0x00000000 | Resp[3]: 0x00000900 [ 57.452426] mmc1: sdhci: Host ctl2: 0x0000000b [ 57.457050] mmc1: sdhci: ADMA Err: 0x00000001 | ADMA Ptr: 0x0000000ffffff248 [ 57.464423] mmc1: sdhci: ============================================ [ 57.471101] mmc1: sdhci: ffffff200: DMA 0x0000000fffe00000, LEN 0x0000, Attr=0x21 [ 57.478853] mmc1: sdhci: ffffff20c: DMA 0x0000000fffe10000, LEN 0x0000, Attr=0x21 [ 57.486596] mmc1: sdhci: ffffff218: DMA 0x0000000fffe20000, LEN 0x0000, Attr=0x21 [ 57.494339] mmc1: sdhci: ffffff224: DMA 0x0000000fffe30000, LEN 0x0000, Attr=0x21 [ 57.502082] mmc1: sdhci: ffffff230: DMA 0x0000000fffe40000, LEN 0x0000, Attr=0x21 [ 57.509829] mmc1: sdhci: ffffff23c: DMA 0x0000000fffe50000, LEN 0x0000, Attr=0x21 [ 57.517573] mmc1: sdhci: ffffff248: DMA 0x0000000fffe60000, LEN 0x0000, Attr=0x21 [ 57.525314] mmc1: sdhci: ffffff254: DMA 0x0000000fffe70000, LEN 0x0000, Attr=0x21 [ 57.533061] mmc1: sdhci: ffffff260: DMA 0x0000000000000000, LEN 0x0000, Attr=0x03 <snip>... [ 57.607761] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT) [ 57.614490] pstate: 80400009 (Nzcv daif +PAN -UAO) [ 57.619438] pc : __arm_lpae_unmap+0x470/0x478 [ 57.623937] lr : __arm_lpae_unmap+0x2e0/0x478 [ 57.628429] sp : ffffffc014b4b710 [ 57.631862] x29: ffffffc014b4b740 x28: ffffffc014b4b9c0 [ 57.637339] x27: 0000000fffe00000 x26: ffffff8173f73000 [ 57.642818] x25: ffffffc010e184db x24: 0000000000000003 [ 57.648297] x23: 0000000000000000 x22: 0000000000000000 [ 57.653773] x21: 0000000fffe00000 x20: ffffff8174162400 [ 57.659248] x19: 0000000000000000 x18: 000000000012aa44 [ 57.664725] x17: 0000000000000020 x16: 0000000000000001 [ 57.670206] x15: 0000000000000010 x14: 0000000000000010 [ 57.675685] x13: 000000000000003f x12: 00000000ffffffff [ 57.681168] x11: 00000000000001ff x10: 0000000000000000 [ 57.686649] x9 : 0000000000000009 x8 : 000000000000000c [ 57.692122] x7 : 0000000000000000 x6 : 000000000000003f [ 57.697613] x5 : ffffff8173f73000 x4 : 0000000000000003 [ 57.703100] x3 : 0000000000001000 x2 : 0000000fffe00000 [ 57.708581] x1 : ffffffc014b4b9c0 x0 : ffffff8174162400 [ 57.714067] Call trace: [ 57.716620] __arm_lpae_unmap+0x470/0x478 [ 57.720770] __arm_lpae_unmap+0x2e0/0x478 [ 57.724916] __arm_lpae_unmap+0x2e0/0x478 [ 57.729059] arm_lpae_unmap+0x54/0x70 [ 57.732857] arm_smmu_unmap+0x64/0xa4 [ 57.736658] __iommu_unmap+0xb8/0x1f0 [ 57.740447] iommu_unmap_fast+0x38/0x48 [ 57.744413] __iommu_dma_unmap+0x88/0x108 [ 57.748566] iommu_dma_unmap_sg+0x90/0xa4 [ 57.752732] sdhci_post_req+0x5c/0x78 [ 57.756598] mmc_test_start_areq+0x10c/0x120 [mmc_test] [ 57.761989] mmc_test_area_io_seq+0x150/0x264 [mmc_test] [ 57.767484] mmc_test_rw_multiple+0x174/0x1c0 [mmc_test] [ 57.772976] mmc_test_rw_multiple_sg_len+0x44/0x6c [mmc_test] [ 57.778913] mmc_test_profile_sglen_wr_nonblock_perf+0x6c/0x94 [mmc_test] [ 57.785923] mtf_test_write+0x238/0x3cc [mmc_test] [ 57.790899] full_proxy_write+0x6c/0xa8 [ 57.794880] __vfs_write+0x54/0x18c [ 57.798496] vfs_write+0xe4/0x1a4 [ 57.801934] ksys_write+0x7c/0xe4 [ 57.805372] __arm64_sys_write+0x20/0x2c [ 57.809452] el0_svc_common+0xa4/0x154 [ 57.813329] el0_svc_compat_handler+0x2c/0x38 [ 57.817845] el0_svc_compat+0x8/0x10
On 11/02/20 11:36 am, Veerabhadrarao Badiganti wrote: > Supply a separate sg list for each of the request in non-blocking > IO test cases where two requests will be issued at same time. > > Otherwise, sg memory may get unmapped when a request is done while > same memory is being accessed by controller from the other request, > and it leads to iommu errors with below call stack: > > __arm_lpae_unmap+0x2e0/0x478 > arm_lpae_unmap+0x54/0x70 > arm_smmu_unmap+0x64/0xa4 > __iommu_unmap+0xb8/0x1f0 > iommu_unmap_fast+0x38/0x48 > __iommu_dma_unmap+0x88/0x108 > iommu_dma_unmap_sg+0x90/0xa4 > sdhci_post_req+0x5c/0x78 > mmc_test_start_areq+0x10c/0x120 [mmc_test] > mmc_test_area_io_seq+0x150/0x264 [mmc_test] > mmc_test_rw_multiple+0x174/0x1c0 [mmc_test] > mmc_test_rw_multiple_sg_len+0x44/0x6c [mmc_test] > mmc_test_profile_sglen_wr_nonblock_perf+0x6c/0x94 [mmc_test] > mtf_test_write+0x238/0x3cc [mmc_test] > > Signed-off-by: Veerabhadrarao Badiganti <vbadigan@codeaurora.org> > --- > drivers/mmc/core/mmc_test.c | 38 ++++++++++++++++++++++++++++++++------ > 1 file changed, 32 insertions(+), 6 deletions(-) > > diff --git a/drivers/mmc/core/mmc_test.c b/drivers/mmc/core/mmc_test.c > index 492dd45..69bdf60 100644 > --- a/drivers/mmc/core/mmc_test.c > +++ b/drivers/mmc/core/mmc_test.c > @@ -71,6 +71,7 @@ struct mmc_test_mem { > * @sg_len: length of currently mapped scatterlist @sg > * @mem: allocated memory > * @sg: scatterlist > + * @sg_areq: scatterlist for non blocking request > */ > struct mmc_test_area { > unsigned long max_sz; > @@ -82,6 +83,7 @@ struct mmc_test_area { > unsigned int sg_len; > struct mmc_test_mem *mem; > struct scatterlist *sg; > + struct scatterlist *sg_areq; > }; > > /** > @@ -836,7 +838,9 @@ static int mmc_test_start_areq(struct mmc_test_card *test, > } > > static int mmc_test_nonblock_transfer(struct mmc_test_card *test, > - struct scatterlist *sg, unsigned sg_len, > + struct scatterlist *sg, > + struct scatterlist *sg_areq, > + unsigned int sg_len, > unsigned dev_addr, unsigned blocks, > unsigned blksz, int write, int count) > { > @@ -867,6 +871,7 @@ static int mmc_test_nonblock_transfer(struct mmc_test_card *test, > prev_mrq = &rq2->mrq; > > swap(mrq, prev_mrq); > + swap(sg, sg_areq); > dev_addr += blocks; > } > > @@ -1396,7 +1401,7 @@ static int mmc_test_no_highmem(struct mmc_test_card *test) > * Map sz bytes so that it can be transferred. > */ > static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz, > - int max_scatter, int min_sg_len) > + int max_scatter, int min_sg_len, bool nonblock) > { > struct mmc_test_area *t = &test->area; > int err; > @@ -1411,6 +1416,20 @@ static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz, > err = mmc_test_map_sg(t->mem, sz, t->sg, 1, t->max_segs, > t->max_seg_sz, &t->sg_len, min_sg_len); > } > + > + if (err || !nonblock) > + goto err; > + > + if (max_scatter) { > + err = mmc_test_map_sg_max_scatter(t->mem, sz, t->sg_areq, > + t->max_segs, t->max_seg_sz, > + &t->sg_len); > + } else { > + err = mmc_test_map_sg(t->mem, sz, t->sg_areq, 1, t->max_segs, > + t->max_seg_sz, &t->sg_len, min_sg_len); > + } I suggest getting the length to a local and confirming it is the same + if (max_scatter) { + err = mmc_test_map_sg_max_scatter(t->mem, sz, t->sg_areq, + t->max_segs, t->max_seg_sz, + &sg_len); + } else { + err = mmc_test_map_sg(t->mem, sz, t->sg_areq, 1, t->max_segs, + t->max_seg_sz, &sg_len, min_sg_len); + } + if (!err && sg_len != t->sg_len) + err = -EINVAL; > + > +err: > if (err) > pr_info("%s: Failed to map sg list\n", > mmc_hostname(test->card->host)); > @@ -1458,15 +1477,16 @@ static int mmc_test_area_io_seq(struct mmc_test_card *test, unsigned long sz, > sz = max_tfr; > } > > - ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len); > + ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len, nonblock); > if (ret) > return ret; > > if (timed) > ktime_get_ts64(&ts1); > if (nonblock) > - ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_len, > - dev_addr, t->blocks, 512, write, count); > + ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_areq, > + t->sg_len, dev_addr, t->blocks, 512, write, > + count); > else > for (i = 0; i < count && ret == 0; i++) { > ret = mmc_test_area_transfer(test, dev_addr, write); > @@ -1584,6 +1604,12 @@ static int mmc_test_area_init(struct mmc_test_card *test, int erase, int fill) > goto out_free; > } > > + t->sg_areq = kmalloc_array(t->max_segs, sizeof(*t->sg), GFP_KERNEL); Needs to be freed in mmc_test_area_cleanup() > + if (!t->sg_areq) { > + ret = -ENOMEM; > + goto out_free; > + } > + > t->dev_addr = mmc_test_capacity(test->card) / 2; > t->dev_addr -= t->dev_addr % (t->max_sz >> 9); > > @@ -2468,7 +2494,7 @@ static int __mmc_test_cmds_during_tfr(struct mmc_test_card *test, > if (!(test->card->host->caps & MMC_CAP_CMD_DURING_TFR)) > return RESULT_UNSUP_HOST; > > - ret = mmc_test_area_map(test, sz, 0, 0); > + ret = mmc_test_area_map(test, sz, 0, 0, use_areq); > if (ret) > return ret; > >
diff --git a/drivers/mmc/core/mmc_test.c b/drivers/mmc/core/mmc_test.c index 492dd45..69bdf60 100644 --- a/drivers/mmc/core/mmc_test.c +++ b/drivers/mmc/core/mmc_test.c @@ -71,6 +71,7 @@ struct mmc_test_mem { * @sg_len: length of currently mapped scatterlist @sg * @mem: allocated memory * @sg: scatterlist + * @sg_areq: scatterlist for non blocking request */ struct mmc_test_area { unsigned long max_sz; @@ -82,6 +83,7 @@ struct mmc_test_area { unsigned int sg_len; struct mmc_test_mem *mem; struct scatterlist *sg; + struct scatterlist *sg_areq; }; /** @@ -836,7 +838,9 @@ static int mmc_test_start_areq(struct mmc_test_card *test, } static int mmc_test_nonblock_transfer(struct mmc_test_card *test, - struct scatterlist *sg, unsigned sg_len, + struct scatterlist *sg, + struct scatterlist *sg_areq, + unsigned int sg_len, unsigned dev_addr, unsigned blocks, unsigned blksz, int write, int count) { @@ -867,6 +871,7 @@ static int mmc_test_nonblock_transfer(struct mmc_test_card *test, prev_mrq = &rq2->mrq; swap(mrq, prev_mrq); + swap(sg, sg_areq); dev_addr += blocks; } @@ -1396,7 +1401,7 @@ static int mmc_test_no_highmem(struct mmc_test_card *test) * Map sz bytes so that it can be transferred. */ static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz, - int max_scatter, int min_sg_len) + int max_scatter, int min_sg_len, bool nonblock) { struct mmc_test_area *t = &test->area; int err; @@ -1411,6 +1416,20 @@ static int mmc_test_area_map(struct mmc_test_card *test, unsigned long sz, err = mmc_test_map_sg(t->mem, sz, t->sg, 1, t->max_segs, t->max_seg_sz, &t->sg_len, min_sg_len); } + + if (err || !nonblock) + goto err; + + if (max_scatter) { + err = mmc_test_map_sg_max_scatter(t->mem, sz, t->sg_areq, + t->max_segs, t->max_seg_sz, + &t->sg_len); + } else { + err = mmc_test_map_sg(t->mem, sz, t->sg_areq, 1, t->max_segs, + t->max_seg_sz, &t->sg_len, min_sg_len); + } + +err: if (err) pr_info("%s: Failed to map sg list\n", mmc_hostname(test->card->host)); @@ -1458,15 +1477,16 @@ static int mmc_test_area_io_seq(struct mmc_test_card *test, unsigned long sz, sz = max_tfr; } - ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len); + ret = mmc_test_area_map(test, sz, max_scatter, min_sg_len, nonblock); if (ret) return ret; if (timed) ktime_get_ts64(&ts1); if (nonblock) - ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_len, - dev_addr, t->blocks, 512, write, count); + ret = mmc_test_nonblock_transfer(test, t->sg, t->sg_areq, + t->sg_len, dev_addr, t->blocks, 512, write, + count); else for (i = 0; i < count && ret == 0; i++) { ret = mmc_test_area_transfer(test, dev_addr, write); @@ -1584,6 +1604,12 @@ static int mmc_test_area_init(struct mmc_test_card *test, int erase, int fill) goto out_free; } + t->sg_areq = kmalloc_array(t->max_segs, sizeof(*t->sg), GFP_KERNEL); + if (!t->sg_areq) { + ret = -ENOMEM; + goto out_free; + } + t->dev_addr = mmc_test_capacity(test->card) / 2; t->dev_addr -= t->dev_addr % (t->max_sz >> 9); @@ -2468,7 +2494,7 @@ static int __mmc_test_cmds_during_tfr(struct mmc_test_card *test, if (!(test->card->host->caps & MMC_CAP_CMD_DURING_TFR)) return RESULT_UNSUP_HOST; - ret = mmc_test_area_map(test, sz, 0, 0); + ret = mmc_test_area_map(test, sz, 0, 0, use_areq); if (ret) return ret;
Supply a separate sg list for each of the request in non-blocking IO test cases where two requests will be issued at same time. Otherwise, sg memory may get unmapped when a request is done while same memory is being accessed by controller from the other request, and it leads to iommu errors with below call stack: __arm_lpae_unmap+0x2e0/0x478 arm_lpae_unmap+0x54/0x70 arm_smmu_unmap+0x64/0xa4 __iommu_unmap+0xb8/0x1f0 iommu_unmap_fast+0x38/0x48 __iommu_dma_unmap+0x88/0x108 iommu_dma_unmap_sg+0x90/0xa4 sdhci_post_req+0x5c/0x78 mmc_test_start_areq+0x10c/0x120 [mmc_test] mmc_test_area_io_seq+0x150/0x264 [mmc_test] mmc_test_rw_multiple+0x174/0x1c0 [mmc_test] mmc_test_rw_multiple_sg_len+0x44/0x6c [mmc_test] mmc_test_profile_sglen_wr_nonblock_perf+0x6c/0x94 [mmc_test] mtf_test_write+0x238/0x3cc [mmc_test] Signed-off-by: Veerabhadrarao Badiganti <vbadigan@codeaurora.org> --- drivers/mmc/core/mmc_test.c | 38 ++++++++++++++++++++++++++++++++------ 1 file changed, 32 insertions(+), 6 deletions(-)