Skip to contentSkip to conversation

ASoC: topology: Propagate error appropriately

PATCH In review Amadeusz Sławiński 5 participants 24 messages Mar 27, 2020 – Apr 9, 2020

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

v2 cover ASoC: topology: Propagate error appropriately

v1:
Check if kstrdup succeeded.

v2:
Remove unneeded freeing, which is performed in another place by dobj handlers.

Additionally for functions which have return status which was ignored, perform success checks and handle failures in appropriate way.

Amadeusz Sławiński (6):
ASoC: topology: Add missing memory checks
ASoC: topology: Check return value of soc_tplg_create_tlv
ASoC: topology: Check return value of soc_tplg_*_create
ASoC: topology: Check soc_tplg_add_route return value
ASoC: topology: Check return value of pcm_new_ver
ASoC: topology: Check return value of soc_tplg_dai_config

 sound/soc/soc-topology.c | 113 ++++++++++++++++++++++++++++++---------
 1 file changed, 88 insertions(+), 25 deletions(-)
Signature · 1 line
2.17.1
#1 lore raw

PATCH v2 1/6 ASoC: topology: Add missing memory checks

kstrdup is an allocation function and it can fail, so its return value should be checked and handled appropriately.

In order to check all cases, we need to modify set_stream_info to return a value, so check that everything went correctly when doing kstrdup(). Later add proper checks and error handlers.

---

v2:
Remove unneeded freeing, which is performed in another place by dobj handlers.

 sound/soc/soc-topology.c | 62 +++++++++++++++++++++++++++++++---------
 1 file changed, 49 insertions(+), 13 deletions(-)
sound/soc/soc-topology.c +49 -13
index 1f81cd2d29cf..434e152ac8ee 100644
@@ -1766,10 +1766,13 @@ static int soc_tplg_dapm_complete(struct soc_tplg *tplg)
 	return 0;
 }
 
-static void set_stream_info(struct snd_soc_pcm_stream *stream,
+static int set_stream_info(struct snd_soc_pcm_stream *stream,
 	struct snd_soc_tplg_stream_caps *caps)
 {
 	stream->stream_name = kstrdup(caps->name, GFP_KERNEL);
+	if (!stream->stream_name)
+		return -ENOMEM;
+
 	stream->channels_min = le32_to_cpu(caps->channels_min);
 	stream->channels_max = le32_to_cpu(caps->channels_max);
 	stream->rates = le32_to_cpu(caps->rates);
@@ -1777,6 +1780,8 @@ static void set_stream_info(struct snd_soc_pcm_stream *stream,
 	stream->rate_max = le32_to_cpu(caps->rate_max);
 	stream->formats = le64_to_cpu(caps->formats);
 	stream->sig_bits = le32_to_cpu(caps->sig_bits);
+
+	return 0;
 }
 
 static void set_dai_flags(struct snd_soc_dai_driver *dai_drv,
@@ -1812,20 +1817,29 @@ static int soc_tplg_dai_create(struct soc_tplg *tplg,
 	if (dai_drv == NULL)
 		return -ENOMEM;
 
-	if (strlen(pcm->dai_name))
+	if (strlen(pcm->dai_name)) {
 		dai_drv->name = kstrdup(pcm->dai_name, GFP_KERNEL);
+		if (!dai_drv->name) {
+			ret = -ENOMEM;
+			goto err;
+		}
+	}
 	dai_drv->id = le32_to_cpu(pcm->dai_id);
 
 	if (pcm->playback) {
 		stream = &dai_drv->playback;
 		caps = &pcm->caps[SND_SOC_TPLG_STREAM_PLAYBACK];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (pcm->capture) {
 		stream = &dai_drv->capture;
 		caps = &pcm->caps[SND_SOC_TPLG_STREAM_CAPTURE];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (pcm->compress)
@@ -1835,11 +1849,7 @@ static int soc_tplg_dai_create(struct soc_tplg *tplg,
 	ret = soc_tplg_dai_load(tplg, dai_drv, pcm, NULL);
 	if (ret < 0) {
 		dev_err(tplg->comp->dev, "ASoC: DAI loading failed\n");
-		kfree(dai_drv->playback.stream_name);
-		kfree(dai_drv->capture.stream_name);
-		kfree(dai_drv->name);
-		kfree(dai_drv);
-		return ret;
+		goto err;
 	}
 
 	dai_drv->dobj.index = tplg->index;
@@ -1860,6 +1870,14 @@ static int soc_tplg_dai_create(struct soc_tplg *tplg,
 		return ret;
 	}
 
+	return 0;
+
+err:
+	kfree(dai_drv->playback.stream_name);
+	kfree(dai_drv->capture.stream_name);
+	kfree(dai_drv->name);
+	kfree(dai_drv);
+
 	return ret;
 }
 
@@ -1916,11 +1934,20 @@ static int soc_tplg_fe_link_create(struct soc_tplg *tplg,
 	if (strlen(pcm->pcm_name)) {
 		link->name = kstrdup(pcm->pcm_name, GFP_KERNEL);
 		link->stream_name = kstrdup(pcm->pcm_name, GFP_KERNEL);
+		if (!link->name || !link->stream_name) {
+			ret = -ENOMEM;
+			goto err;
+		}
 	}
 	link->id = le32_to_cpu(pcm->pcm_id);
 
-	if (strlen(pcm->dai_name))
+	if (strlen(pcm->dai_name)) {
 		link->cpus->dai_name = kstrdup(pcm->dai_name, GFP_KERNEL);
+		if (!link->cpus->dai_name) {
+			ret = -ENOMEM;
+			goto err;
+		}
+	}
 
 	link->codecs->name = "snd-soc-dummy";
 	link->codecs->dai_name = "snd-soc-dummy-dai";
@@ -2436,13 +2463,17 @@ static int soc_tplg_dai_config(struct soc_tplg *tplg,
 	if (d->playback) {
 		stream = &dai_drv->playback;
 		caps = &d->caps[SND_SOC_TPLG_STREAM_PLAYBACK];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (d->capture) {
 		stream = &dai_drv->capture;
 		caps = &d->caps[SND_SOC_TPLG_STREAM_CAPTURE];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (d->flag_mask)
@@ -2454,10 +2485,15 @@ static int soc_tplg_dai_config(struct soc_tplg *tplg,
 	ret = soc_tplg_dai_load(tplg, dai_drv, NULL, dai);
 	if (ret < 0) {
 		dev_err(tplg->comp->dev, "ASoC: DAI loading failed\n");
-		return ret;
+		goto err;
 	}
 
 	return 0;
+
+err:
+	kfree(dai_drv->playback.stream_name);
+	kfree(dai_drv->capture.stream_name);
+	return ret;
 }
 
 /* load physical DAI elements */
Signature · 1 line
2.17.1
maintainer
#2 lore raw

Applied "ASoC: topology: Add missing memory checks" to the asoc tree

The patch

ASoC: topology: Add missing memory checks

has been applied to the asoc tree at

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

From abc3caac24501008465fdb55c5e89e16d58d5a3d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?=
<[email protected]>
Date: Fri, 27 Mar 2020 16:47:24 -0400
Subject: [PATCH] ASoC: topology: Add missing memory checks
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

kstrdup is an allocation function and it can fail, so its return value should be checked and handled appropriately.

In order to check all cases, we need to modify set_stream_info to return a value, so check that everything went correctly when doing kstrdup(). Later add proper checks and error handlers.

---
 sound/soc/soc-topology.c | 62 +++++++++++++++++++++++++++++++---------
 1 file changed, 49 insertions(+), 13 deletions(-)
sound/soc/soc-topology.c +49 -13
index 87f75edba3dc..73fc304c9aca 100644
@@ -1766,10 +1766,13 @@ static int soc_tplg_dapm_complete(struct soc_tplg *tplg)
 	return 0;
 }
 
-static void set_stream_info(struct snd_soc_pcm_stream *stream,
+static int set_stream_info(struct snd_soc_pcm_stream *stream,
 	struct snd_soc_tplg_stream_caps *caps)
 {
 	stream->stream_name = kstrdup(caps->name, GFP_KERNEL);
+	if (!stream->stream_name)
+		return -ENOMEM;
+
 	stream->channels_min = le32_to_cpu(caps->channels_min);
 	stream->channels_max = le32_to_cpu(caps->channels_max);
 	stream->rates = le32_to_cpu(caps->rates);
@@ -1777,6 +1780,8 @@ static void set_stream_info(struct snd_soc_pcm_stream *stream,
 	stream->rate_max = le32_to_cpu(caps->rate_max);
 	stream->formats = le64_to_cpu(caps->formats);
 	stream->sig_bits = le32_to_cpu(caps->sig_bits);
+
+	return 0;
 }
 
 static void set_dai_flags(struct snd_soc_dai_driver *dai_drv,
@@ -1812,20 +1817,29 @@ static int soc_tplg_dai_create(struct soc_tplg *tplg,
 	if (dai_drv == NULL)
 		return -ENOMEM;
 
-	if (strlen(pcm->dai_name))
+	if (strlen(pcm->dai_name)) {
 		dai_drv->name = kstrdup(pcm->dai_name, GFP_KERNEL);
+		if (!dai_drv->name) {
+			ret = -ENOMEM;
+			goto err;
+		}
+	}
 	dai_drv->id = le32_to_cpu(pcm->dai_id);
 
 	if (pcm->playback) {
 		stream = &dai_drv->playback;
 		caps = &pcm->caps[SND_SOC_TPLG_STREAM_PLAYBACK];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (pcm->capture) {
 		stream = &dai_drv->capture;
 		caps = &pcm->caps[SND_SOC_TPLG_STREAM_CAPTURE];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (pcm->compress)
@@ -1835,11 +1849,7 @@ static int soc_tplg_dai_create(struct soc_tplg *tplg,
 	ret = soc_tplg_dai_load(tplg, dai_drv, pcm, NULL);
 	if (ret < 0) {
 		dev_err(tplg->comp->dev, "ASoC: DAI loading failed\n");
-		kfree(dai_drv->playback.stream_name);
-		kfree(dai_drv->capture.stream_name);
-		kfree(dai_drv->name);
-		kfree(dai_drv);
-		return ret;
+		goto err;
 	}
 
 	dai_drv->dobj.index = tplg->index;
@@ -1860,6 +1870,14 @@ static int soc_tplg_dai_create(struct soc_tplg *tplg,
 		return ret;
 	}
 
+	return 0;
+
+err:
+	kfree(dai_drv->playback.stream_name);
+	kfree(dai_drv->capture.stream_name);
+	kfree(dai_drv->name);
+	kfree(dai_drv);
+
 	return ret;
 }
 
@@ -1916,11 +1934,20 @@ static int soc_tplg_fe_link_create(struct soc_tplg *tplg,
 	if (strlen(pcm->pcm_name)) {
 		link->name = kstrdup(pcm->pcm_name, GFP_KERNEL);
 		link->stream_name = kstrdup(pcm->pcm_name, GFP_KERNEL);
+		if (!link->name || !link->stream_name) {
+			ret = -ENOMEM;
+			goto err;
+		}
 	}
 	link->id = le32_to_cpu(pcm->pcm_id);
 
-	if (strlen(pcm->dai_name))
+	if (strlen(pcm->dai_name)) {
 		link->cpus->dai_name = kstrdup(pcm->dai_name, GFP_KERNEL);
+		if (!link->cpus->dai_name) {
+			ret = -ENOMEM;
+			goto err;
+		}
+	}
 
 	link->codecs->name = "snd-soc-dummy";
 	link->codecs->dai_name = "snd-soc-dummy-dai";
@@ -2436,13 +2463,17 @@ static int soc_tplg_dai_config(struct soc_tplg *tplg,
 	if (d->playback) {
 		stream = &dai_drv->playback;
 		caps = &d->caps[SND_SOC_TPLG_STREAM_PLAYBACK];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (d->capture) {
 		stream = &dai_drv->capture;
 		caps = &d->caps[SND_SOC_TPLG_STREAM_CAPTURE];
-		set_stream_info(stream, caps);
+		ret = set_stream_info(stream, caps);
+		if (ret < 0)
+			goto err;
 	}
 
 	if (d->flag_mask)
@@ -2454,10 +2485,15 @@ static int soc_tplg_dai_config(struct soc_tplg *tplg,
 	ret = soc_tplg_dai_load(tplg, dai_drv, NULL, dai);
 	if (ret < 0) {
 		dev_err(tplg->comp->dev, "ASoC: DAI loading failed\n");
-		return ret;
+		goto err;
 	}
 
 	return 0;
+
+err:
+	kfree(dai_drv->playback.stream_name);
+	kfree(dai_drv->capture.stream_name);
+	return ret;
 }
 
 /* load physical DAI elements */
Signature · 1 line
2.20.1
#3 lore raw

PATCH v2 2/6 ASoC: topology: Check return value of soc_tplg_create_tlv

In reply to Amadeusz Sławiński

Function soc_tplg_create_tlv can fail, so we should check if it succeded or not and proceed appropriately.

---

v2:
Added this patch

 sound/soc/soc-topology.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)
sound/soc/soc-topology.c +14 -2
index 434e152ac8ee..c3811dd66b68 100644
@@ -894,7 +894,13 @@ static int soc_tplg_dmixer_create(struct soc_tplg *tplg, unsigned int count,
 		}
 
 		/* create any TLV data */
-		soc_tplg_create_tlv(tplg, &kc, &mc->hdr);
+		err = soc_tplg_create_tlv(tplg, &kc, &mc->hdr);
+		if (err < 0) {
+			dev_err(tplg->dev, "ASoC: failed to create TLV %s\n",
+				mc->hdr.name);
+			kfree(sm);
+			continue;
+		}
 
 		/* pass control to driver for optional further init */
 		err = soc_tplg_init_kcontrol(tplg, &kc,
@@ -1355,7 +1361,13 @@ static struct snd_kcontrol_new *soc_tplg_dapm_widget_dmixer_create(
 		}
 
 		/* create any TLV data */
-		soc_tplg_create_tlv(tplg, &kc[i], &mc->hdr);
+		err = soc_tplg_create_tlv(tplg, &kc[i], &mc->hdr);
+		if (err < 0) {
+			dev_err(tplg->dev, "ASoC: failed to create TLV %s\n",
+				mc->hdr.name);
+			kfree(sm);
+			continue;
+		}
 
 		/* pass control to driver for optional further init */
 		err = soc_tplg_init_kcontrol(tplg, &kc[i],
Signature · 1 line
2.17.1
maintainer
#4 lore raw

Applied "ASoC: topology: Check return value of soc_tplg_create_tlv" to the asoc tree

The patch

ASoC: topology: Check return value of soc_tplg_create_tlv

has been applied to the asoc tree at

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

From 482db55ae87f3749db05810a38b1d618dfd4407c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?=
<[email protected]>
Date: Fri, 27 Mar 2020 16:47:25 -0400
Subject: [PATCH] ASoC: topology: Check return value of soc_tplg_create_tlv
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Function soc_tplg_create_tlv can fail, so we should check if it succeded or not and proceed appropriately.

---
 sound/soc/soc-topology.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)
sound/soc/soc-topology.c +14 -2
index 73fc304c9aca..f37a72aebb5a 100644
@@ -894,7 +894,13 @@ static int soc_tplg_dmixer_create(struct soc_tplg *tplg, unsigned int count,
 		}
 
 		/* create any TLV data */
-		soc_tplg_create_tlv(tplg, &kc, &mc->hdr);
+		err = soc_tplg_create_tlv(tplg, &kc, &mc->hdr);
+		if (err < 0) {
+			dev_err(tplg->dev, "ASoC: failed to create TLV %s\n",
+				mc->hdr.name);
+			kfree(sm);
+			continue;
+		}
 
 		/* pass control to driver for optional further init */
 		err = soc_tplg_init_kcontrol(tplg, &kc,
@@ -1355,7 +1361,13 @@ static struct snd_kcontrol_new *soc_tplg_dapm_widget_dmixer_create(
 		}
 
 		/* create any TLV data */
-		soc_tplg_create_tlv(tplg, &kc[i], &mc->hdr);
+		err = soc_tplg_create_tlv(tplg, &kc[i], &mc->hdr);
+		if (err < 0) {
+			dev_err(tplg->dev, "ASoC: failed to create TLV %s\n",
+				mc->hdr.name);
+			kfree(sm);
+			continue;
+		}
 
 		/* pass control to driver for optional further init */
 		err = soc_tplg_init_kcontrol(tplg, &kc[i],
Signature · 1 line
2.20.1
#5 lore raw

PATCH v2 3/6 ASoC: topology: Check return value of soc_tplg_*_create

In reply to Amadeusz Sławiński

Functions soc_tplg_denum_create, soc_tplg_dmixer_create, soc_tplg_dbytes_create can fail, so their return values should be checked and error should be propagated.

---

v2:
Added this patch

 sound/soc/soc-topology.c | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)
sound/soc/soc-topology.c +12 -6
index c3811dd66b68..860bced933d6 100644
@@ -1124,6 +1124,7 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,
 	struct snd_soc_tplg_hdr *hdr)
 {
 	struct snd_soc_tplg_ctl_hdr *control_hdr;
+	int ret;
 	int i;
 
 	if (tplg->pass != SOC_TPLG_PASS_MIXER) {
@@ -1152,25 +1153,30 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,
 		case SND_SOC_TPLG_CTL_RANGE:
 		case SND_SOC_TPLG_DAPM_CTL_VOLSW:
 		case SND_SOC_TPLG_DAPM_CTL_PIN:
-			soc_tplg_dmixer_create(tplg, 1,
-					       le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_dmixer_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
 			break;
 		case SND_SOC_TPLG_CTL_ENUM:
 		case SND_SOC_TPLG_CTL_ENUM_VALUE:
 		case SND_SOC_TPLG_DAPM_CTL_ENUM_DOUBLE:
 		case SND_SOC_TPLG_DAPM_CTL_ENUM_VIRT:
 		case SND_SOC_TPLG_DAPM_CTL_ENUM_VALUE:
-			soc_tplg_denum_create(tplg, 1,
-					      le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_denum_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
 			break;
 		case SND_SOC_TPLG_CTL_BYTES:
-			soc_tplg_dbytes_create(tplg, 1,
-					       le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_dbytes_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
 			break;
 		default:
 			soc_bind_err(tplg, control_hdr, i);
 			return -EINVAL;
 		}
+		if (ret < 0) {
+			dev_err(tplg->dev, "ASoC: invalid control\n");
+			return ret;
+		}
+
 	}
 
 	return 0;
Signature · 1 line
2.17.1
#6 lore raw

Re: [PATCH v2 3/6] ASoC: topology: Check return value of soc_tplg_*_create

Quoted 48 lines · Amadeusz Sławiński

Functions soc_tplg_denum_create, soc_tplg_dmixer_create, soc_tplg_dbytes_create can fail, so their return values should be checked and error should be propagated.

---

v2:
Added this patch

  sound/soc/soc-topology.c | 18 ++++++++++++------
  1 file changed, 12 insertions(+), 6 deletions(-)
sound/soc/soc-topology.c +6 -6
index c3811dd66b68..860bced933d6 100644
@@ -1124,6 +1124,7 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,
  	struct snd_soc_tplg_hdr *hdr)
  {
  	struct snd_soc_tplg_ctl_hdr *control_hdr;
+	int ret;
  	int i;
 
  	if (tplg->pass != SOC_TPLG_PASS_MIXER) {
@@ -1152,25 +1153,30 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,
  		case SND_SOC_TPLG_CTL_RANGE:
  		case SND_SOC_TPLG_DAPM_CTL_VOLSW:
  		case SND_SOC_TPLG_DAPM_CTL_PIN:
-			soc_tplg_dmixer_create(tplg, 1,
-					       le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_dmixer_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
  			break;
  		case SND_SOC_TPLG_CTL_ENUM:
  		case SND_SOC_TPLG_CTL_ENUM_VALUE:
  		case SND_SOC_TPLG_DAPM_CTL_ENUM_DOUBLE:
  		case SND_SOC_TPLG_DAPM_CTL_ENUM_VIRT:
  		case SND_SOC_TPLG_DAPM_CTL_ENUM_VALUE:
-			soc_tplg_denum_create(tplg, 1,
-					      le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_denum_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
  			break;
  		case SND_SOC_TPLG_CTL_BYTES:
-			soc_tplg_dbytes_create(tplg, 1,
-					       le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_dbytes_create(tplg, 1,

sound/soc/soc-topology.c static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg, · line 1178

Quoted 10 lines
sound/soc/soc-topology.c +5 -0
+					le32_to_cpu(hdr->payload_size));
  			break;
  		default:
  			soc_bind_err(tplg, control_hdr, i);
  			return -EINVAL;
  		}
+		if (ret < 0) {
+			dev_err(tplg->dev, "ASoC: invalid control\n");
+			return ret;
+		}

Sounds good, but this happens in a loop, so would all the memory previously allocated by denum/dbytes/dmixer_create leak, or is it freed automatically somewhere else?

Quoted 5 lines
sound/soc/soc-topology.c +1 -0
+
  	}
 
  	return 0;
#7 lore raw

Re: [PATCH v2 3/6] ASoC: topology: Check return value of soc_tplg_*_create

Quoted 56 lines · Pierre-Louis Bossart
Quoted 53 lines · Amadeusz Sławiński

Functions soc_tplg_denum_create, soc_tplg_dmixer_create, soc_tplg_dbytes_create can fail, so their return values should be checked and error should be propagated.

---

  v2:
   Added this patch

  sound/soc/soc-topology.c | 18 ++++++++++++------

  1 file changed, 12 insertions(+), 6 deletions(-)
sound/soc/soc-topology.c +1 -0
index c3811dd66b68..860bced933d6 100644
@@ -1124,6 +1124,7 @@ static int soc_tplg_kcontrol_elems_load(struct 
 soc_tplg *tplg,
       struct snd_soc_tplg_hdr *hdr)
   {
       struct snd_soc_tplg_ctl_hdr *control_hdr;
+    int ret;
       int i;
       if (tplg->pass != SOC_TPLG_PASS_MIXER) {
@@ -1152,25 +1153,30 @@ static int soc_tplg_kcontrol_elems_load(struct 
 soc_tplg *tplg,

          case SND_SOC_TPLG_CTL_RANGE:
          case SND_SOC_TPLG_DAPM_CTL_VOLSW:
          case SND_SOC_TPLG_DAPM_CTL_PIN:
-            soc_tplg_dmixer_create(tplg, 1,
-                           le32_to_cpu(hdr->payload_size)); +            ret = soc_tplg_dmixer_create(tplg, 1, +                    le32_to_cpu(hdr->payload_size)); break;
          case SND_SOC_TPLG_CTL_ENUM:
          case SND_SOC_TPLG_CTL_ENUM_VALUE:
          case SND_SOC_TPLG_DAPM_CTL_ENUM_DOUBLE:
          case SND_SOC_TPLG_DAPM_CTL_ENUM_VIRT:
          case SND_SOC_TPLG_DAPM_CTL_ENUM_VALUE:
-            soc_tplg_denum_create(tplg, 1,
-                          le32_to_cpu(hdr->payload_size)); +            ret = soc_tplg_denum_create(tplg, 1, +                    le32_to_cpu(hdr->payload_size)); break;
          case SND_SOC_TPLG_CTL_BYTES:
-            soc_tplg_dbytes_create(tplg, 1,
-                           le32_to_cpu(hdr->payload_size)); +            ret = soc_tplg_dbytes_create(tplg, 1, +                    le32_to_cpu(hdr->payload_size)); break;
          default:
              soc_bind_err(tplg, control_hdr, i);

Quoted 9 lines
Quoted 6 lines

              return -EINVAL;
          }
+        if (ret < 0) {
+            dev_err(tplg->dev, "ASoC: invalid control\n"); +            return ret;
+        }

Sounds good, but this happens in a loop, so would all the memory previously allocated by denum/dbytes/dmixer_create leak, or is it freed automatically somewhere else?

Well, now that error is propagated, snd_soc_tplg_component_remove() should be called by snd_soc_tplg_component_load() in case of errors while parsing. From quick look it seems like it should be able to free it up correctly by calling remove_enum/bytes/mixer.

Thanks,
Amadeusz

#8 lore raw

Re: [PATCH v2 3/6] ASoC: topology: Check return value of soc_tplg_*_create

Quoted 40 lines
Quoted 30 lines
Quoted 30 lines

      if (tplg->pass != SOC_TPLG_PASS_MIXER) {

(quoted diff) +0 -0
@@ -1152,25 +1153,30 @@ static int 
 soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,

          case SND_SOC_TPLG_CTL_RANGE:
          case SND_SOC_TPLG_DAPM_CTL_VOLSW:
          case SND_SOC_TPLG_DAPM_CTL_PIN:
-            soc_tplg_dmixer_create(tplg, 1,
-                           le32_to_cpu(hdr->payload_size)); +            ret = soc_tplg_dmixer_create(tplg, 1, +                    le32_to_cpu(hdr->payload_size)); break;
          case SND_SOC_TPLG_CTL_ENUM:
          case SND_SOC_TPLG_CTL_ENUM_VALUE:
          case SND_SOC_TPLG_DAPM_CTL_ENUM_DOUBLE:
          case SND_SOC_TPLG_DAPM_CTL_ENUM_VIRT:
          case SND_SOC_TPLG_DAPM_CTL_ENUM_VALUE:
-            soc_tplg_denum_create(tplg, 1,
-                          le32_to_cpu(hdr->payload_size)); +            ret = soc_tplg_denum_create(tplg, 1, +                    le32_to_cpu(hdr->payload_size)); break;
          case SND_SOC_TPLG_CTL_BYTES:
-            soc_tplg_dbytes_create(tplg, 1,
-                           le32_to_cpu(hdr->payload_size)); +            ret = soc_tplg_dbytes_create(tplg, 1, +                    le32_to_cpu(hdr->payload_size)); break;
          default:
              soc_bind_err(tplg, control_hdr, i);

Quoted 9 lines
Quoted 6 lines

              return -EINVAL;
          }
+        if (ret < 0) {
+            dev_err(tplg->dev, "ASoC: invalid control\n"); +            return ret;
+        }

Sounds good, but this happens in a loop, so would all the memory previously allocated by denum/dbytes/dmixer_create leak, or is it freed automatically somewhere else?

Quoted 1 line

Well, now that error is propagated, snd_soc_tplg_component_remove() should be called by snd_soc_tplg_component_load() in case of errors while parsing. From quick look it seems like it should be able to free it up correctly by calling remove_enum/bytes/mixer.

I am not sure what you meant by 'should be called', if it's a recommendation for a future change or a description of the existing behavior. Just to be clear, are you saying the existing code will take care of this error flow or that a new patch is needed?

#9 lore raw

Re: [PATCH v2 3/6] ASoC: topology: Check return value of soc_tplg_*_create

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines

Sounds good, but this happens in a loop, so would all the memory previously allocated by denum/dbytes/dmixer_create leak, or is it freed automatically somewhere else?

Well, now that error is propagated, snd_soc_tplg_component_remove() should be called by snd_soc_tplg_component_load() in case of errors while parsing. From quick look it seems like it should be able to free it up correctly by calling remove_enum/bytes/mixer.

I am not sure what you meant by 'should be called', if it's a recommendation for a future change or a description of the existing behavior. Just to be clear, are you saying the existing code will take care of this error flow or that a new patch is needed?

Existing code should handle this properly.
No new code is needed.

#10 lore raw

Re: [PATCH v2 3/6] ASoC: topology: Check return value of soc_tplg_*_create

Quoted 11 lines · Amadeusz Sławiński
Quoted 8 lines
Quoted 5 lines
Quoted 3 lines

Sounds good, but this happens in a loop, so would all the memory previously allocated by denum/dbytes/dmixer_create leak, or is it freed automatically somewhere else?

Well, now that error is propagated, snd_soc_tplg_component_remove() should be called by snd_soc_tplg_component_load() in case of errors while parsing. From quick look it seems like it should be able to free it up correctly by calling remove_enum/bytes/mixer.

I am not sure what you meant by 'should be called', if it's a recommendation for a future change or a description of the existing behavior. Just to be clear, are you saying the existing code will take care of this error flow or that a new patch is needed?

Existing code should handle this properly.
No new code is needed.

Sounds good, thanks.

maintainer
#11 lore raw

Applied "ASoC: topology: Check return value of soc_tplg_*_create" to the asoc tree

In reply to Amadeusz Sławiński

The patch

ASoC: topology: Check return value of soc_tplg_*_create

has been applied to the asoc tree at

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

From 2ae548f30d7f6973388fc3769bb3c2f6fd13652b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?=
<[email protected]>
Date: Fri, 27 Mar 2020 16:47:26 -0400
Subject: [PATCH] ASoC: topology: Check return value of soc_tplg_*_create
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Functions soc_tplg_denum_create, soc_tplg_dmixer_create, soc_tplg_dbytes_create can fail, so their return values should be checked and error should be propagated.

---
 sound/soc/soc-topology.c | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)
sound/soc/soc-topology.c +12 -6
index f37a72aebb5a..3ada769cf823 100644
@@ -1124,6 +1124,7 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,
 	struct snd_soc_tplg_hdr *hdr)
 {
 	struct snd_soc_tplg_ctl_hdr *control_hdr;
+	int ret;
 	int i;
 
 	if (tplg->pass != SOC_TPLG_PASS_MIXER) {
@@ -1152,25 +1153,30 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg,
 		case SND_SOC_TPLG_CTL_RANGE:
 		case SND_SOC_TPLG_DAPM_CTL_VOLSW:
 		case SND_SOC_TPLG_DAPM_CTL_PIN:
-			soc_tplg_dmixer_create(tplg, 1,
-					       le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_dmixer_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
 			break;
 		case SND_SOC_TPLG_CTL_ENUM:
 		case SND_SOC_TPLG_CTL_ENUM_VALUE:
 		case SND_SOC_TPLG_DAPM_CTL_ENUM_DOUBLE:
 		case SND_SOC_TPLG_DAPM_CTL_ENUM_VIRT:
 		case SND_SOC_TPLG_DAPM_CTL_ENUM_VALUE:
-			soc_tplg_denum_create(tplg, 1,
-					      le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_denum_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
 			break;
 		case SND_SOC_TPLG_CTL_BYTES:
-			soc_tplg_dbytes_create(tplg, 1,
-					       le32_to_cpu(hdr->payload_size));
+			ret = soc_tplg_dbytes_create(tplg, 1,
+					le32_to_cpu(hdr->payload_size));
 			break;
 		default:
 			soc_bind_err(tplg, control_hdr, i);
 			return -EINVAL;
 		}
+		if (ret < 0) {
+			dev_err(tplg->dev, "ASoC: invalid control\n");
+			return ret;
+		}
+
 	}
 
 	return 0;
Signature · 1 line
2.20.1
#12 lore raw

PATCH v2 4/6 ASoC: topology: Check soc_tplg_add_route return value

In reply to Amadeusz Sławiński

Function soc_tplg_add_route can propagate error code from callback, we should check its return value and handle fail in correct way.

---

v2:
Added this patch

 sound/soc/soc-topology.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
sound/soc/soc-topology.c +3 -1
index 860bced933d6..78d10b9b725b 100644
@@ -1284,7 +1284,9 @@ static int soc_tplg_dapm_graph_elems_load(struct soc_tplg *tplg,
 		routes[i]->dobj.index = tplg->index;
 		list_add(&routes[i]->dobj.list, &tplg->comp->dobj_list);
 
-		soc_tplg_add_route(tplg, routes[i]);
+		ret = soc_tplg_add_route(tplg, routes[i]);
+		if (ret < 0)
+			break;
 
 		/* add route, but keep going if some fail */
 		snd_soc_dapm_add_routes(dapm, routes[i], 1);
Signature · 1 line
2.17.1
maintainer
#13 lore raw

Applied "ASoC: topology: Check soc_tplg_add_route return value" to the asoc tree

The patch

ASoC: topology: Check soc_tplg_add_route return value

has been applied to the asoc tree at

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

From 6856e887eae3efc0fe56899cb3f969fe063171c5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?=
<[email protected]>
Date: Fri, 27 Mar 2020 16:47:27 -0400
Subject: [PATCH] ASoC: topology: Check soc_tplg_add_route return value
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Function soc_tplg_add_route can propagate error code from callback, we should check its return value and handle fail in correct way.

---
 sound/soc/soc-topology.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
sound/soc/soc-topology.c +3 -1
index 3ada769cf823..cb43994089de 100644
@@ -1284,7 +1284,9 @@ static int soc_tplg_dapm_graph_elems_load(struct soc_tplg *tplg,
 		routes[i]->dobj.index = tplg->index;
 		list_add(&routes[i]->dobj.list, &tplg->comp->dobj_list);
 
-		soc_tplg_add_route(tplg, routes[i]);
+		ret = soc_tplg_add_route(tplg, routes[i]);
+		if (ret < 0)
+			break;
 
 		/* add route, but keep going if some fail */
 		snd_soc_dapm_add_routes(dapm, routes[i], 1);
Signature · 1 line
2.20.1
#14 lore raw

PATCH v2 5/6 ASoC: topology: Check return value of pcm_new_ver

In reply to Amadeusz Sławiński

Function pcm_new_ver can fail, so we should check it's return value and handle possible error.

---

v2:
Added this patch

 sound/soc/soc-topology.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
sound/soc/soc-topology.c +3 -1
index 78d10b9b725b..dade27588e51 100644
@@ -2135,7 +2135,9 @@ static int soc_tplg_pcm_elems_load(struct soc_tplg *tplg,
 			_pcm = pcm;
 		} else {
 			abi_match = false;
-			pcm_new_ver(tplg, pcm, &_pcm);
+			ret = pcm_new_ver(tplg, pcm, &_pcm);
+			if (ret < 0)
+				return ret;
 		}
 
 		/* create the FE DAIs and DAI links */
Signature · 1 line
2.17.1
maintainer
#15 lore raw

Applied "ASoC: topology: Check return value of pcm_new_ver" to the asoc tree

The patch

ASoC: topology: Check return value of pcm_new_ver

has been applied to the asoc tree at

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

From b3677fc3d68dd942c92de52f0bd9dd8b472a40e6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?=
<[email protected]>
Date: Fri, 27 Mar 2020 16:47:28 -0400
Subject: [PATCH] ASoC: topology: Check return value of pcm_new_ver
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Function pcm_new_ver can fail, so we should check it's return value and handle possible error.

---
 sound/soc/soc-topology.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
sound/soc/soc-topology.c +3 -1
index cb43994089de..818657b06799 100644
@@ -2135,7 +2135,9 @@ static int soc_tplg_pcm_elems_load(struct soc_tplg *tplg,
 			_pcm = pcm;
 		} else {
 			abi_match = false;
-			pcm_new_ver(tplg, pcm, &_pcm);
+			ret = pcm_new_ver(tplg, pcm, &_pcm);
+			if (ret < 0)
+				return ret;
 		}
 
 		/* create the FE DAIs and DAI links */
Signature · 1 line
2.20.1
#16 lore raw

PATCH v2 6/6 ASoC: topology: Check return value of soc_tplg_dai_config

In reply to Amadeusz Sławiński

Function soc_tplg_dai_config can fail, check for and handle possible failure.

---

v2:
Added this patch

 sound/soc/soc-topology.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
sound/soc/soc-topology.c +7 -2
index dade27588e51..9d6b266ccef8 100644
@@ -2524,7 +2524,7 @@ static int soc_tplg_dai_elems_load(struct soc_tplg *tplg,
 {
 	struct snd_soc_tplg_dai *dai;
 	int count;
-	int i;
+	int i, ret;
 
 	count = le32_to_cpu(hdr->count);
 
@@ -2539,7 +2539,12 @@ static int soc_tplg_dai_elems_load(struct soc_tplg *tplg,
 			return -EINVAL;
 		}
 
-		soc_tplg_dai_config(tplg, dai);
+		ret = soc_tplg_dai_config(tplg, dai);
+		if (ret < 0) {
+			dev_err(tplg->dev, "ASoC: failed to configure DAI\n");
+			return ret;
+		}
+
 		tplg->pos += (sizeof(*dai) + le32_to_cpu(dai->priv.size));
 	}
 
Signature · 1 line
2.17.1
maintainer
#17 lore raw

Applied "ASoC: topology: Check return value of soc_tplg_dai_config" to the asoc tree

The patch

ASoC: topology: Check return value of soc_tplg_dai_config

has been applied to the asoc tree at

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

From dd8e871d4e560eeb8d22af82dde91457ad835a63 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?=
<[email protected]>
Date: Fri, 27 Mar 2020 16:47:29 -0400
Subject: [PATCH] ASoC: topology: Check return value of soc_tplg_dai_config
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Function soc_tplg_dai_config can fail, check for and handle possible failure.

---
 sound/soc/soc-topology.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
sound/soc/soc-topology.c +7 -2
index 818657b06799..33e8d189ba2f 100644
@@ -2524,7 +2524,7 @@ static int soc_tplg_dai_elems_load(struct soc_tplg *tplg,
 {
 	struct snd_soc_tplg_dai *dai;
 	int count;
-	int i;
+	int i, ret;
 
 	count = le32_to_cpu(hdr->count);
 
@@ -2539,7 +2539,12 @@ static int soc_tplg_dai_elems_load(struct soc_tplg *tplg,
 			return -EINVAL;
 		}
 
-		soc_tplg_dai_config(tplg, dai);
+		ret = soc_tplg_dai_config(tplg, dai);
+		if (ret < 0) {
+			dev_err(tplg->dev, "ASoC: failed to configure DAI\n");
+			return ret;
+		}
+
 		tplg->pos += (sizeof(*dai) + le32_to_cpu(dai->priv.size));
 	}
 
Signature · 1 line
2.20.1
Reviewed-by Ranjani Sridharan on the v2 series
#18 lore raw
In reply to Amadeusz Sławiński
Quoted 10 lines · Amadeusz Sławiński

v1:
Check if kstrdup succeeded.

v2:
Remove unneeded freeing, which is performed in another place by dobj handlers.

Additionally for functions which have return status which was ignored, perform success checks and handle failures in appropriate way.

Quoted 7 lines

Amadeusz Sławiński (6):
ASoC: topology: Add missing memory checks
ASoC: topology: Check return value of soc_tplg_create_tlv
ASoC: topology: Check return value of soc_tplg_*_create
ASoC: topology: Check soc_tplg_add_route return value
ASoC: topology: Check return value of pcm_new_ver
ASoC: topology: Check return value of soc_tplg_dai_config

Thanks, Amadeusz. LGTM

Reviewed-by on the v2 series
#19 lore raw
In reply to Amadeusz Sławiński
Quoted 10 lines · Amadeusz Sławiński

v1:
Check if kstrdup succeeded.

v2:
Remove unneeded freeing, which is performed in another place by dobj handlers.

Additionally for functions which have return status which was ignored, perform success checks and handle failures in appropriate way.

Quoted 7 lines

Amadeusz Sławiński (6):
ASoC: topology: Add missing memory checks
ASoC: topology: Check return value of soc_tplg_create_tlv
ASoC: topology: Check return value of soc_tplg_*_create
ASoC: topology: Check soc_tplg_add_route return value
ASoC: topology: Check return value of pcm_new_ver
ASoC: topology: Check return value of soc_tplg_dai_config

Looks good to me

We probably want Ranjani to double-check this series though.

Quoted 4 lines
  sound/soc/soc-topology.c | 113 ++++++++++++++++++++++++++++++---------
  1 file changed, 88 insertions(+), 25 deletions(-)
#20 lore raw
Quoted 21 lines · Pierre-Louis Bossart
Quoted 10 lines · Amadeusz Sławiński

v1:
   Check if kstrdup succeeded.

v2:
   Remove unneeded freeing, which is performed in another place by dobj handlers.

   Additionally for functions which have return status which was ignored, perform success checks and handle failures in appropriate way.

Quoted 7 lines

Amadeusz Sławiński (6):
   ASoC: topology: Add missing memory checks
   ASoC: topology: Check return value of soc_tplg_create_tlv
   ASoC: topology: Check return value of soc_tplg_*_create
   ASoC: topology: Check soc_tplg_add_route return value
   ASoC: topology: Check return value of pcm_new_ver
   ASoC: topology: Check return value of soc_tplg_dai_config

Quoted 6 lines

Looks good to me

We probably want Ranjani to double-check this series though.

Hi Ranjani, can you take another look, I would like for this to get merged before I forget about it ;)

Quoted 4 lines
Quoted 4 lines

  sound/soc/soc-topology.c | 113 ++++++++++++++++++++++++++++++---------

  1 file changed, 88 insertions(+), 25 deletions(-)
#21 lore raw
Quoted 9 lines
Quoted 7 lines

Looks good to me

We probably want Ranjani to double-check this series though.

Hi Ranjani, can you take another look, I would like for this to get merged before I forget about it ;)

Ranjani provided her Reviewed-by tag on March 30 - likely our emails crossed.

#22 lore raw
Quoted 11 lines · Pierre-Louis Bossart
Quoted 9 lines
Quoted 7 lines

Looks good to me

We probably want Ranjani to double-check this series though.

Hi Ranjani, can you take another look, I would like for this to get merged before I forget about it ;)

Ranjani provided her Reviewed-by tag on March 30 - likely our emails crossed.

That's probably what happened, I only asked, because "double-check" comment above caused me to think it may need another look. Thanks, for confirming it's good.

#23 lore raw
Quoted 15 lines · Amadeusz Sławiński
Quoted 13 lines · Pierre-Louis Bossart
Quoted 12 lines
Quoted 8 lines

Looks good to me

Reviewed-by: Pierre-Louis Bossart <
[email protected]>

We probably want Ranjani to double-check this series though.

Hi Ranjani, can you take another look, I would like for this to get
merged before I forget about it ;)

Quoted 8 lines
Quoted 3 lines

Ranjani provided her Reviewed-by tag on March 30 - likely our emails
crossed.

That's probably what happened, I only asked, because "double-check" comment above caused me to think it may need another look. Thanks, for
confirming it's good.

Sorry for the confusion, Amadeusz/Pierre. Yes, I didnt want to reply and add noise as I had already reviewed the series.

Thanks,
Ranjani